--- Log opened Thu Jun 14 00:00:57 2018 00:08 < TheWild> and what the hell that Windows requires drivers for wifi adapter? On Ubuntu it works out of the box. 00:09 < darkmeson> afair, all recent versions of the sata spec require hotplugging, but I can't say I've ever had a problem doing it on any revision 00:09 < guideX> what does it mean when ntpq -p output (under the when column) has - 00:11 < jim> TheWild, as you might know, ubuntu inherits a lot of stuff from debian, including the way debian builds kernels, they're configured to compile everything possible as drivers, then they're packaged with the kernel image 00:12 < darkmeson> A COW filesystem might behave extremely bad if you do it, but you run that risk every time the kernel OOPSes or the power goes out anyway 00:13 < TheWild> but Windows? Sir, this is Windows. Windows was famous for having out of the box support for various crazy hardware. 00:13 < darkmeson> s/COW filesystem/mounted COW filesystem/ 00:13 < djph> TheWild: "Out of the box" 00:14 < djph> Windows is probably more famous for their rollercoaster release cycle, "wait til SP1", and "Embrace Extend Extinguish" 00:14 < rascul> get out of my box 00:14 < djph> rascul: funny, that's what she said last night... 00:14 < darkmeson> The Midori site says Debian has a packaged version, but I'm not seeing one in the testing repos even with "main contrib non-free". Am I missing something or is their information outdated? 00:15 < TheWild> oh, I remember funny issue when I was installing Windows 95. Installation was booted from CD. At some point of installation it needed to restart, but after restart it couldn't find drivers for CD-ROM, so it asked me to point to directory where drivers were, which were 00:15 < TheWild> on CD 00:16 < TheWild> (game is over. Format the disk, copy CD-ROM drivers to disk, boot installation again). 00:16 < darkmeson> I had that issue 00:17 < phogg> TheWild: you're supposed to copy the .inf to a floppy 00:17 < phogg> my favorite part of a Windows installer is the NT4 install routine where it gets to the "Detecting CPU architecture stage" 00:17 < bls> yeah, duh, everybody knew you always copied the cdrom drivers to the floppy, after moving highmem 00:18 < phogg> after a binary executable had been running for 5 minutes 00:18 < darkmeson> In fact, that's why I started installing DOS first, then copying over the installation directory and doing the installation from there 00:18 < triceratux> darkmeson: i can see it in the webmorrors http://yandex-asia.archive.parrotsec.org/debian/pool/main/m/midori/ 00:18 < phogg> darkmeson: if you were clever you did that. If you're clever you still do that with windows, only no DOS. 00:19 < Elladan> Windows is famous for coming with the computer when you buy it, with the right drivers installed already. Linux computers also have this property. 00:19 < darkmeson> I'm even more clever, because I haven't used Windows for more than basic testing in about a decade ;) 00:19 < n-iCe> Finally, my bluetooth headset is working, what a pain 00:20 < rascul> the best feature of win95 was that you could install linux instead 00:20 < domhnall> Foxboron: ok think im looking in right place now, how about burpsuite 1.7.34? 00:21 < Foxboron> domhnall: why did you move the conversation to ##linux? 00:21 < domhnall> oh shit 00:21 < domhnall> sorry 00:22 < domhnall> Foxboron: that was not intentional. 00:23 < Foxboron> domhnall: nps. But burpsuite has a restrictive license. so can't redistribute the binary 00:23 < Foxboron> so AUR 00:23 < domhnall> Ok. 00:23 < darkmeson> n-iCe: it was a lot easier before blueman went defunct 00:25 < darkmeson> Oh, it looks like someone must've resurrected it in the interim. Nice! 01:17 < phantomcircuit> the kernel seems to maintain all of these wireless regulatory rules 01:18 < phantomcircuit> they all seem insane, anybody know if there's a real reason for the kernel to enforce these rules? 01:18 < bls> phantomcircuit: if the kernel didn't, linux would be banned from supporting wireless cards in a lot of countries 01:19 < searedvandal> and that would be less than ideal 01:23 < sacules> i personally like my distros working with wifi out of the box 01:24 < phogg> I personally like to plug in a wire so my network isn't unreliable. 01:24 < searedvandal> same here. but sometimes that wireless is needed 01:25 < sacules> mostly if it's for somebody just getting into gnu/linux 01:25 < phogg> Speaking topically the alternative is to have what Intel used to provide: a proprietary userspace "management" daemon which enforces regulatory compliance rules. It was not fun and didn't work well. 01:28 < rjdp9736_> an interesting argument between Linus and Andy Tanenbaum https://groups.google.com/forum/m/#!topic/comp.os.minix/wlhw16QWltI%5B1-25%5D 01:29 < koala_man> a classic 01:29 < dgs> that the one from the 90s? 01:29 < rjdp9736_> yes! 01:29 < koala_man> "linux is obsolete", yes 01:29 < dgs> is tanebaum even still around? 01:29 < rjdp9736_> haha 01:30 < rjdp9736_> dgs, yes still whining 01:30 < dgs> i can see how he'd be a bit pissy after some kid designed a homebrew that wasn't as good as his "proper" one 01:30 < dgs> and that kids os went on to pretty much take over the world 01:34 < lupine> good old history: > As a result of my occupation, I think I know a bit about where operating are going in the next decade or so 01:34 < degva> hi all, is there a way to resend a received package to another server? meaning, say I have two servers (B and C), and every package I receive to B, C will also receive it. 01:34 < lupine> degva: which packaging system? apt, yum, etc? 01:35 < phogg> degva: the short answer is "of course" 01:35 < phogg> lupine: Oh? Care to share? Are they getting replaced by our new AI overlords? 01:35 < lupine> typically it will be easier to have a third machine locally that both A and B pull from 01:35 < degva> lupine: package received from another server, as like a ping 01:36 < lupine> ohhh, you mean packet? 01:36 < phogg> package != packet 01:36 < degva> packet, yes, sorry 01:36 < phogg> they're nearly the same thing in English but we're speaking computer here. 01:36 < lupine> yeah, you can use various services to do that, or iptables 01:36 < degva> haha 01:36 < lupine> I often use rinetd or socat if I just want to send the packets on to the second server, and am uninterested in processing them locally 01:36 < degva> oh right, iptables! 01:37 < degva> lupine: thank you :) 01:37 < lupine> np 01:37 < lupine> just don't ask me to draw up the iptables rules. I've not done that for ages :D 01:38 < degva> haha well, I've learned it on school, but I get the concept 01:45 < truthadjustr> when doing `cp` how do we tell if some bytes are still in cache, ie not yet written to disk? 01:46 < phogg> truthadjustr: you guess, or you run sync. 01:47 < truthadjustr> phogg: no /proc or /sys/ tells? 01:47 < bls> the overhead of even instrumenting something like that would be awful 01:48 < phogg> truthadjustr: there are and that's going to be slightly more successful 01:48 < phogg> truthadjustr: all you can do is flush *everything*, though. ANd *nothing* forces the disk to also flush its cache 01:49 < truthadjustr> phogg: already tried `sync`, still mdsum not matching.. 01:49 < bls> what problem are you really trying to solve? 01:49 < truthadjustr> happens lots of times when i do `cp -av` .. the way i fixed it is cp individually each file 01:50 < phogg> truthadjustr: echo 3 > /proc/sys/vm/drop_caches ; blockdev --flushbufs /dev/your-device # there are some other things you can try, too 01:50 < truthadjustr> bls: `cp -av /mnt/stuff/ /home/john/stuff` produces md5sum mismatch for huge files (25G) 01:50 < phogg> truthadjustr: that should never, ever happen 01:51 < phogg> the userspace view of the filesystem is *consistent* even if the buffers are not flushed 01:51 < truthadjustr> after i found out a file is corrupted, i cp individually.. to fix it. 01:51 < bls> right, userspace has no clue where it's reading from 01:51 < phogg> truthadjustr: if that happens to you then you have a hardware problem. Write errors are occurring and nobody knows it. 01:51 < truthadjustr> event `sync` does not help 01:51 < phogg> bits being flipped on disk, or whatever 01:52 < bls> yeah, start triaging your drives and FSs, then memory 01:52 < phogg> truthadjustr: re-copying only solves it by luckily not failing that time 01:52 < truthadjustr> phogg: don't think so. If i cp invidivually, it is ok 01:52 < phogg> given that it happens to you for mass recursive copies I'll bet heat is a factor 01:52 < phogg> truthadjustr: Trust me, you *CANNOT* get your results without a hardware error. 01:53 < bls> cp/the FS/VFS/block layers don't magically slip into "best effort" mode just because it's been given more than one file to copy 01:54 < phogg> truthadjustr: have you looked at a SMART report lately? 01:54 < phogg> It may not be the disk, but it may be the disk. 01:54 < truthadjustr> phogg: will look into it.. 01:55 < truthadjustr> phogg: the source is a mounted ntfs usb drive 01:55 < bls> heh, even more suspect 01:55 < truthadjustr> i hope it's not my thinkpad nix 01:55 < phogg> truthadjustr: you may also want to try the program badblocks 01:57 < ryouma> when i mv a file from dir a to dir b, the ctime gets updated. i did not expect this. has this always been the case? shouldn't mv that is not across fses preserve all 4 file times (abcm)? 01:58 < phogg> ryouma: ctime is change time, which is change in *metadata*, contrast mtime which is change in file bytes. 01:58 < phogg> ryouma: the file's parent dir is part of its metadata 01:59 < phogg> most filesystems don't support btime in any case 01:59 < ryouma> weird. where is ctime stored? inode or dir? and why is that metadata? is grandparent metadata? 01:59 < phogg> ryouma: inode 01:59 < ryouma> 16:59 most filesystems don't support btime in any case -- ext4 does, even as of a while back. but the kernel does not until recently. 02:00 < phogg> ryouma: mv is basically just link+unlink. Adding link references to the inode bumps ctime. At least that's how I understand it. 02:00 < ryouma> the inode links to the dir? idgi. 02:00 < ryouma> oh i see the adding of the link does it. but i thought it added it to the dir. 02:01 < phogg> ryouma: pretty sure the file's inode still gets updated. I am not a kernel hacker, but I guess a quick trip through the vfs source would make this clear. 02:04 < ryouma> i look at c maybe once every 15 years 02:06 < phantomcircuit> bls, i guess, but the NO-IR flags seems to be beyond what the fcc actually requires 02:07 < ryouma> ctime annoys me. it is almost useful. but basically not for almost anything. 02:07 < ryouma> inb4 hence btime 02:19 < jamiejackson> i folks. i'm automating a vpn connect with expect, but it's exiting immediately, before it waits for me to authenticate with mfa. it should wait for the highlighted part: https://gist.github.com/jamiejackson/caa42fe7c36b0373e572fbf74d463a8d#file-expect-L18 02:20 < jamiejackson> happen to know what i'm doing wrong? 02:21 < jamiejackson> i forgot about autoexpect. i'll try that first 02:30 < jamiejackson> autoexpect gave me an ending of `expect eof`, but i think that's premature. i think i want to ignore that (if that makes sense). possible? 02:32 < jamiejackson> i'm also curious why this channel is so quiet. it's usually a ridiculous torrent. 02:33 < cmj> lean back 02:34 < cmj> you can also try ##networking 02:36 < darkmeson> jamiejackson: afair, it's expecting eof because the program hung up 02:37 < jamiejackson> interactively, it waits for me to deal with mfa on my cell phone, then it continues to "STATUS::Login succeed", etc. any clue what the difference might be, darkmeson 02:39 < darkmeson> That was mostly gibberish to me, and I don't see anything in the backbuffer offering further explanation 02:40 < BenderRodriguez> How cryptographically secure are LUKS encrypted file systems 02:40 < darkmeson> If you're running autoexpect to create a profile, it's obviously going to be entirely interactive because it has to see how you want it to answer at the different prompts 02:41 < darkmeson> afair, it will end and 'expect eof' if the program you were running exited or you typed 'exit' 02:41 < koala_man> BenderRodriguez: highly 02:42 < BenderRodriguez> koala_man: I've been doing some reading and I found a few that reference GPU based brute force attacks that can comprimise LUKS passphrases 02:42 < BenderRodriguez> in light of this, is it a useless endeavour to use it? 02:43 < jamiejackson> there's something different when running this with expect/autoexpect than when i run it as a human, darkmeson. that's what i meant. i don't know why [auto]expect bails out at that point. 02:43 < darkmeson> If that's not what you want, then run autoexpect by itself (no arguments), run everything you want to run, THEN type exit and insert your own conditional logic if necessary 02:45 < kerframil> BenderRodriguez: no. in short, employ a strong passphrase (obviously) and argon2 as your key derivation function. 02:45 < koala_man> BenderRodriguez: all disk encryption systems can be broken with brute force attacks, the point is that it takes forever 02:45 < darkmeson> BenderRodriguez: look at the tail end of cryptsetup --help. It uses fairly strong aes crypto with fairly strong sha hashing by default 02:45 < jamiejackson> darkmeson. thanks. trying... 02:45 < koala_man> what's the current default? 02:45 < kerframil> BenderRodriguez: argon2 is designed so as to render GPU attacks slow 02:46 < kerframil> koala_man: it's argon2i for LUKS2 02:46 < koala_man> nice 02:46 < kerframil> koala_man: and pbkdf2 for LUKS1. you can you check your defaults with cryptsetup --help. 02:47 < darkmeson> BenderRodriguez: If you want to worry about something, worry about Android 7+'s switch to file-based encryption 02:47 < darkmeson> It in theory has the same serious flaws as encfs and friends 02:50 < ryouma> what flaws? 02:50 < ryouma> i mean i know it's not widely used 02:50 < ryouma> does encfs have i mean 02:51 < darkmeson> https://defuse.ca/audits/encfs.htm 02:51 < ryouma> is argon2 default? 02:53 < darkmeson> There's also a big, fat warning with further reading presented when you install encfs on a Debian-like distro, I believe 02:55 < darkmeson> ryouma: as I said earlier, check cryptsetup --help 02:55 < darkmeson> As far as defaults in general go, the sooner you stop relying on them, the safer you'll be 03:09 < nekoseam> https://www.reddit.com/r/linux/comments/8lx1rm/ththanks_aptitude/ How can people think they can fool anybody like this 03:10 < nekoseam> Ubuntu user claims that trying to install steam removed a ton of programs. When in reality it's literally the output when you run aptitude autoremove --dry-run 03:11 < mgolisch> no idea 03:11 < mgolisch> people on reddit are weird 03:11 < ALowther> I created a partition, then I created a file system for the partition. Then I deleted the partition, then recreated it and the file system was still there. How do I delete the file system with the partition? 03:12 < koala_man> ALowther: you can't delete it, you can only overwrite it 03:13 < ALowther> koala_man: So if I create a new filesystem, it will for sure overwrite it? 03:13 < ALowther> (On the same partition) 03:13 < koala_man> ALowther: yes 03:13 < ALowther> koala_man: Okay, thanks. fdisk just says there is another file system there, but doesn't say it will overwrite it, so I wasn't sure. 03:13 < koala_man> but it won't go through and overwrite every single bit on the partition, only what the fs needs for its own bookkeeping 03:26 < nekoseam> sigh 03:29 < truthadjustr> i think, the cause of the `cp` error from a mounted ntfs usb drive is: `mount /dev/sdb1 /mnt` 03:30 < truthadjustr> i did `ntfs-3g /dev/sdb1 /mnt` and the cp is now accurate... 03:31 < truthadjustr> i thought, using the `mount` was reliable. I guess, `ntfs-3g` should be used for ntfs 03:39 < kerframil> mount -t ntfs-3g 03:40 < lnnb> ewwww 03:40 < dannylee> https://www.youtube.com/watch?v=K5jvUXij7nU 03:40 < lnnb> hello daniel 03:40 < dannylee> linux is really great 03:41 < dubhdara> what flavor are you running of linux 03:41 < dannylee> openSuSe 03:41 < dubhdara> manjaro budgie here 03:41 < dannylee> c0000l 03:42 < dubhdara> Suse is good too 03:42 < dannylee> i use openSuSe for over ten years 03:42 < dannylee> i like it 03:43 < dubhdara> which version leap or tumbleweed 03:43 < dannylee> tumle weed 03:43 < dubhdara> the latest and greatest 03:43 < dannylee> that way i have gnome and kde an lxde 03:44 < dannylee> i just install every thing 03:44 < dubhdara> well if it works 03:44 < dannylee> and it works 03:44 < dannylee> i have a graphical root tooop 03:45 < dannylee> open SuSE is really great..no mistakes 03:45 < dannylee> i like Gnome 03:46 < dubhdara> i tried it a while back but the leap version it was to stuffy 03:46 < Dreaman> open suse kde 03:46 < Dreaman> not base gnome 03:46 < dubhdara> might have to try the tumbleweed branch 03:46 < dannylee> i all ready did that 03:47 < Dreaman> fedora base 03:47 < dannylee> i open KDE just for the settings 03:47 < dannylee> ok i still have fedora on my other machine 03:47 < dannylee> 27 03:48 < dannylee> i still love redhat 03:48 < Dreaman> i love debian base systems 03:48 < dubhdara> fedora seems like it went downhill after removing core 03:48 < Dreaman> easy 03:48 < dannylee> ok its write for you 03:48 < dubhdara> i love the CentOS though 03:48 < dannylee> redhat 03:48 < dannylee> but no software 03:49 < Dreaman> nikolov@ubuntu-ivan:~$ inxi -F 03:49 < Dreaman> System: Host: ubuntu-ivan Kernel: 4.17.1-041701-generic x86_64 bits: 64 03:49 < Dreaman> Desktop: Gnome 3.28.1 Distro: Ubuntu 18.04 LTS 03:49 < Dreaman> simple 03:49 < dannylee> i have a copy of gnome ubuntu 03:49 < dannylee> as a back up 03:50 < dannylee> if i get piss i might install it 03:50 < dannylee> suse is really the best 03:51 < mgolisch> no idea, not used that in atleast a decade 03:54 < dubhdara> whoami 03:54 < lnnb> u r dubhdara 03:55 < ryouma> -> ##philosophy 03:56 < ansraliant> TIL there is a philosophy channel 03:56 < ansraliant> and they are talking about civ6. Seems about right 04:00 < fundies> yo what linus' cell i need help fixing linux 04:01 < esselfe> 418-555-3264 04:01 < Pentode> 867-5309 04:01 < lnnb> niieeiiieeiien 04:01 < nchambers> jenni jenni 04:04 < updated> fundies: if you mention his name three times in front of the mirror he'll show up 04:06 < Pentode> that just might be more frightening than bloody mary 04:06 < fundies> https://pastebin.com/raw/rLVYY3gc can anyone here help me debug this? I tried asking #radeon but its been fairly silient for hours 04:06 < fundies> those erors fully lock up my computer to the point it cant even reboot 04:07 < fundies> yet strangely I can ssh 04:07 < fundies> I've seen some posts about it on tracker but no real fixes 04:10 < ansraliant> I'm no expect on amd cards and stuff. But it seems there is some problems communicating with it. I know this might sound cliche, but have you tried reinstalling the driver? 04:11 < fundies> ansraliant, I've upgraded to 4.17 to test it so basically 04:12 < fundies> ansraliant, this only happens under certain conditions 04:12 < fundies> i cant really narrow them but i can trigger it reliably 04:22 < ollien> I might be going nuts, but I have authorized_keys with permissions 600, and .ssh with permissions 700, and I still can't ssh in with my public key. I'm not really sure what else to try. Can someone help me out? 04:23 < kerframil> ollien: use ssh -v, and read the log data from sshd 04:24 < ollien> kerframil: Yeah, I have. It offers the public key, then authentication fails 04:25 < kerframil> ollien: so you've read the log messages produced by sshd as well? 04:25 < ansraliant> fundies: the only error logs you have is that dmesg? 04:25 < ollien> ah sorry, misread your message. Looking into sshd logs now 04:26 < Elladan> ollien, how about the permissions on your home dir. 04:26 < fundies> ansraliant, correct 04:26 < kerframil> ollien: the log facility for these messages is typically auth or authpriv 04:26 < ollien> Elladan: Well, this is the funny thing. The .ssh folder is symlinked to another location on disk 04:26 < ollien> must that containing directory also have 700 perms? 04:27 < Elladan> ollien, I'm not sure that ssh will be willing to follow symlinks at all 04:27 < Elladan> ollien, what permissions does it have? 04:27 < Elladan> And also your home dir? 04:27 < ollien> Elladan: the home dir is 744 04:28 < ollien> the .ssh dir is 700 04:28 < ansraliant> fundies: I found this https://bugs.freedesktop.org/show_bug.cgi?id=98874. Maybe you already read it. But if not, maybe it helps 04:28 < Elladan> ollien, try 755 for home 04:28 < kerframil> ollien: 744 for a dir is very dubious. yes, try 755. and read the logs still. 04:28 < ollien> err, sorry, 755 is a typo 04:28 < ollien> *744 was a typo 04:29 < ollien> kerframil: yep. sorry, got sidetracked. Looks like it doesn't like the perms of the directory holding the symlink pointer 04:29 < fundies> ansraliant, ive read evever search result for those errors 04:29 < ollien> gonna try 700ing that 04:29 < fundies> thanks though 04:29 < kerframil> ollien: interesting. what does it report exactly? 04:29 < Elladan> ollien, also you didn't say what the permissions are on your key. 04:29 < ollien> Elladan: locally? 600 04:29 < ollien> kerframil: Jun 14 02:29:48 planet-express-ship sshd[1527]: Authentication refused: bad ownership or modes for directory /home/nick 04:30 < ollien> which ... that's frustrating 04:30 < Elladan> Is it owned by you? :-) 04:30 < ollien> well here's the peculiar thing 04:30 < ollien> I'm trying to ssh a user 'git', and symlink the .ssh directory to a .ssh directory within /home/nick 04:30 < ollien> but I imagine that might be futile 04:30 < ollien> given what I'm seeing now 04:30 < ollien> so yes, each /home/ directory is owned by their respective users 04:31 < ollien> but git does not own /home/nick, no 04:31 < kerframil> ollien: more the point, if the user git doesn't own .ssh, then a mode of 700 renders it non-traversible 04:31 < kerframil> more to* 04:31 < ollien> kerframil: git does own .ssh 04:31 < ollien> and its parent directory 04:31 < kerframil> ollien: oh, I see 04:32 < kerframil> ollien: it might help to clarify if you pastebin: stat /{,home,home/nick,home/nick/.ssh} 04:33 < ollien> kerframil: https://gist.github.com/ollien/43b12ad51d70822e1d58f5ea1d97945a 04:36 < kerframil> ollien: it's important to walk the entire directory chain (gogs and gogs/git are missing). 04:37 < oryois> Today Challenge : It seems to be a complicated cipher, can you break it? https://s3-eu-west-1.amazonaws.com/hubchallenges/crypto/genfei.zip , First Solver prize : licenses for (Burpsuite,ida7pro,binaryninja,avast antivirus,200$ amazon gift card) 04:37 < kerframil> ollieh: anyway, my theory is that sshd will complain if any directory in said chain is not owned by the user that you are authenticating as 04:37 < kerframil> sorry, that was for ollien 04:37 < ollien> kerframil: yeah, that was my theory 04:37 < ollien> https://gist.github.com/ollien/9d4f07d26fd723bff7bf83438e4f33ee 04:38 < ollien> oh well. I'll find some way to do this 04:38 < ollien> ...tomorrow. The sleepiness is def impacting my thinking about this 04:38 < kerframil> ollien: you could disable StrictModes just for the git user. I would recommend any other approach that does not entail disabling StrictModes, however. 04:39 < kerframil> ollien: man sshd_config 04:39 < ollien> kerframil: yeah, I have some ideas. It's just not the most productive time of night :) 04:39 < ollien> thanks :) 04:41 < kerframil> ollien: actually, not quite "any directory" but you get the idea. anything from the designated home directory downwards. 05:24 < darkmeson> fundies: I don't have issues with them on my radeon systems, but a server system that has a generic gpu that uses the ast module can't cope with 4.15+ either 05:26 < darkmeson> It's fairly common in supermicro systems, so maybe the combination of the two will yield better search results 05:37 < sinatrablue> hmm now that i have this linux box setup, what should i do with it? 05:39 < Bashing-om> sinatrablue: linux is a mighty big hammer, how you swing it is up to you :P 05:40 < darkmeson> The sad part is, the OS doesn't really matter much anymore 05:41 < darkmeson> They're all pwned in the same ways by the same web browsers having the same lax security and poor isolation models 05:41 < fundies> darkmeson, this only happens with games as far as i can tell 05:42 < fundies> darkmeson, and i need amd dc stuffs to run thisc game 05:42 < Elladan> Android has the right idea about isolation IMO. 05:43 < Elladan> Nobody cares about users if you're the only user, but you don't want your web browser to be able to access your password manager. 05:43 < sinatrablue> Bashing-om: this box has more than enough power for a webserver, but probably not a good idea on a residential connection 05:43 < darkmeson> fundies: Then I'd try to rule out thermal issues if I were you. Set up lm-sensors if you haven't and see what 'sensors' has to say 05:43 < Elladan> sinatrablue, what do you want to do? 05:44 < fundies> darkmeson, I can trigger it from certain effects in game 05:44 < fundies> its bug in code for sure 05:44 < sinatrablue> Elladan: not sure, i have this 1800x computer but not much to do with it 05:44 < sinatrablue> looking for things to do 05:44 < darkmeson> fundies: you said you could keep ssh access, so you should be able to do 'watch sensors' in a ssh session 05:45 < fundies> darkmeson, i know its not a temp issue 05:45 < darkmeson> Elladan: Android's APP isolation is good, but it's NETWORK isolation is beyond horrible 05:46 < Elladan> darkmeson, I didn't say it was the right implementation :-D 05:46 < fundies> darkmeson, I can boot up game press button to render the 1 effect and itll crash 05:46 < fundies> or i can play for hours without that effect 05:47 < darkmeson> They don't want you to have root, but also don't want you to have access to the firewall features every other OS has, and they don't even fence apps off from network access except via a proxy when a proxy is set 05:47 < Dagmar> You don't need access to the firewall 05:47 < Rainb> help 05:47 < Dagmar> 99% of the users will just _screw everything up_ anyway 05:47 < Rainb> my enabled systemctl service never runs at startup 05:47 < Rainb> https://pastebin.com/SK2p7sRx 05:47 < Rainb> why is this? 05:48 < Rainb> at boot-time* 05:48 < darkmeson> fundies: It might be a broken GL stack. Is this a wine game? 05:48 < Dagmar> ...and if you're running an app that's going to talk to hosts on the network against your will, it's time to start questioning why the hell you're downloading sketchy APKs from untrustworthy sources 05:49 < fundies> darkmeson, yes 05:49 < darkmeson> Dagmar: even F-Droid apps help themselves to the network without asking 05:49 < Dagmar> So what 05:49 < Dagmar> If you installed an app, and it uses the network, you either know this and are fine with it, or you have made a terrible mistake 05:50 < fundies> darkmeson, its overwatch specifically 05:50 < darkmeson> The user is supposed to be able to have absolute control over their own device. It's silly to have to work around such an obvious and contrived flaw with VPN trickery and so forth 05:50 < fundies> darkmeson, others can run it without theses crashes just have fps dips 05:51 < Elladan> Android isn't built around the idea that the user owns their device. 05:51 < fundies> darkmeson, this bug seems related to amgdgpu driver only 05:51 < Dagmar> Incorrect 05:52 < Dagmar> Android is built around the idea that the user owns their device, but that they need the operating system to help them with technical decisions 05:52 < darkmeson> Elladan: If they don't people disabling network functionality to get around the ads, they need to make the apps check for network access and refuse to work if they can't connect. But I'm sure we both know why it is the way it is. 05:53 < Dagmar> It's the reason the permissions framework keeps getting more and more granular, and doesn't show you things like "snmp_mem_t (a)llow or (d)eny?" 05:53 < Dagmar> It shows you things like "requests access to the camera" and "requests access to your browser history" 05:53 < Elladan> darkmeson, there are various reasons why it works that way. It isn't just the ads: you don't have to go far to find people who argue that users should never have root or bootloader access to any device. 05:53 < Dagmar> Clear, human-readable concepts that are easy to grasp. 05:53 < Elladan> However, this is not really relevant to the question of what a good security model would be like. 05:54 < Dagmar> Users aren't allowed access to an unrestricted security context under normal operating conditions, because if the users can get at that, someone's going to find a way to trick the users into giving their program access to that unrestricted context... at which point the game is over 05:54 < Elladan> I'm sure we'd agree that an ideal setup would give you effective app isolation, and also good tools for detecting and controlling network access. 05:54 < darkmeson> fundies: Presumably it's running under Wine, and Wine has a pretty good debugging framework you might be able to make use of. Do a web search for WINEDEBUG= and/or inquire in #winehq. They should be able to help you out (although it might be a few hours since it's late for most of them) 05:55 < Elladan> Dagmar, that's... not the real reason. 05:55 < Dagmar> Yes, that is the real reason 05:55 < Elladan> It's not, and I'm sure you know that. 05:55 < Dagmar> It *is*. 05:55 < Dagmar> THe moment some malware gains the ability to write to the flash storage directly, you can't trust the device anymore 05:56 < Dagmar> You're basically f**ked at that point 05:56 < fundies> darkmeson, i asked there and they pointed me towards my gou driver 05:56 < fundies> gpu* 05:56 < fundies> darkmeson, i think its their stance that no matter how bad their code is it shouldnt crash the driver 05:56 < lnnb> do i have malware if some weird cloud backup service starts every time i mtp an audio file to my phone? 05:56 < fundies> :P 05:56 < Dagmar> Just the same as when some worm gains root access on your server, you are now _done_ and the only safe way to restore verity to the system is to _restore from backups_ 05:57 < darkmeson> Elladan: they intentionally removed the inet permission without any rhyme or reason given, really. That shows EXACTLY what they after. Data exfiltration to the max 05:57 < darkmeson> Steal anything that isn't bolted down, and make sure not to provide even the most basic OS constructs that might get in the way of it 05:57 < kopper> Dagmar: Provided that your backups don't have the same worm 05:57 < Dagmar> ..but at least on a PC you can pop the top, boot trusted media, and be at least somewhat assured that no one's started modifying EEPROM BIOS images for the malware on a widespread basis yet 05:57 < Dagmar> You do NOT have that same flexibility on a phone 05:58 < Dagmar> Users brick their s**t all the time as it is just trying to sideload APKs 05:58 < Dagmar> Heck I've personally seen two pieces of malware (CHinese app stores, go figure) that would deliberately erase the flash if you tried to remove them 05:59 < Dagmar> I.e., a _very_ bricked phone 05:59 < darkmeson> fundies: well, rightfully, if it IS the driver crashing, the driver has bugs. The right WINEDEBUG= might help you figure out if it's the driver or the lib stack though 06:01 < darkmeson> Alternatively, some distros have wine packages and 'wine-development'. Switching between the two might also help determine the nature of the bug 06:01 < Dagmar> One of those two devices was restorable because the chipset actually has a not-quite-documented method of booting off the microSD card 06:01 < Dagmar> The other one had to go back to the factory 06:03 < darkmeson> It's possible with a lot of devices if you know which connections to short and whatnot, but therein lies the problem 06:04 < Dagmar> Dude 06:04 < Dagmar> When was the last time you opened a cell phone? 06:04 < Dagmar> This is no longer something one can work on with a 15w soldering pencil 06:04 < darkmeson> It's rarely documented and very time-consuming to track down, even for someone with the right skill set 06:04 < Dagmar> I have to use a freakin' jeweler's loupe for these things on the regular 06:04 < Elladan> It's not particularly difficult to build a secure reflash facility into a device, if you actually cared about these sorts of problems. 06:05 < darkmeson> Even then they're liable to end up with a pile of bricks before they get it right, and it's just not worth it 06:05 < Dagmar> ...and it's just WILDLY beyond reasonable to expect random users to open their $800 phone and start cutting pins and so forth 06:05 < Elladan> However, it's basically irrelevant since hostile APKs don't have much trouble getting root regardless through local privilege escalation bugs. 06:06 < Dagmar> Elladan: Really. Then explain why it's only now there with Android 7 devices 06:06 < Dagmar> ...and why it's only _recently_ come about for PCs in the form of SecureBoot and UEFI 06:06 < Elladan> The point of not giving the user access to the device is all about walled gardens and "trusted" software e.g. DRM and related stuff, that is to prevent the user from exfiltrating their own data. 06:06 < darkmeson> There's no cutting really, just a shorting here and there 06:07 < Dagmar> Bullshit. 06:07 < Dagmar> Users have the access they need to use the device to it's fullest. Advanced users can get away with replacing the OS with some third-party thing like LineageOS. 06:08 < darkmeson> Elladan: they're now baking the official Facebook app into roms and you can't disable it without using adb 06:08 < Dagmar> The vast majority of these users can't even flash an image onto a thumbdrive, let alone pick the right firmware for their device 06:08 < Triffid_Hunter> Dagmar: I've had cyanogen (now lineage) on every smartphone I've ever owned.. stock roms never give me the options I want 06:08 < darkmeson> If you're keeping track, that means there's not just one, but two different frameworks that have been caught red-handed stealing location data 24/7, among other things 06:09 < darkmeson> And they can't be disabled without special know-how. How is that in the user's best interests, exactly? 06:09 < Triffid_Hunter> I always keep an eye on the location icon in my status bar, if it's on and I don't know why, I start force stopping stuff until it goes away 06:10 < Dagmar> Triffid_Hunter: I've been with sprint for some time, and their "special sauce" is horrible. I'm actually about to go back to using my S5 because this S7 can't be properly unlocked for a new image 06:10 < Dagmar> It actually runs worse than the S5 06:10 < Elladan> Triffid_Hunter, imagine an app that only turned on location when you weren't looking. 06:10 < Dagmar> darkmeson: What "special know how" does it take to touch the GPS icon on the notification tray to turn it off? 06:10 < Dagmar> That switch can not be overridden. 06:11 < kirk781> The official Facebook app is cancer 06:11 < Triffid_Hunter> Dagmar: I'm gonna get a xiaomi mix2 soon.. it has /all/ the radio bands and the lineage folks tell me it works great 06:11 < Dagmar> kirk781: And fire is hot 06:11 < kirk781> I can disable it on my phone but not remove is directly 06:11 < T-Rog> The Arch wiki's guide on getting joysticks working is pretty confusing. Can anyone help me out? 06:11 < Dagmar> Triffid_Hunter: I've considered it, but... China 06:11 < dgs> Dagmar: for what it's worth, I've recently gone back to my s5 after dropping my s7.... they are still really good phones 06:11 < kirk781> For some reason, Samsung thought integrating Flipboard will also be a good thing 06:12 < Dagmar> Yeah, I'm more than slightly disgusted that substantially better hardware performs worse than a model two generations old, just because of vendor crap and TouchWiz overcompensating 06:12 < darkmeson> Dagmar: any time it's enabled, those frameworks are logging and reporting back 06:13 < Dagmar> darkmeson: Prove it 06:13 < darkmeson> btw, in the off chance this is news to anyone here. https://www.wired.com/2013/03/anonymous-phone-location-data/ 06:13 < Dagmar> darkmeson: I have spent pretty much the last two years crawling up Androids sphincter all the way through to the mouth 06:14 < kirk781> Why do companies spend so much resources on developing 'extra' apps for their phones when they know that all it would do is delay actual OS upgrades in the future? 06:14 < Dagmar> This on top of 20+ years of using Slackware and several other Linux distros 06:14 < kirk781> Dagmar, what's your favorite distro? 06:14 < Dagmar> I decided it was time to see where Android had gotten to, so I got up to speed properly 06:14 < fundies> darkmeson, i logged the mesa from run to crash n its 40mb text file lol 06:14 < kirk781> And what distro caused you to pull your hairs off your head? 06:14 < Dagmar> kirk781: Whichever one isn't pissing me off at the moment. 06:14 < fundies> darkmeson, heres tail of it https://pastebin.com/raw/6nte2a2M 06:15 < Dagmar> There's distros that are good for some things and not others 06:15 < darkmeson> fundies: you did WINEDEBUG=-all didn't you? :) 06:15 < fundies> darkmeson, nah 06:15 < darkmeson> Dagmar: you can find it for yourself with a simple web search 06:15 < fundies> darkmeson, i did mesagl=dump 06:16 < Dagmar> darkmeson: I can also find articles about mind control sattelites and detailed lists of the chemicals they put in jet fuel to pacify the populace 06:16 < Dagmar> It doesn't make these things real 06:16 < ayecee> that's what they want you to think 06:16 < Dagmar> heh 06:17 < Dagmar> Go spend just six months tearing apart an AOSP build and a Lineage build, until you can reliably swap out the kernel and init 06:17 < Dagmar> Then come tell us how much automatic logging you found 06:18 < Dagmar> All the apps that log or report GPS information _say so outright_ 06:18 < Dagmar> Just enabling GPS doesn't actually do *anything* 06:18 < Dagmar> That chip doesn't get woken up until it's actually being queried for location data 06:18 < Dagmar> ...or at least, that part of the chipset 06:19 < Dagmar> That information isn't automatically logged anywhere either unless the app requesting it has set something up to do so 06:19 < poutine> A lot of apps have valid uses for GPS but have been shown to abuse it outside of its valid use 06:19 < Dagmar> Some stuff _might_ get leaked into logcat, but that's about as reliable as claiming that people are using the dmesg ringbuffer to spy on you 06:19 < darkmeson> Dagmar: baiting doesn't work on me 06:20 < Dagmar> darkmeson: Apparently facts don't work on you, either 06:20 < Dagmar> Go learn some Java and write some apps so you'll learn how these mechanisms actually work 06:20 < darkmeson> either you care to know, or you stick your head in the sand and pretend it doesn't happen 06:20 < darkmeson> you'll probably be fine either way, unless you become a someone 06:20 < Dagmar> I have stuck my head so far up Lineage's butt I could see what it had for lunch 06:20 < Dagmar> I am not a man who sticks his head in sand 06:20 < Triffid_Hunter> Dagmar: heh I live in china.. fwiw xiaomi is trying really hard to be a good company, but there's plenty of reasons I choose phones I can put lineage on and they're not unique to this country :P 06:21 < fhdrin> 06:21 < Dagmar> Triffid_Hunter: Ah okay. If you're in China I guess there's no point in worrying too much about whatever games the gov't might get up to. They'll have plenty of avenues other than just the phone 06:21 < darkmeson> Dagmar: Or maybe YOU should gain root, install some Xposed modules, and see what your apps are REALLY doing 06:21 < Dagmar> Are you slow? 06:22 < ayecee> no need for that 06:22 < Dagmar> Dude, my S5 runs a build of Lineage that was pretty heavily modified, by _me_ 06:22 < Dagmar> I am not joking when I say I have looked into this stuff _in detail_ 06:22 < ayecee> it's reasonable to doubt internet credentials 06:22 < darkmeson> I can guarantee you'll be shocked at the lengths many common apps go to to exfiltrate your data, even after you've denied those permissions you think save you 06:22 < Dagmar> I've spent 20+ dealing with Slackware and having to build most of the fancy stuff myself, including replacing and debugging kernels and practically every binary on the system 06:23 < Dagmar> Android was _not_ some great stretch 06:23 < Dagmar> I wanted to know how it worked, so I opened it up and looked inside 06:24 < Dagmar> Now i've got detailed reasons to chew on some Samsung people if I ever get ahold of them, and I already knew Sprint were idiots 06:24 * darkmeson really hates that kids get summer off 06:24 < Dagmar> Their "connection optimizer" app is literally snake oil 06:25 < ayecee> well, figuratively 06:25 < Dagmar> ayecee: No, the new OED listing. ;) 06:25 < Dagmar> It can _optionally_ report back when the signal goes to hell, but it actually doesn't 06:26 < ayecee> the one where literally doesn't mean literally anymore 06:26 < Dagmar> Precisely 06:26 < Dagmar> The app just doesn't _do_ anything useful and sucks up a service slot 06:26 < storge> ah, android apps 06:27 < Dagmar> Android already handles the hand-offs between wifi and 3g/lte/whatever and has done so fairly well since before they decided they needed to invent such a silly thing 06:27 * storge kisses his hand like a chef 06:28 < Dagmar> 'cuz basically... Linux kernel and networking already solved these problems quite a while back 06:29 < Triffid_Hunter> Dagmar: I always enable the aggressive handoff option in developer settings.. stops it having to reconnect to 3g/4g when wifi drops out 06:29 < Dagmar> IT works 06:30 < Dagmar> Probably annoys the carriers a bit, but it's a miniscule load on their end 06:30 < hyperair> Triffid_Hunter: o/ you use lineage too? what's good to get these days? my phone's got one foot in the grave with lineage dropping support for it. 06:31 < Dagmar> I'm eyeing the Moto X lines and OnePlus for various hardware reasons 06:31 < Dagmar> It's very clear the workload I'm asking of the device needs 4Gb RAM or more 06:31 < Triffid_Hunter> hyperair: got my eye on xiaomi mix2, waiting for next payday 06:32 < hyperair> Triffid_Hunter: ah, xiaomi after all then. i had my eyes on the xiaomi mi 5 or mi 6 06:32 * hyperair has small hands 06:32 < Dagmar> hyperair: At this point you might consider doing the same thing I am... just stalking the various cell enthusiast sites for whose new shiny runs Lineage well, and work backwards from there 06:32 < Triffid_Hunter> hyperair: mix2 has /all/ the radio bands and official lineage support 06:33 < hyperair> Dagmar: sounds good. i've been stalking download.lineageos.org and working backward from there 06:33 < hyperair> Triffid_Hunter: but it's a 6" phone, no? 06:33 < Triffid_Hunter> I'm not terribly pleased about no SD card slot and no headphone port though, but they seem to be increasingly rare on modern phones :/ 06:33 < hyperair> urgh 06:33 < hyperair> i do like my headphone port 06:33 < Triffid_Hunter> hyperair: yeah but that's no problem for me, my handspan is 25cm :P 06:33 < hyperair> naise 06:34 < Dagmar> I can do without the headphone port, but without a slot there needs to be 64Gb-128Gb storage, just to keep up with what I've got 06:34 < darkmeson> Not really 06:34 < hyperair> Dagmar: true that. i need some extra storage for ebooks too 06:34 < Triffid_Hunter> Dagmar: it comes in 64/128/256G internal storage variants, there's also a 8G ram option with the 128G storage 06:34 < Dagmar> I actually encoded several hundred of my own CDs 06:34 < hyperair> o okay 256GB ought to be enough for anybody 06:34 < Dagmar> There's no real sense in streaming that all the time. It'd be a waste of bandwidth 06:35 < Dagmar> 48Gb of the 64Gb SD I have are just those mp3s 06:35 < hyperair> i think it should be enough for me too, but i always hit the storage limit somehow or other 06:35 < Triffid_Hunter> Dagmar: yeah I always did my own encoding back when CDs/DVDs were a thing.. I like my compressed music in ogg vorbis VBR rather than mp3 ;) 06:35 < hyperair> yeah vorbis vbr is great 06:35 < Dagmar> 2Gb of it is an image that sometimes even works that I'm trying to use to be able to boot PCs off the phone 06:36 < Triffid_Hunter> I have a 64G SD in my current phone, it's pretty full.. having a torrent client on my phone probably doesn't help with that though ;) 06:36 < Dagmar> Triffid_Hunter: I went back and re-encoded about 200 of them after LAME added the new vbr mechanisms 06:36 < darkmeson> hyperair: it probably is if you're trusting it at an appropriate level 06:36 < hyperair> darkmeson: trusting what? 06:36 < hyperair> darkmeson: my phone? i don't trust it enough to stick ssh keys on 06:36 < Triffid_Hunter> Dagmar: heh did you run into the thing where lots of players can't work out the length with vbr mp3? pretty sure ogg just lists it in metadat 06:36 < Dagmar> I was already telling it to be as accurate as possible, so they werne't exactly small mp3s 06:37 < Dagmar> I've seen a few. I think you're right about the length being written as a field in a metadata tag\ 06:41 < darkmeson> hyperair: this was kind of the point. there's only so many mp3s you can listen to, and they can't be trusted for much else 06:42 < darkmeson> Tablets lacking a cell modem, maybe. not phones 06:42 < hyperair> darkmeson: true that 06:43 < hyperair> darkmeson: but i do play some mobile games with loads of downloaded media 06:43 < hyperair> darkmeson: and i do store a pretty large collection of epubs on my phone 06:44 < hyperair> plus mainstream apps like facebook, whatsapp, chrome, etc all get heavier over time 06:44 < Dagmar> I occasionally use mine like a portable data safe 06:44 < darkmeson> You don't need them 06:44 < Dagmar> ...which is part of what got me started messing with the software to make it present itself as a removeable drive to boot PCs 06:45 < Dagmar> It's _slow_ but at least marginally useful for hasty forensics 06:45 < darkmeson> There are lightweight webview wrappers around most services in F-Droid, and they notify you all the same 06:45 < hyperair> darkmeson: that's arguable 06:46 < Dagmar> Facebook's web notifications are actually notably less useful than the app 06:46 < hyperair> whatsapp lives on the phone, so it's not like you can have a light webview wrapper unless you have a phone acting as a server for it stashed somewhere 06:46 < Dagmar> ...and the app is a terrible example of why you don't try and reinvent push notifications on your own 06:46 < darkmeson> You ARE aware of the plethora of USB exploits, right? 06:46 < hyperair> darkmeson: eh use a usb condom and you're fine 06:47 < hyperair> sorry i meant usb fast charge adapter 06:47 < Dagmar> I would not call a single PoC a "plethora" 06:47 < hyperair> the cheapo things that short data pins together 06:47 < darkmeson> Plugging a phone into any and every computer might not be the best idea, especially given the usage and the many ways it could already be compromised 06:48 < Dagmar> Also, the vuln for that PoC was patched some months back 06:48 < Dagmar> Do you typically try to boot a PC from external media by plugging it into the thing while it's still on? 06:48 < Dagmar> 'cuz that's not really how that works, man 06:49 < darkmeson> Dagmar: Go to a security channel and get schooled, please. I have neither the time nor inclination to teach you myself 06:49 < Dagmar> Dude my name is in newspaper articles from before the century turned 06:50 < Dagmar> I don't need to "get schooled" 06:50 < Dagmar> I had my first cert in that field probably before you were born, judging by your attitude 06:50 * Aph3x-WL makes popcorn 06:50 < darkmeson> Also, I'm not entirely sure you're not just trying to troll me anyway 06:50 * hyperair steals some popcorn from Aph3x-WL 06:51 < CrazyTux> hello, I am not able to use KDE connect on Ubuntu Mate 18.04. 06:51 < Dagmar> I'm not going to tote around a whole system just to grab malware off a friend or a relative's PC 06:51 < Aph3x-WL> you could have just asked, i made plenty :P 06:51 < Dagmar> They're not of interest to nation states, so cold-booting the PC to boot from my phone is pretty darned safe 06:51 < CrazyTux> not able to browse files on my android phone. 06:52 < Dagmar> ...and no one was talking about plugging their phone into random PCs all over the place except _you_ 06:52 < darkmeson> Dagmar: just take it to a security channel before you insert your foot any further into your mouth 06:52 < Dagmar> I'll just simplify things by plonking you 06:53 < Dagmar> CrazyTux: That functionality is pretty basic. Be sure you've enabled MTP on your phone and that there's at least some packages installed with "mtp" in the name that your preferred file manager actually uses 06:53 < Dagmar> If you're somehow sitll using the USB removable device stuff, you should consider a newer phone 06:53 < Dagmar> It never worked very sanely on the devices where that was still an option 06:54 < CrazyTux> both my phone and laptop were purchased last year. 06:54 < Dagmar> Okay, so _definitely_ ensure that MTP mode is enabled on the phone 06:54 < CrazyTux> Dagmar, how to do that? 06:54 < Dagmar> That's the mechanism it uses to expose files to a PC now, and it's both old and common 06:54 < Aph3x-WL> does kde connect even work outside of kde/plasma? 06:54 < CrazyTux> Dagmar, I use an android phone. 06:55 < Dagmar> Under Settings, but it should also basically be accessible from the notification of "hey I'm charging here" that appears when you plug into a host port 06:55 < swift110> me too CrazyTux 06:55 < Dagmar> Just tap the notification and you should get the option to enable MTP for that session 06:55 < Dagmar> Assuming the DE knows about MTP, it should just appear as if you'd plugged in a camera or something 06:56 < Dagmar> Ubuntu Mate definitely has those packages 06:56 < CrazyTux> when I had Kubuntu Desktop installed I didn't have any problem. 06:56 < Dagmar> Okay then you've either gotten MTP turned off in the file manager, or one of your assumptions (like MTP being automatically enabled) is no longer correct 06:56 < CrazyTux> but, I am using just Ubunt Mate now with this app kde connect installed. 06:57 < Aph3x-WL> mate is not kde, so maybe it's not setup to do that out of the box 06:57 < Dagmar> Very possible 06:58 < Dagmar> It's possible libmtp and friends just weren't installed for some reason 07:01 < CrazyTux> I am clueless. 07:02 < swift110> hmm 07:04 < Elladan> CrazyTux, this is going to sound dumb, but try plugging your phone into different USB ports and so forth. 07:05 < Elladan> Also try a different cable. 07:05 < CrazyTux> Elladan, doesn't kde connect work through wifi? 07:05 < swift110> good idea Elladan 07:05 < Elladan> Oh, sorry, I was looking up and seeing talk about android MTP. 07:05 < Elladan> I don't know anything about KDE connect. 07:06 < CrazyTux> Elladan, I am not able to browse my android phone using KDE connect. 07:06 < CrazyTux> I am using Ubuntu Mate 18.04 with KDE Connect installed on it. 07:06 < CrazyTux> I can send and receive the files between my laptop and phone though. 07:08 < Aph3x-WL> do you have kde connect installed on your phone as well? 07:08 < screwsss> hey everyone 07:08 < Elladan> CrazyTux, you might check firewall settings if appropriate. 07:11 < bipul> Good morning. 07:12 < Dagmar> Yeah he does definitely need the KDE connect app running on the phone, but I assumed that was not part of the "browse files" problem 07:12 < Dagmar> Android's idea of display plugging is basically "HDMI or you're on your own" and that's only if the hardware supports it 07:13 < domhnall> CrazyTux: hate to ask you repeate the end goal but could you? 07:13 < Dagmar> The other way around _always_ requires an app of some sort 07:13 < Dagmar> (which is irksome because you can get some older phones now for less than the cost of a USB display) 07:16 < ayecee> would be so nice to use one of those old cheap tablets for that, since they're almost useless for anything else 07:16 < Dagmar> CrazyTux: Do you mean you're trying to simply make KDE Connect talk to your phone? 07:17 < Dagmar> ayecee: I have a wingray Xoom, and I've yet to find a solution for it that doesn't at least half-suck 07:17 < ayecee> doing pretty good if it only half-sucks 07:18 < Dagmar> I'm half-tempted to try it the way GPU vendors do... just send a video stream to a player app 07:18 < Dagmar> ..but for what I've been doing it's basically been easier to just _write a freakin' app_ that polls the data I need to see 07:19 < ayecee> nod 07:20 < Dagmar> I have an image for the S3 I've been tempted to try, since it's supposed to have X 07:20 < bipul> web crawler? 07:21 < CrazyTux> what is the difference between using sudo apt and sudo tasksel? 07:22 < bipul> I'm just trying to make a web crawler that search specific strings from random website on internet, It works on regular expression. 07:22 < oerheks> tasksel installs a service, apt installs a package, which can be a service 07:22 < ayecee> tasksel presents metapackages represented as tasks 07:22 < CrazyTux> ok 07:23 < ayecee> i don't know if those same task metapackages are available through apt as such 07:23 < oerheks> sure, lamp 07:23 < ayecee> Dagmar: you mentioned S5 and Lineage. is there an image for that? any downsides compared to running stock? 07:24 < ayecee> i remember when i tried jailbreaking an iphone, the ebook app could no longer open books, presumably because it no longer had access to the keys in whatever is the equivalent of a TPM 07:30 < CrazyTux> I think Airdroid is a better option. 07:31 < CrazyTux> can be used on any DE on any distro easily. 07:31 < darkmeson> If it's anything like the process for Android phones, you lose everything when you unlock the bootloader anyway 07:32 < ayecee> that's to be expected 07:33 < darkmeson> well, yes. it's a security feature 07:36 < darkmeson> I'd always intended to Androidify an Apple device, but it's hard to justify the price of even an older model when there are better things on the market, cheaper 07:37 < ayecee> also, apple ios does things pretty well, if you're okay with the walled garden 07:38 < sauvin> That "walled garden" is precisely why I tell people NOT to buy iPoop. 07:38 < darkmeson> That's where we ran into trouble 07:39 < darkmeson> The wifi debacle on the first gen iphone cured me of any desire to run their software stack 07:39 < Elladan> I found airdroid to be unreliable these days, and switched to just using an ssh server on the phone. 07:39 < ayecee> sauvin: the wisdom giveth, and the childish language taketh away 07:40 < darkmeson> Elladan: termux-based, or? 07:40 < sauvin> ayecee, it's a religious thing, literally. If you already understand, no words are needed, and if you don't already understand, no amout of foul language is going to matter. 07:41 < CrazyTux> in terms of usability which DE is the most convenient? 07:41 < sauvin> Define "convenient". 07:41 < Disconsented> subjective to a point 07:41 < ayecee> sauvin: right up there with micro$hit. don't do that. 07:41 < CrazyTux> one that can be installed and is ready to use instantly. 07:42 < Elladan> darkmeson, uh, "SimpleSSHD" dunno maybe it's pwned me and sold all my data to aliens. 07:42 < sauvin> CrazyTux, too many for me to count. You're going to have to narrow it down a bit. 07:42 < CrazyTux> sauvin, kde or mate or gnome? 07:42 < CrazyTux> or cinnamon? 07:43 < sauvin> I'll go this far for you, and no further until you decide to be a bit more specific: I personally use KDE. 07:43 < CrazyTux> sauvin, but some say KDE is resource hungry and buggy too. 07:44 < sauvin> Some say that. It's said to be pretty heavy, but it's also very, very powerful. Try it and see - if you don't like it, zap it. 07:44 < Elladan> I use Cinnamon. It's buggy but doesn't make me want to punch things. 07:44 < CrazyTux> ok 07:44 < CrazyTux> sauvin, btw, which distro you use? 07:45 < sauvin> CrazyTux, Ubuntu and Debian. 07:45 < CrazyTux> sauvin, some say OpenSuse has the best KDE implementation. 07:46 < sauvin> That's possible. 07:47 < ayecee> some say holding your nose when you sneeze causes brain damage 07:47 < CrazyTux> lol 07:47 < Elladan> There isn't any "best" DE, though there are certainly a lot of worst ones. What people like is pretty personal. 07:48 < searedvandal> so what's the worst ones? 07:48 < CrazyTux> Elladan, I was asking from a newbie's perspective. 07:49 < pingfloyd> searedvandal: gnome or kde 07:49 < pingfloyd> I'd say hands down unity, but that one is gone now 07:49 < Elladan> Though honestly none of the modern ones seem to work as well as Gnome 2 did with Compiz back in 2010. 07:49 < ayecee> also ratpoison 07:49 < Elladan> searedvandal, ... Enlightenment? :-D 07:49 < ayecee> and fvwm95 07:49 < pingfloyd> yeah, ratpoison is pretty horrible 07:49 < CrazyTux> from the perspective of productivity, for a newbie and an end user. 07:49 < pingfloyd> it's designer suffers from OCPD or something 07:50 < pingfloyd> he's obsessed about not touching his mouse 07:50 < CrazyTux> pingfloyd, I agree. Ratpoison kills. 07:50 < Elladan> KDE, Gnome, Cinnamon, and Mate will all work fine for a new user interested in being productive. 07:50 < CrazyTux> lol 07:50 < ayecee> which is to say that worst is also pretty subjective 07:50 < Elladan> xfce will work too, but it's more geared towards low-end machines. 07:51 < CrazyTux> Elladan, ok. 07:51 < pingfloyd> I'm all for more usage of the keyboard in the gui, but there's a certain point where things become unreasonable. 07:52 < pingfloyd> I wouldn't be surprised if ratpoison has a "feature" where you can move the mouse cursor around through keyboard comb chains 07:52 < searedvandal> haha 07:53 < searedvandal> sounds fun 07:53 < CrazyTux> suppose one is asked to install a linux distro in an office where the people working don't have any knowledge of linux and are least interested in tinkering with the OS and just want their work done. 07:53 < pingfloyd> I bet if you say to them, "what about programs like Gimp?", they would reply "We do not recognize the existence of gimp.". 07:54 < darkmeson> Elladan: that's a shame. I was hoping you'd be able to tell me how well it works. I've not made the shift from SSHDroid yet, and it DOES try to phone home. 07:56 < hexnewbie> “You use GIMP?!” “Yeah.” “Can you help me fix my Amazon instance?” 07:56 < CrazyTux> which distro and which DE for the above case? 07:56 < pingfloyd> Elladan: I'd say that xfce is more geared toward gui minimalism 07:56 < pingfloyd> Elladan: i.e., having the standard and expected feature of a desktop but not much more. 07:56 < pingfloyd> *features 07:57 < pingfloyd> features get added to xfce for instance, but they're usually features that are beyond being the latest fad. 07:58 < pingfloyd> ubiquitous current "must have" features instead 07:58 < domhnall> CrazyTux: Pretty basic case for any binary distro that uses an installer. OpenSuse Leap, comes to mind with KDE. 07:58 < pingfloyd> Elladan: it just tends to work out better for low-end hardware because it isn't so bloated in comparison to the others. 07:59 < pingfloyd> LXDE is even lighter though 07:59 < hexnewbie> Features don't actually make Xfce light (I had to switch to LXDE when I fount Xfce nearly seemed as heavy as KDE), but at least it's not extremely buggy because they don't have “features before bugfixes” attitude. Then I went back to KDE because I found all the missing features were essential, and I can deal with constant crashing. 07:59 < hexnewbie> s/Features/Lack of features/ 07:59 < pingfloyd> I guess supposedly LXQT is supposed to be lighter as well, but when I tried, it sure didn't feel lighter than xfce 07:59 < CrazyTux> domhnall, ok 08:00 * darkmeson has been able to make just about every DE look and act the same way, with varying degrees of difficulty 08:00 < mrig> Hi, C code is beautiful :) 08:00 < pingfloyd> hexnewbie: yeah, the other part of what makes it light is sanity in its design 08:00 < ayecee> could be that the wm is not normally the heaviest part of normal operation 08:00 < ayecee> or de 08:01 < domhnall> CrazyTux: imo, only feature (application) important on KDE is akregator. 08:01 < hexnewbie> The web browser is the heaviest part. 08:01 < pingfloyd> your web browser is usually much heavier 08:01 < pingfloyd> unless you use some light browser 08:01 < pingfloyd> like midori or dillo 08:01 < domhnall> midori still around? cool 08:01 < darkmeson> qt DOES annoy me though because it requires two different tools to set the theme, and even then it doesn't always correctly apply the dark themes 08:01 < pingfloyd> yep 08:01 < CrazyTux> domhnall, ok 08:01 < pingfloyd> midori is great if you're stuck with a low-end system 08:02 < ayecee> tbf though, unity's default requirement for 3d acceleration made it pretty heavy on older systems 08:02 < CrazyTux> domhnall, so recommend OpenSuse with KDE? 08:02 < domhnall> last use of midori for me was constant crashes 08:02 < Elladan> pingfloyd, that's a reasonable way to look at it, yeah. It does have some more resource-intensive features like compositing, too. 08:02 * darkmeson just installed midori a few hours ago 08:02 < domhnall> CrazyTux: no, I said it's a great candidate for your use case mentioned.... 08:02 < papul> whats the best and quickest way to test some OS in linux? VirtualBox? 08:02 < ayecee> probably yes 08:02 < CrazyTux> domhnall, ok 08:02 < domhnall> I make no recommendations on ##linux unless it's a customer 08:03 < ayecee> aren't you a principled sob 08:03 < domhnall> heh 08:03 < CrazyTux> domhnall, consider me as your customer. 08:03 < Elladan> darkmeson, as an ssh server for a phone it works perfectly fine. You either load a key into it, or it'll show a one-time password when you connect. 08:04 < Elladan> darkmeson, ... and I've never had any trouble with transferring gigabytes of data like I do with other tools. 08:04 < pingfloyd> over the years, I used many different DEs, and gave many of the second tries later on. The one it finally landed on for me, after a second try, was xfce. 08:04 < pingfloyd> it was like having gnome without the insane mentality 08:04 < Kirball> There’s a Mr Robot Badge PCB on eBay from last years defcon personalized and autographed by int80 of Dual Core. If anyone’s interested. 08:04 < pingfloyd> mentality like designed to be non-configurable 08:04 < hexnewbie> papul: QEMU/KVM often with virt-manager as the GUI frontend is often recommended here. But it's not as user friendly, so even if it is not best, VirtualBox *will* be way easier and quicker (and perhaps your only reasonable choice). 08:05 < Elladan> darkmeson, Also there aren't any ads. But I mean, I didn't capture its network traffic or disassemble it so I can't vouch for its benign nature. 08:05 < pingfloyd> I always liked gnome more or less, but it has some deal breakers mostly from the gnome project's philosophy. 08:05 < hexnewbie> papul: It also depends on whether you want to try out a server OS, or a desktop OS. 08:05 < pingfloyd> like I think they're half-way on the right track in their thinking 08:06 < pingfloyd> the modular mentality is one of their good points 08:06 < darkmeson> papul: httpo://copy.sh/v86/ :) 08:07 * pingfloyd wonders how long before gnome has a ribbon 08:07 < darkmeson> https://copy.sh/v86/ 08:07 < domhnall> CrazyTux: If that's the case, consider hardware specs and space requirements 08:07 < pingfloyd> "The Ribbon" is one of the biggest UI abominations ever invented 08:07 < hexnewbie> When GNOME gets a ribbon, they'll make sure the tabs on the ribbon get 30px padding on top and bottom, and the buttons in the ribbon below have 60px padding on all sides. 08:08 < pingfloyd> from a UI perspective that's about the only thing that is bad about Windows 10 08:09 < CrazyTux> domhnall, laptops with Intel Core i3 2.0 Ghz quad core cpu and 4 GB RAM and 1 TB HDD. 08:09 < hexnewbie> The ribbon interface was quite nice in Quanta Plus, and is quite nice in Bluefish right now. I wish Quanta was still alive, and Kyle or TeXstudio had similar UI design to BlueFish and Quanta 08:10 < Elladan> I honestly don't really care that much about a lot of the GUI UI things people talk about. I mean, I mostly just use terminals, text editors, and web pages. 08:10 < _xor> Hey guys, I have a table-top multi-touch screen. Is there a list of applications that I can browse through to experiment with? 08:10 < Elladan> The things that really bother me tend to be bugs more than anything else. 08:11 < darkmeson> papul: in seriousness though, virt-manager and kvm or xen if you don't need graphical acceleration, or virtualbox if you do 08:11 < hexnewbie> Unrelated, after Firefox switched to GTK 3, the GTK 3 idiotic padding has made some of my web pages (ones with very tall tables, with few hundred rows), 3-4 times larger in height, as every row has because an input field in the table now has padding larger than the font. 08:12 < papul> hexnewbie, darkmeson, I am trying out in virtualbox for now. Let's see 08:12 < papul> it's some server OS. YunoHost 08:12 < hexnewbie> Tried ~/.config/gtk-3.0/gtk.css to fix the web pages, but I can edit the padding of everything *but* the boxes, and someone told me that was deprecated now as well 08:12 < _xor> I haven't seriously messed with Linux in years. It's interesting seeing the landscape these days. 08:14 < darkmeson> We actually have several html5-based DEs that can run in a browser today, but that's made somewhat less amazing by the fact that the browser can run a whole OS including a GUI (see the previous link) 08:14 < hexnewbie> I don't think GNU/Linux will be ever ready for anything if it keeps deprecating the only way to do something. Like my 15 year quest to get a laptop's touchpad to work. Every 1-2 year, the workaround to fix the insane defaults gets deprecated, cuz (Xorg.conf, hal, udev rules, xinput, now that's gone though). Yuck. 08:14 < Elladan> I mean just in terms of basic usability, like if the lock screen has a bug where it doesn't show a password box when you open your laptop sometimes. I know how to break out and kill things from the console, but any normal user would be holding down the power button. 08:16 < hexnewbie> First synaptics touchpads worked out of the box, then the synaptics authors decided that it was bad design for them to work by default, so you needed Xorg.conf, then you had to remove that because it interfered with autodetection and use hald instead, then hal got removed so you use udev rules, then those got removed and now you use xinput commands in ~/.xsessionrc (which doesn't work in GNOME), and now Wayland gets rid of that. How can this 08:16 < hexnewbie> be a usable OS? 08:16 < kerframil> hexnewbie: I got back into linux on the desktop lately and am struggling with that also. I don't know how to disable coasting while using a trackpoint, for instance, nor how to make wheel scrolling behave as I expect. I launch imwheel when using an external mouse but it has limitations; I really would prefer for these options to be exposed in gnome-settings so that I can just set it and forget it. 08:17 < kerframil> hexnewbie: the acceleration curve feels a little off in libinput too, or so it seem to me 08:17 < _xor> I'm a FreeBSD guy. I appreciate both Linux & BSD. The best way I've heard the differences described is: "BSD is what you get when a bunch of geeks get together and engineer a UNIX OS. Linux is what you get when you get a bunch of geeks together and grow a UNIX OS."" 08:17 < Elladan> hexnewbie, pragmatically that's something the installer should have configured for you automatically. 08:17 < pingfloyd> hexnewbie: yeah, things change at a fast pace in linux 08:18 < pingfloyd> hexnewbie: that reality about it, isn't friendly for beginners for sure. 08:18 < Elladan> hexnewbie, however I can understand your frustration that it isn't standardized in some way that makes looking up information easy. 08:18 < pingfloyd> hexnewbie: this is why it is hard to recommend to the average user. Really it's power users that naturally become sticklers about their computing that I think it appeals to most. 08:18 < hexnewbie> kerframil: Yeah, I purged libinput immediately to go back to evdev, because after playing with libinput's settings for 15 minutes, I couldn't find anything resembling the 4x polynomial deceleration of evdev... So my mouse with libinput seemed unusable to me. 08:19 < CrazyTux> can we install and use any application on windows when we install windows in virtualbox on a linux distro? 08:19 < Elladan> CrazyTux, as long as it runs with the virtual machine's emulated hardware. 08:20 < pingfloyd> Elladan: closest things we have to standards are BSD, System V, and posix 08:20 < darkmeson> Elladan: some of us have been at it so long that it's nearly impossible to put ourselves in new users' shoes. that's a goodly part of the problem 08:20 < CrazyTux> Elladan, ok 08:20 < pingfloyd> Elladan: outside of that, it's like the wild west 08:20 < Elladan> pingfloyd, that's not exactly true, but it's close enough. 08:20 < pingfloyd> okay and LSB 08:20 < pingfloyd> and HFS 08:20 < pingfloyd> I'm sure there's others, but those are the main things 08:21 < Elladan> darkmeson, I don't think time does that. I mean, I've been doing this for a long time too and it's intuitively obvious to me that "break to console and start hunting for a process to kill" is not an acceptable user experience. 08:21 < pingfloyd> we have several competing "standards" at all levels 08:22 < CrazyTux> domhnall, your recommendation? 08:22 < darkmeson> CrazyTux: make sure you've installed the guest extensions if you expect any multimedia apps to perform well at all, and make sure you enabled the graphical acceleration option in the UI also 08:22 < pingfloyd> between those 3 I'm kind of partial to system v 08:23 < darkmeson> Elladan: it was a metaphor, and that's what it meant ;) 08:23 < Elladan> pingfloyd, distros, particularly Debian, also set de-facto standards for doing things. However, it's obviously true that there needs to be a lot more standardization. 08:23 < CrazyTux> darkmeson, ok. 08:23 < pingfloyd> Elladan: yeah, but those are definitely de-facto ones 08:23 < pingfloyd> separate class from the first 3 08:24 < Elladan> CrazyTux, re vms, the main thing to keep in mind is that the VM will probably not support GPU acceleration, or at least not very well. 08:24 < Elladan> CrazyTux, this means things like games or some CAD tools and the like won't run. 08:24 < CrazyTux> Elladan, ok 08:25 < CrazyTux> Elladan, I don't intend to run games on it. 08:25 < Elladan> There are often workarounds, but they're not really of the user-friendly sort. 08:25 < domhnall> CrazyTux: hm? sorry missed a part, cant scroll as on er...hackbook. what's the vendor? HP? 08:25 < Elladan> Any sort of "ordinary" program like a word processor will work fine. 08:25 < darkmeson> I never thought I have to use this term but, it'd help if people stopped the NIH and stopped poettering everything up 08:25 < CrazyTux> Elladan, I want to be able to install SAP ERP on Windows that is installed in VirtualBox on a linux distro. 08:25 < CrazyTux> domhnall, asus. 08:26 < CrazyTux> domhnall, and some are of HP too. 08:26 < pingfloyd> Elladan: here's a good link to pass on to beginners since it gives a good, straight to the point, run down on the standards. https://www.youtube.com/watch?v=jowCUo_UGts 08:27 < Elladan> CrazyTux, software of that nature should be fine. 08:27 < CrazyTux> Elladan, ok 08:27 < darkmeson> I don't really like to pick on him, but he DOES seem to be the perfect example of someone who doesn't even bother to understand why things are the way they are before coding something completely random, conflicting, and often poorly-thought-out 08:27 < domhnall> CrazyTux: https://en.opensuse.org/HCL:Laptops 08:27 < Elladan> CrazyTux, if you intend to run anything important in the VM, make sure you understand how the VM guarantees data integrity and have the cache settings configured appropriately. 08:27 < CrazyTux> domhnall, ok. 08:28 < CrazyTux> Elladan, ok. I will try to understand that. 08:28 < Elladan> That's not really an issue for web browsing or whatnot, but SAP ERP sounds like the sort of software where you might care about the data. 08:29 < CrazyTux> Elladan, yes. 08:29 < CrazyTux> domhnall, thanks a lot for that info. 08:30 < domhnall> CrazyTux: I'll bill you later :) 08:30 < CrazyTux> domhnall, yeah. Sure. why not? :) 08:31 < pingfloyd> CrazyTux: if you have vt-d, I'd recommend using qemu-kvm instead 08:31 < pingfloyd> along with virtio 08:31 < CrazyTux> pingfloyd, what is vt -d? 08:32 < CrazyTux> pingfloyd, sorry, I am a linux newbie. Am not aware of what these things are. 08:32 < pingfloyd> CrazyTux: Intel Virtualization Technology for Directed I/O 08:32 < Elladan> I've found that virtualbox works better than qemu-kvm at running a guest GUI. 08:32 < pingfloyd> or having the AMD equiv if you're using AMD of course 08:33 < Elladan> qemu-kvm is more geared towards running a server in your VM. 08:33 < CrazyTux> pingfloyd, the laptops run on Intel Core i3 cpus. 08:34 < Elladan> I don't think vt-d is relevant to doing light GUI usage in a guest VM. 08:35 < darkmeson> Elladan: It's because of the guest extensions 08:35 < pingfloyd> Elladan: it is 08:35 < Elladan> How? 08:35 < darkmeson> at least with windows guests 08:36 < pingfloyd> the display driver you can use 08:36 < darkmeson> afaik, they basically use wined3d to convert directx to opengl, which they can then pipe to the gpu easier across existing channels 08:36 < pingfloyd> in your windows guest, you want to be sure to install the virtio drivers 08:37 < pingfloyd> also should enable the hyper-v elements 08:37 < darkmeson> or at least that's how it used to work, back in the day 08:37 < darkmeson> they might have some additional zero-copy sliding buffer system by now 08:38 < Elladan> My understanding of VT-d was that it was relevant if you wanted to give the VM direct access to a piece of hardware. 08:38 < darkmeson> it is, but that's not what virtualbox is doing 08:38 < Elladan> That's never going to be the case for a basic sort of VM setup, it's either for high performance server deployments or if you need the guest to own your GPU for some reason. 08:39 < darkmeson> come to think of it, they DO have that virtio gpu code now, but I have yet to see it used anywhere 08:40 < pingfloyd> Elladan: also, you're forgetting about vga passthrough 08:40 < CrazyTux> thanks a lot for all this info. 08:40 < pingfloyd> that requires vt-d/iommu 08:40 < Elladan> Just getting 3d acceleration in the guest via virtio / guest extensions doesn't involve direct hardware access. 08:40 < darkmeson> Elladan: the latter falls under pci and vga passthrough, and just about all virt stacks support it now, I think 08:40 < domhnall> Hope you can digest it easily 08:41 < Elladan> pingfloyd, I didn't forget it. 08:41 < pingfloyd> it gets you around "95% native performance". 08:42 < Elladan> I just don't think that stuff is really new-user material, except maybe if they really want to run Windows games or something. 08:44 < pingfloyd> well, I'd say kvm is a bit more of challenge to setup properly than vbox 08:44 < pingfloyd> but not that much more 08:44 < pingfloyd> you're essentially having to do the same things, just under different terminology 08:44 < darkmeson> I guess they're calling the new method virgl, and some people are using it already. https://www.reddit.com/r/archlinux/comments/7nmceg/kvmqemu_with_virtiogpu_virgl_support_enabled/ 08:45 < domhnall> CrazyTux: would like to apologise also, Just figure how to scroll up logs... 08:45 * darkmeson might have to dust off a Windows vm and see how well it works 08:46 < Elladan> I'm using kvm on a server. It mostly works fine, but using a guest GUI is a pain. 08:47 < pingfloyd> it works with windows 10 for me pretty well. 08:47 < pingfloyd> (windows 10 guest) 08:48 < Elladan> Virtualbox seems less buggy and much faster at rendering a GUI. The vbox gui also seems easier for a new user to navigate to me. 08:48 < pingfloyd> Elladan: which guest? 08:48 < domhnall> no seems about it...it is easier than vmware and qemu 08:48 < Elladan> Various linux guests. 08:48 < Elladan> I don't have any use for windows. 08:49 < cmj> kvm++ 08:49 < pingfloyd> it's only easier in that it is all configured in one app 08:49 < pingfloyd> with kvm it's more about configuring your linux system since it's part of the kernel. 08:49 < pingfloyd> but there's apps that help for that such as virt-manager 08:49 < Elladan> I won't say I've found vbox very reliable over the years, just that it seems simple and gui-oriented guests worked well. 08:50 < pingfloyd> I think there's no way around having to run virsh commands though at some point 08:50 < pingfloyd> but you could say practically the same about VBoxModify 08:51 < Elladan> Yeah, kvm isn't bad or anything. It just seems like a better solution of you're running a server. 08:52 < Elladan> I also get rather poor IO performance with kvm, but I think I saw some IO path optimizations in a new version I haven't bothered to upgrade to. 08:52 < Elladan> (I seem to recall virtualbox cheats at IO benchmarks by defaulting to fsync=off, or it used to anyway) 08:53 < darkmeson> I mostly use TurboVNC within my guests since its algorithm isn't terrible, and it can even pull off fullscreen multimedia in some instances 08:54 < pingfloyd> darkmeson: I've been able to watch 1080p fine in the windows 10 guest 08:55 < pingfloyd> even on a $300 craptop 08:56 < darkmeson> That's probably with both being machine-local 08:56 < Elladan> I've had serious rendering / refresh problems even in recent versions of virt-manager + kvm, and things like the mouse pointer being off by 50 pixels etc. 08:56 < darkmeson> Mine is from a server to another system over GbE 08:58 < Elladan> I mean I got it working with some random trial-and-error, but there were definite issues. 08:59 < darkmeson> The slide-down bar at the top of virt-manager in full screen mode has been transparent on the last three distros I've used 08:59 < darkmeson> That made trying to get out of fullscreen mode fun until I started getting an idea where I needed to be clicking 09:00 < darkmeson> virt-manager has always been horribly buggy like that though, and no surprise it's written in Python 09:00 < pingfloyd> that's when you use virsh 09:00 < pingfloyd> eventually you don't even bother with virt-manager 09:01 < pingfloyd> would be nice to see virt-manager cleaned up a bit though anyway 09:01 < pocketmon> how can i install eclipse cdt? 09:02 < darkmeson> I usually don't anyway 09:02 < pocketmon> me? darkmeson ? 09:03 < darkmeson> I have serial consoles set up, but Android never managed to attract anyone willing to maintain a VNC server over the long haul, so I use virt-manager for one specific app that I have to run within an android-x86 vm 09:04 < pingfloyd> isn't that just an eclipse plugin? 09:04 < darkmeson> pocketmon: don't get carried away. there's lag you know. what distro? 09:04 < pocketmon> oh then how can i install eclipse? 09:05 < darkmeson> what distro? 09:05 < pingfloyd> pocketmon: use your dist's package manager 09:05 < pocketmon> apt-get pingfloyd 09:05 < pingfloyd> yeah 09:06 < pingfloyd> try apt-cache search -n eclipse 09:06 < pingfloyd> to find the package name 09:06 < pingfloyd> then apt-get install packagename 09:06 < pingfloyd> probably eclipse anyway 09:06 < pingfloyd> which is probably a meta-package 09:07 < Elladan> On my distro there's an "eclipse-cdt" package. 09:07 < storge> cdt? 09:07 < pocketmon> apt-get install eclipse-cdt this is ok? 09:07 < nindustries> Hi, I noticed that executable files keep working when I cat random bytes to them. (e.g. $(cat /dev/urandom | head -n1) > executable 09:07 < pingfloyd> there's tons of packages for eclipse on mine 09:07 < pocketmon> i want to use asm ide 09:08 < nindustries> Does that mean that the bytes will be contained in the app (referenced in the header), but they will not interfere with the inner workings? 09:08 < Elladan> pingfloyd, yeah pocketmon asked about eclipse cdt specifically. 09:08 < Elladan> pocketmon, it is likely that will install it for you. You may need sudo. 09:08 < storge> i don't see a -cdt in debian buster 09:09 < storge> unless its in one of the others. apt-cache show $package 09:09 < Elladan> nindustries, when you do that, you're deleting the executable that was there and replacing it with some random garbage. 09:09 < storge> what is cdt? 09:09 < pingfloyd> Elladan: then asked about installing eclipse, and then about an asm ide 09:09 < Elladan> nindustries, since the running app has a reference to the old deleted file, it sticks around until the app is closed. 09:10 < Elladan> nindustries, to be slightly more clear, you're deleting the file and then creating a new file with the same name. 09:10 < paulcarroty> Absolute path to 'lspci' is '/sbin/lspci', so running it may require superuser privileges (eg. root). 09:10 < paulcarroty> crap, opensuse is funny 09:10 < nindustries> Elladan: sorry, I meant >> 09:10 < pingfloyd> storge: try apt-cache search eclipse cdt (stretch has it for sure) 09:10 < darkmeson> Ubuntu has it, but Debian doesn't 09:10 < pingfloyd> darkmeson: Stretch has it 09:11 < pingfloyd> has a bunch of eclipse-cdt-* packages 09:12 < Elladan> nindustries, with >> you're appending junk to the end of the executable. Many executable binary formats will not use this junk, but some will stop working. 09:12 < Elladan> nindustries, however, whether a running instance will notice the new data depends on many factors. 09:12 < darkmeson> storge: it's the version for C/C++ 09:12 < nindustries> In my test cases, I tried running it AFTER appending the junk data 09:13 < darkmeson> pingfloyd: out of curiosity, does it have midori too? buster doesn't have that either 09:13 < pingfloyd> yep 09:13 < darkmeson> interesting 09:14 < domhnall> pocketmon: might be useful to check out sasm also. doubt it's in repos for debian though 09:14 < Elladan> nindustries, if the binary format is e.g. ELF (i.e. a C/C++ program), the program has a header that, in effect, points to the different pieces in a structured way. So some extra junk at the end is unlikely to matter. 09:14 < pingfloyd> Candidate: 0.5.11-ds1-4+b1 09:14 < storge> darkmeson: they'll find their way in. example: etherape wasn't in buster, now it is. 09:14 < nindustries> Elladan: In my test cases I used programs compiled from Golang 09:15 < Elladan> nindustries, I'm not familiar with how the golang runtime works. 09:15 < pingfloyd> that's the nature of Testing branch 09:15 < storge> yep 09:15 < Elladan> nindustries, as an example of something that would fail, a python script would become very unhappy if you appended random junk to the end. 09:15 < nindustries> I'm asking in #go-nuts , but thanks for the explanation Elladan 09:15 < nindustries> Ahh, yes. 09:15 < nindustries> Because that just contains a basic file header & 'text'? 09:16 < pingfloyd> nindustries: if you're trying to hack executables, learn assembler and how to use a hex editor instead. 09:16 < Elladan> Yes. So the random junk at the end would be parsed as runnable code. 09:17 < domhnall> yum..entropy 09:18 < darkmeson> pingfloyd: It's actually kind of funny because I'd used Slackware, then jumped to Redhat, then Mandrake (iirc), Ubuntu , and various others, but never Debian until recently 09:22 < pingfloyd> darkmeson: I've used many dists, but started and ended with deb. 09:22 < pingfloyd> all the others pale in comparison to me 09:22 < pingfloyd> so I always end up coming back to debian 09:23 < storge> likewise, me also 09:23 < pingfloyd> every now and then I go on a spree of checking out new dists coming out 09:23 < pingfloyd> I'm going to check out void next 09:23 < pingfloyd> it sounds interesting 09:24 < pingfloyd> and not using systemd sounds like a plus with it 09:24 < domhnall> pingfloyd: when trying out, do you usually install or try vm first? 09:24 < darkmeson> I guess I just tried to avoid deb-based distros as much as possible because the packages were so much more of a challenge to make and all of the limitations and a few of the philosophical choices made in-place upgrades excruciating 09:24 < storge> pingfloyd: in the last week i downloaded void, alpine, and a couple others 09:24 < pingfloyd> domhnall: usually installed. Now I got something with vt-d, so I'll probably play around in vm 09:25 < adtac> people use alpine for their desktop? o.O 09:25 < storge> adtac: i'm going to try it at least 09:25 < Elladan> The only distro I've really cared to play with in the last couple years was Fedora, when they got wayland working. 09:25 < pingfloyd> darkmeson: there packages can be a pain to make if you want to make official ones. 09:26 < pingfloyd> darkmeson: if you just need a package for your own uses, you can just use something like checkinstall 09:26 < darkmeson> None of that is really a problem anymore since I can easily make a chroot, container, or vm with a completely different distro to be lazy and not have to build any packages to begin with 09:26 < pingfloyd> or forego making a package to begin with and say round out of its source tree 09:26 < adtac> Elladan: for whatever reason, fedora's GNOME is always smoother than the vanilla GNOME you get from Arch's packages, no idea why 09:26 < pingfloyd> s/round/run/ 09:27 < darkmeson> Back when I was making debs, I mostly just cheated and made rpms, then used alien. or sometimes I'd use checkinstall 09:27 < Elladan> I don't really like gnome 3's UI, though I will say it seems tolerable in a way that unity... wasn't. 09:27 < pingfloyd> adtac: I think fedora does a lot of patching to gnome for it 09:27 < pingfloyd> adtac: fedora is pretty gnome bias 09:27 < adtac> pingfloyd: are those patches available to be applied in other gnome installs? 09:28 < Elladan> But I really just wanted to see the glorious wayland in operation :-D 09:28 < nindustries> Woops, network bork. 09:29 < surrealpie> how can i enable ssh in batch mode 09:29 < Elladan> I don't think I saw any obvious graphics tearing, for pretty much the first time since I was using an Amiga 500 in the '90s. So yay wayland? :-) 09:29 < pingfloyd> adtac: this is probably your best bet for that https://src.fedoraproject.org/cgit 09:30 < surrealpie> i tried uncommenting batchmod = yes , in /etc/ssh/ssh_config 09:30 < darkmeson> It still kind of annoys me because even today I have to go into /var/lib/dpkg/info every once in a while and move some .post* scripts out of the way temporarily to unwedge package operations 09:30 < darkmeson> It's worth it to avoid systemd on my baremetal installations though ;) 09:31 < domhnall> sounds like a job for *bsd 09:31 < darkmeson> Elladan: when you say gnome 3, are you talking about gnome shell? 09:31 < pingfloyd> Elladan: that's kind of what was a catalyst for wayland 09:32 < pingfloyd> Elladan: how X can't do vsync correctly 09:32 < Elladan> darkmeson, yes. 09:32 < Elladan> darkmeson, I mean the actual gnome libs are used by a lot of DE's. 09:32 < adtac> I have the weirdest bug ever in chrome + xorg: every fourth window I open will have a very weird screen render issue like that infamous XP bug where if you move your window around the border will create patterns 09:32 < adtac> *every* fourth window 09:33 < storge> adtac: it's all your fault 09:33 < darkmeson> They actually did finally cave and put the work into fallback mode that it needed to become a first class citizen again 09:33 < pingfloyd> YOU BROKE IT! 09:33 < adtac> gave me another reason to move to ff lol 09:33 < Elladan> adtac, I think your computer hates you. 09:34 < darkmeson> gnome shell was their unrealistic expectation that a desktop workflow would ever be able to be adapted to a tablet interface 09:34 < adtac> a piece of structured rock hates me, a new low in life 09:34 < darkmeson> so obviously it's going to be somewhat lacking if you're not using it on a tablet ;) 09:34 < pingfloyd> so bloated 09:34 < Elladan> pingfloyd, yeah. To be fair to X, I see painfully bad tearing on Windows and Mac too. :-) 09:34 < pingfloyd> I use dmenu 09:35 < pingfloyd> I don't want superfluous animations and decorations 09:35 < pingfloyd> I just want the program I want to run, to run without delay 09:35 < pingfloyd> in minimal keystrokes 09:35 < darkmeson> adtac: I hate to break it to you, but firefox isn't any better 09:35 < Aph3x-WL> yes it is 09:36 < pingfloyd> ^ 09:36 < darkmeson> I'm typing this through an app called Riot, and it manages to bring firefox to its knees by itself 09:36 < Aph3x-WL> because riot is garbage 09:36 < pingfloyd> ^ 09:37 < Elladan> I stopped using Firefox years ago when Chrome got to version, like, 0.0000001 on Linux, because it would segfault on about every 4th web page and take down the whole browser. 09:37 < ikonia> I see Crazytux came back for this weeks round of "what distro is best" or "what desktop is best" 09:38 < darkmeson> It's actually due for a reload even now because it freezes up for a good 20 seconds here and there despite ample available resources 09:38 < Elladan> ... while chrome lacked features like "a menu" and stuff, but had process isolation. These days they both seem like programs that function. 09:39 < Elladan> Heh, the modern web: A place where every web page you visit might try to burn out your CPU with a javascript bitcoin miner delivered through someone's hacked 4-level-deep ad network account. 09:40 < pingfloyd> the whole web-app concept is flawed 09:40 < darkmeson> Aph3x-WL: actually, it's the browser. It kept pegging a core on a single, main thread but even that went away in 60 and I have all of the other obvious things like hardware acceleration disabled 09:40 < storge> pingfloyd: do you use surf too? 09:41 < darkmeson> It's not the concept that flawed, it's the browser security 09:41 < pingfloyd> storge: no 09:41 < adtac> I hate to further the electron hate circlejerk, but what braindead person thought it'd be a good idea to have *multiple* browsers with separate runtimes to display websites without a url bar 09:41 < pingfloyd> storge: I don't like most suckless projects, but dmenu serves a good purpose well. 09:42 < darkmeson> Somehow browser developers think that enabling random third parties to run arbitrary code with minimal isolation is a good thing 09:42 < Elladan> darkmeson, surely you jest. Every page having permission to load a program onto your computer which initiates arbitrary TCP connections to localhost (or anywhere else), what could go wrong? 09:42 < adtac> pingfloyd: st is pretty cool, I'm running it 09:42 < pingfloyd> storge: maybe dwm is the other one that is okay 09:42 < pingfloyd> adtac: I don't like st for anything other than running roguelikes. 09:42 < darkmeson> adtac: electron had a pretty serious vulnerability very recently, actually 09:43 < darkmeson> https://copy.sh/v86/ 09:43 < darkmeson> 'nuff said 09:43 < pingfloyd> adtac: it's atrocious from a usability standpoint as a terminal emulator. (Doesn't even have scrolling) 09:43 < Aph3x-WL> darkmeson: the browser does what it's supposed to and it does it well. a poorly coded, poorly optimized, piece of software running inside that browser bringing it "to its knees" is the fault of the coders of the software running inside, not the browser itself 09:43 < pingfloyd> adtac: I think suckless got really myopic with st 09:44 < Aph3x-WL> it's meant to browse the web, not to run a nodejs OS inside it 09:44 < darkmeson> here's the github for that btw. https://github.com/copy/v86 09:44 < cmj> jesus 09:44 < pingfloyd> adtac: maybe not though. I think it probably is perfect for a minimalist with OCD. 09:44 < adtac> pingfloyd: not having scrolling was a design decision 09:44 < adtac> pingfloyd: they decided tmux filled that gap well enough 09:45 < pingfloyd> I know 09:45 < pingfloyd> but I think it is a terrible design decision since it throws aways bare essential functionality 09:45 < pingfloyd> I think they're drawing the line to tight. 09:46 < pingfloyd> I mean we could go all the way back to where you had to enter opcode bit by bit 09:46 < pingfloyd> so where does one draw the line? 09:46 < pingfloyd> I think basic taken for granted functionality is a good one 09:46 < adtac> there's a patch somewhere around that adds scrolling, if you're interested, but the suckless devs didn't take it in 09:46 < pingfloyd> adtac: I know 09:47 < darkmeson> Aph3x-WL: ok, let me make it a little less controversial. firefox still doesn't do proper multiprocess just yet, and it also still fails to manage extensions in an appropriate way 09:47 < pingfloyd> adtac: I like xfce-terminal better altogether anyway. ST is still great for playing roguelikes in text mode. 09:47 < adtac> pingfloyd: I get your point, but there *is* a market for st, and it fills that niche 09:47 < adtac> not that there's anything wrong with other terminals 09:47 < pingfloyd> adtac: yeah, I guess where they chose to draw the line is probably where that market likes 09:48 < pingfloyd> adtac: it's fine though, because it's not a big investment for me either way since I like xfce-terminal anyway. 09:48 < darkmeson> Just to qualify the latter point, it looked like a futex storm was going on in the thread pegging the core, which afair is the same damned thing I've always had happen with firefox, even under the old engine 09:48 < pingfloyd> I like its balance between a barebones VTE based terminal emulator and super bloated gnome terminal 09:49 < darkmeson> Extensions conflicting and stepping all over each other because they didn't have a clear and defined order of precedence, basically 09:49 < pingfloyd> darkmeson: I've had firefox freeze, but it seems to less often than other competing web browsers 09:50 < storge> ^ 09:50 < pingfloyd> I'd like to see mozilla iron that out the next few versions 09:50 < storge> and rarely if ever, anymore 09:50 < storge> (for me) 09:50 < pingfloyd> I'd rather they do that than add any new features really 09:51 < pingfloyd> I've had freezes on this laptop lately and I'm still trying to nail it down. 09:51 < pingfloyd> I just haven't had the time and energy to 09:51 < Elladan> I haven't had chrome freeze in a long time. 09:52 < pingfloyd> the main reason I don't use chrome is because its google 09:52 < Elladan> Yeah, that's a potential issue. 09:52 < pingfloyd> and google is shady with their practices on users 09:53 < paulcarroty> google is a gigantic privacy pomp 09:53 < Elladan> I trust them more than I'd trust other megacorps with my data, but it's getting less and less warranted. 09:53 < domhnall> heh, at least geolocation is off by default 09:53 < darkmeson> The bad part is that most of the extensions that I use are things that really should've been integrated right into the browser and UI long ago like uMatrix, ScriptSafe/NoScript, and Cookie Autodelete/Self-Destructing Cookies 09:53 < domhnall> for now...maybe? 09:53 < pingfloyd> who to trust less? Oracle, Microsoft, Apple, or Google? 09:53 < storge> i have better luck with chrome in mapping and gis sites, but i use firefox for basically everything else. and since i only use map/gis at work, firefox is essentially my first choice.. 09:54 < morf> i don't trust none of them 09:54 < morf> what are you mad? 09:54 < darkmeson> So it's pointless to even test without them because what'd I'd be left with was something completely unsafe for use anyway 09:54 < pingfloyd> it's like a Presidential Election 09:54 < paulcarroty> folks, what about window manages today? They all are died or something is active now? I'd like to use something fast and functional. 09:54 < morf> they are corporations ... they make money... they don't give a fuck 09:54 < darkmeson> s/was/would be/ 09:54 < pingfloyd> which crook do we trust the least 09:54 < storge> morf: +1 well they do, but their goals are not yours 09:54 < paulcarroty> to trust who doesn't sell private data 09:55 < morf> also let's remember one quote "Corporation: An ingenious device for obtaining profit without individual responsibility." 09:55 < pingfloyd> paulcarroty: have you tried icewm? 09:55 < pingfloyd> paulcarroty: are you needing some light, functional, and fast? 09:55 < storge> morf: that's exactly what i mean, well put. 09:55 < darkmeson> or lxde or xfce 09:56 < paulcarroty> pingfloyd, installed, but there's no clock and sound regulation 09:56 < pingfloyd> paulcarroty: that's going to be other programs 09:56 < paulcarroty> lxde/xfce is nearly to dead now 09:56 < pingfloyd> paulcarroty: DE's have such programs, but there are also standalone ones. 09:57 < pingfloyd> paulcarroty: you'd probably want to use pavcontrol or alsamixer in your case 09:57 < pingfloyd> *pavucontrol 09:58 < bartmon> how come xscreensaver has no fork that would bring some theming to it? it stands out like a sore thumb with its default appearance... 09:58 < darkmeson> there's also awesome 09:58 < bartmon> paulcarroty, see i3/sway/bspwm 09:58 < Elladan> bartmon, there are gnome and kde screensavers 09:59 < darkmeson> patches are welcome, I'm sure 09:59 < MrElendig> would rather rewrite the whole thing 09:59 < MrElendig> than try to patch ut 09:59 < MrElendig> it 09:59 < Elladan> But also, do people even use screensavers these days? I mean it seems like a waste of electricity. 09:59 < pingfloyd> xscreensaver has a lot of problems 09:59 < paulcarroty> well, sway use Wayland? interesting 09:59 < pingfloyd> which is unfortunate 09:59 < ansraliant> bartmon: how about awesome? isn't it active any more? 10:00 < ansraliant> it was like very popular a while ago 10:00 < paulcarroty> sreensaver is waste of time 10:00 < pingfloyd> I like ones that show feeds 10:00 < MrElendig> awesome isn't awesome 10:01 < paulcarroty> https://github.com/awesomeWM/awesome looks alive 10:01 < iflema> let it powerdown ffs... 10:01 < pingfloyd> there's one that grab random crap off the web. It's funny when porno ends up its collage 10:01 < Elladan> I like the idea of a tiling WM, but honestly I just make everything full screen and use workspaces. 10:01 < Elladan> ... which works in pretty much any WM that isn't terrible. 10:02 < pingfloyd> I use fullscreen too 10:02 < pingfloyd> I really don't ever need tiling for more than two windows at a time anyway 10:02 < Triffid_Hunter> I use a combo.. 6 desktops with most apps on each tiled into a quarter screen grid 10:02 < Elladan> The half-screen snap stuff tends to be fine for when I need 2 windows. 10:02 < darkmeson> I do the same, except my workspaces are mostly vnc sessions 10:03 < pingfloyd> Triffid_Hunter: doesn't that feel like sitting at work? 10:03 < Triffid_Hunter> pingfloyd: I am sitting at work :P 10:03 < pingfloyd> that many screens and it feels like I'm on security duty or something. 10:03 < Elladan> For terminals I want more of course, but you can do that inside the window with terminator or tmux or screen or whatever. 10:03 < adtac> I've been meaning to move to sway for over 4 months now, inertia's a bitch, I got too comfortable in gnome 10:04 < pingfloyd> Elladan: I just open new windows 10:04 < pingfloyd> Elladan: and manage it through the wm 10:04 < adtac> I basically only use 2 windows anyway; tmux with 5-8 windows and a browser 10:04 < pingfloyd> I really hate the extra layer of shortcuts with tmux and screen 10:04 < Elladan> Yeah, I use terminator. 10:05 < notmike> Slackware is still the best distro all-time 10:05 < darkmeson> I actually have 12 desktops atm 10:05 < bartmon> notmike, define best() 10:05 < Elladan> It seems to have shortcuts that are pleasing, compared to using a terminal-splitter like tmux 10:05 < notmike> bartmon: but any definition 10:05 < notmike> By 10:10 * iflema https://imgur.com/a/6Imlh0k 10:11 < paulcarroty> iflema, it killing my eyes 10:11 < paulcarroty> a lot of black color and small fonts 10:13 < pingfloyd> iflema: is the hinting that bad or is it because it is jpg? 10:13 < truthr> when will linux overtake windows as a desktop OS? 10:13 < iflema> png... scrot... never used it before ;) 10:14 < iflema> just hit and it went beep and made apng 10:14 < pingfloyd> on imgur its a jpg 10:14 < iflema> madness 10:16 < pingfloyd> https://i.imgur.com/439gPOc.jpg 10:16 < MrElendig> ptpb > imgurl 10:16 < iflema> muh new old keyboard has no volume controls... pez 10:17 < MrElendig> iflema: 2€ avr/stm board + a pot 10:17 < pingfloyd> this laptop has them in the f'n fn keys 10:18 < iflema> so does mine since 4.10 10:18 < iflema> just wont boot since 4.11 10:18 < iflema> lol 10:19 < iflema> i might look into it one day... 10:19 < iflema> just missed the LTS ;) 10:21 < ansraliant> don't you know what year it is truthr? 10:21 < akd> Hi 10:21 < ansraliant> it's the year of the linux desktop 10:22 < akd> I amhaving an erreor in a bash script 10:22 < akd> : 10:22 < akd> => /root/.npm/_npx/151/lib/node_modules/rollup-umd-scripts/bin/scripts-ci/cmds/../../../internals/declination/create: line 78: /root/oot/.npm/_npx/151/lib/node_modules/rollup-umd-scripts/bin/scripts-ci/cmds/../../../internals/declination/01-react: No such file or directory 10:22 < akd> I dont understanbd how 01-react cannot exist, it is in the same directory as create 10:22 < iflema> MrElendig: I got a Microsoft remote... they make great hardware... 10:22 < pingfloyd> ansraliant: I think truthr still lives in 1995 10:22 < akd> and create is throwing err L78 , when I try to run 10:23 < phre4k> I upgraded to Kernel 4.17 and now my monitor timings over DVI don't seem to be correct, I have a refresh rate of 59.95 working over DP but with DVI only 60.00 works. 10:23 < phre4k> how can I report this? I only get an error on my monitor, not in any journal/log 10:23 < ansraliant> pingfloyd: the windows commertials from that time were awesome 10:23 < iflema> MCE fing with a green button... and the DVD was scribing shit with light... once... 10:23 < pingfloyd> Start me up! 10:23 < pingfloyd> that one? 10:24 < pingfloyd> think they were trying to prove something to Steve Jobs when he said they have no taste. 10:24 < ansraliant> yes, that one 10:24 < pingfloyd> but Steve Jobs had Bono 10:25 < pingfloyd> so he shouldn't throw stones 10:25 < ayecee> i see what you did there 10:25 < ansraliant> https://www.youtube.com/watch?v=IY2j_GPIqRA 10:25 < ansraliant> i don't remember when this came out 10:25 < ansraliant> around 98 I guess 10:26 < proa> do we have to add entries in /etc/exports or /etc/fstab? 10:26 < pingfloyd> now I feel nauseous. Thanks 10:26 < proa> I am on centos. trying to setup nfs mounts 10:26 < ansraliant> there was also the one with jenifer aniston when she was hot and young. Now she is hot and old 10:26 < ayecee> proa: that depends a lot on what you want to accomplish 10:26 < pingfloyd> ansraliant: touche https://www.youtube.com/watch?v=V7PYQCXdX3A 10:27 < proa> ayecee: all i am trying to do is to setup a nfs storage between aws instance and my vagrant box 10:27 < ayecee> still vague 10:28 < pingfloyd> ansraliant: kind of like Microsoft except s/hot/gross/ 10:28 < ayecee> still no clue on which is mounting what 10:29 < proa> the storage on aws needs to appear in vagrant 10:30 < pingfloyd> proa: all you got to do is tell the computer to 10:30 < truthr> pingfloyd, 1995!? 10:30 < truthr> linux has not passed windows for desktop os, or has it? no it hasn't 10:30 < truthr> pingfloyd, great nick btw 10:31 < ayecee> but next year is the year of linux on the desktop 10:31 < pingfloyd> truthr: that's subjective 10:31 < storge> it's always next year 10:31 < truthr> it will happen 10:31 < truthr> it is coming 10:31 < pingfloyd> truthr: I think it has. I can do a lot more with it and have a lot more options. 10:31 < phre4k> the year of the Linux desktop is always $currentyear 10:32 < ayecee> + 1 10:32 < pingfloyd> what determines? 10:32 < phre4k> it works 10:32 < truthr> wait a minute...linux users do not seem to care if most users use windows 10:32 < pingfloyd> then it happened a long time ago 10:32 < truthr> only that they use it 10:32 < truthr> what about the Freedom of the masses 10:33 < ayecee> what about it 10:33 < pingfloyd> how do you give freedom to those that turn their nose at it? 10:33 < truthr> it is hard 10:34 < truthr> if you do not give freedom to the masses, you will eventually lose your freedom. i made that up just now. 10:34 < truthr> probably true in some form 10:35 < ansraliant> #philosophy 10:35 < ansraliant> haha 10:35 < ansraliant> my jokes are bad and I should feel bad 10:35 < ayecee> still doesn't make a year of the linux desktop 10:36 < truthr> i take it most here do not follow the 4 freedoms of Free Software 10:37 < ayecee> you take correctly sir 10:37 < ansraliant> weren't they 3 freedoms? 10:37 < truthr> 4 10:37 < ayecee> 5 10:37 < truthr> https://fsfe.org/freesoftware/basics/4freedoms.en.html 10:37 < ansraliant> 6 10:37 < truthr> get back to your roots 10:38 < ayecee> how's the view from that high horse 10:38 < truthr> pretty damned good 10:38 < pingfloyd> good luck making your system 100% free 10:38 < truthr> yeah, well that is almost impossible 10:38 < ayecee> the only thing the high horse lacks is a mirror 10:38 < pingfloyd> after you have only 100% free software on the system, the next can of worms is all the firmware. 10:39 < truthr> the goal is free software and hardware 10:39 < ayecee> and food 10:39 < pingfloyd> it's a good goal, eventually there will be better choices as long as their is the drive. 10:39 < truthr> yeah you have to make a living 10:40 < pingfloyd> you need groups like fsf and gnu to be a driving force to make headway in that direction. 10:41 < truthr> so support fsf 10:41 < ayecee> no u 10:41 < truthr> i do 10:41 < ayecee> well good, that's taken care of 10:42 < pingfloyd> did you pay via paypal? 10:42 < truthr> i see the universities are doing a bang up job and teaching students the fundamental importance of Free Software 10:42 < truthr> i paid, not saying how 10:42 < ayecee> lies 10:42 < ayecee> present receipts 10:43 < pingfloyd> did you send them a bag of weed or something? 10:43 < truthr> nah, i have been an associate member or some shit for years 10:43 * ayecee bets on some shit 10:44 < MrElendig> universities are so stupid, my local one forces all the students to use a closed source GIS suite, even though no one uses that software in the real life, almost everyone uses qgis instead 10:44 < pingfloyd> scorched 10:44 < ayecee> almost everyone of my friends* 10:44 < MrElendig> so not only are the students forced to pay 400€ (after 80% discount) for this crappy software, they don't learn the software stack that they will actually be using after graduation 10:45 < pingfloyd> the education racket 10:45 < ansraliant> after the goverment universities are a big pile of burocratic crap 10:45 < truthr> education has become for profit shit 10:45 < pingfloyd> put your product across school to win the hearts and minds of the youth 10:45 < MrElendig> the universities here are private 10:45 < truthr> they have sacrificed integrity 10:45 < ayecee> can't even spell bureaucratic 10:45 < pingfloyd> use their career as your sink or swim platform 10:45 < pingfloyd> instead of teaching them the defacto standard 10:46 < pingfloyd> what's best for them 10:46 < MrElendig> even in this communist state (according to fox news) 10:46 < im0nde> Hi I have a problem with the latest kernel from the arch linux repos. When I update to it, my keyboard (or maybe its the usb) doesn't work anymore. I have to chroot from a usb live system and downgrade to restore the functionality 10:46 < MrElendig> im0nde: post logs? 10:46 < MrElendig> also, sure you booted into the correct kernel? 10:47 < ayecee> what happens when you try 10:47 < ayecee> ok, fair enough, keyboard doesn't work anymore 10:47 < pingfloyd> doesn't work how? 10:47 < MrElendig> got a second device you can use to ssh in? 10:47 < ayecee> still would be useful to have dmesg output 10:47 < im0nde> MrElendig: I dont now how to do anything, except live stick 10:48 < im0nde> ayecee: can i get that from chroot? 10:48 < ayecee> im0nde: maybe yes 10:48 < MrElendig> journalctl -b -1 .... 10:48 < ayecee> from /var/log within the chroot 10:48 < pingfloyd> dmesg -w and then plug in the keyboard 10:48 < ayecee> wat 10:49 < im0nde> I cant really unplug the keyboard. Its a laptop 10:49 < ayecee> ofc 10:49 < MrElendig> wouldn't be supriced if this is a case of not having /boot mounted when he updated 10:49 < pingfloyd> great 10:49 < ayecee> doesn't seem likely 10:49 < im0nde> ok let me boot the chroot then. 10:50 < ayecee> normal boot with a usb keyboard may also be an option 10:50 < im0nde> ayecee: no, i tried that. 10:51 < im0nde> I'm starting to get the feeling it freezes at the login prompt (tty) 10:51 < im0nde> ok im in chroot now. 10:52 < pingfloyd> probably should report a bug to arch 10:52 < ayecee> ok, that's more different than keyboard not working 10:53 < im0nde> ayecee: As I said, its hard to tell. Usb devices seem to not work 10:53 < MrElendig> get the log from journalctl, see if ssh works 10:54 < azarus> Does ext4 make sense for standard flash USB drives? Does having a journal decrease its lifespan? 10:55 < MrElendig> no 10:55 < MrElendig> not in a noticable way 10:55 < MrElendig> specially not if you are just using it for file transfer 10:55 < azarus> Then one might use any journaled filesystem for it, for what it matters 10:56 < pingfloyd> how much does it really matter for a usb stick? 10:58 < im0nde> MrElendig: Can I start ssh from chroot? systemd start sshd refuses to run in chroot 10:59 < Metalsutton> Hi allow. I am having a little trouble following the circled steps: https://i.imgur.com/rIUrMc0.png , how do i go about doing this on a windows machine? 10:59 < truthr> ##windows 11:00 < MrElendig> Metalsutton: it literally tells you how 11:01 < MrElendig> Metalsutton: it tells you to use win32 disk imager 11:01 < MrElendig> Metalsutton: which has a whooping 3 buttons, one of which being exit 11:02 < MrElendig> hmm, they have added a cancel button too now 11:03 < MrElendig> Metalsutton: https://sourceforge.net/projects/win32diskimager/ 11:03 < MrElendig> and window mounts things when you plug them in 11:06 < han-solo> Hello guys 11:06 < han-solo> I happened to have wrote echo > /usr/bin/systemctl, and now i cannot use systemctl 11:06 < han-solo> Is there any way i can get the executable back ? 11:07 < MrElendig> yourpackagemanager install systemd 11:07 < Metalsutton> oh I use a DD writer called Rufus ... I thought i could somehow interchange it. 11:07 < han-solo> is it possible to copy one system's 'systemctl' to another machine ? 11:07 < MrElendig> no need, just use your package manager 11:08 < han-solo> MrElendig: Yes, it says systemd is already installed, and since no update,.... 11:09 < MrElendig> you can tell it to reinstall it anyway 11:09 < MrElendig> see the manual for your package manager 11:09 < han-solo> MrElendig: Yeah, i did. Installed package systemd-219-42.el7_4.10.x86_64 (from updates) not available. 11:10 < han-solo> MrElendig: Okay, i'll just update it then, i guess 11:10 < han-solo> Thanks 11:11 < MrElendig> dnf reinstall systemd 11:11 < han-solo> yum reinstall, didn't do anything. I did yum update systemd and worked 11:14 < Metalsutton> MrElendig, Maybe you didnt see the circled bit, the circled bit refers to uImage, I already have the .img file written to the sdcard and the BOOT partition avalible to me. 11:15 < im0nde> MrElendig: I have the log https://ptpb.pw/aeg_ 11:16 < MrElendig> Metalsutton: your cicle included writing the image too 11:17 < MrElendig> Metalsutton: so your problem is that you don't know how to copy a file on windows? 11:17 < catphish> i want to use ssh-keygen and output the key to stdout, is there any way to do this? 11:17 < catphish> i've been unable to work this out for myself, it may not be possible, i'm not sure 11:17 < MrElendig> Metalsutton: drag and drop the file, then press f2 to rename 11:17 < Metalsutton> I know how to copy. But what is uImage? its another file in the boot root folder. 11:17 < MrElendig> Metalsutton: a file 11:17 < Metalsutton> AHHHHHH. ok. 11:17 < MrElendig> file/directory 11:19 < Metalsutton> would it just make sense to rename whatever I want to use to uImage and Script.bin ? 11:19 < catphish> seems it's impossible :( wonder if i can get openssl to generate all the keys i need instead 11:19 < Metalsutton> Acheives the same result i guess. 11:20 < MrElendig> script.bin not Script.bin 11:37 < heeen> can you configure udp in a way to receive all broadcast datagrams on a socket that is bound form two processes 11:40 < heeen> actually, I just noticed this works for broadcast but not for non-broadcast 11:46 < demio> hey guys 11:46 < demio> having some trouble installing windows fonts on alpine 11:46 < demio> and having trouble finding online resources about it 11:46 < demio> im installing fontconfig@edge \ msttcorefonts-installer@edge \ ttf-dejavu@edge \ 11:46 < demio> and running update-ms-fonts && \ fc-cache -f 11:47 < demio> but when i compile GraphicsMagick 11:47 < demio> it doesnt detect the windows fonts 11:56 < BluesKaj> Hiyas all 11:57 < inthl> I have 2 physical hdds which are set up as raid 0 in a software raid, with ext4. is it generally possible to change the blocksize afterwards, aka without re-creating the fs? I already have tons of data on them on a production environment and would not like to cause downtime 11:59 < nostrora> Hello, i have a well customized ZSH in my linux. i want to know if i can use it when i ssh on another computer ? because i don't want to lost my zsh configuration 11:59 < MrElendig> copy it over to the remote 11:59 < MrElendig> I asume you mean .zshrc not that you patched zsh itself 12:02 < nostrora> MrElendig: yes 12:03 < azarus> inthl: no, you can't change that afterwards 12:03 < azarus> inthl: also, what's the current and new block size you want? 12:03 < azarus> normally, e2fsprogs chooses it perfectly 12:06 < inthl> azarus, I set it down from the default, 4096 to 2048, but I think that was a bad idea, since i/o wait is a little too high 12:09 < pentanol> hi all, could anyone suggest me channele for nodejs meteor react? 12:13 < pentanol> anyone around? 12:14 < Armand> Yes 12:14 < kerframil> pentanol: ##webdev, perhaps 12:14 < pentanol> ok, thanks 12:15 < kr_217> Hi all, need little help on tar command.I want tar a files which in /a/b/c.but when i do tar -czvf abc.tar /a/b/c/* it archive the directory path as well but i don't want that.I want to archive the files in /a/b/c? 12:16 < DLange> kr_217: https://unix.stackexchange.com/questions/24870/tar-files-only-no-directories 12:16 < han-solo> tar -C a/b/c -cvzf stuff.tar.gz . 12:17 < han-solo> not the '.' at the end, it's there for a purpose. Don't omit it 12:18 < kerframil> kr_217: it's unclear whether you are trying to avoid recursion of the sub-directories of /a/b/c or whether you are simply trying to including "/a/b/c" as the prefix for the archived pathnames. 12:18 < kerframil> to avoid including* 12:20 < kerframil> kr_217: if the former, +1 to kevin's answer in the stackoverflow thread (the one that uses find without xargs). if the latter, then as per han-solo's answer. 12:21 < azarus> inthl: premature optimizations? :< 12:26 < kerframil> kr_217: actually, the answers that recommend the use of --null and -T are safer, as long as you're using GNU tar (the use of find ... {} + could potentially execute tar more than once, if traversing a very large number of files). 12:28 < inthl> azarus, well, there was theory and there is practice 12:29 < azarus> inthl: in practice, the defaults get you the furthest ;) 12:29 < azarus> but no, can't change the blocksize of ext4 after creation 12:29 < azarus> nor do I know of any other filesystem that could do that 12:29 < inthl> I have 64G of RAM on a system with the mentioned lots of I/O, with about 20G of RAM in use for applications running. how come that there is always about 40G of ram free, ..should't that be used for metadata caching? 12:30 < azarus> inthl: well, do you even have that much metadata? 12:30 < inthl> gazillions of files 12:30 < azarus> or, is it even being accessed frequently enough to make sense to be cached? 12:31 < inthl> well surely about 100 opens/closes every second, over all those files 12:31 < searedvandal> if it's needed, it will cache it 12:31 < azarus> ^^ 12:31 < kerframil> inthl: in general, it means that you haven't been running for long enough and/or page faulted enough data for pagecache to grow sufficiently large. 12:31 < inthl> well but shouldn't it cache the metadata anyway in RAM, until the RAM is full, and probably unused become evicted? 12:31 < inthl> the host (2 of them, both behave similar) are up for 3 weeks by now 12:32 < azarus> inthl is feeling the opposite of this: https://www.linuxatemyram.com/ 12:32 < kerframil> it's not exclusively about uptime, but also the workload that the host is subjected to 12:32 < inthl> yes, the opposite 12:32 < azarus> his ram isn't being eaten and he wants it to be eaten 12:32 < inthl> exactly 12:32 < imado> . 12:32 < searedvandal> that's a first 12:33 < azarus> point of that website is: Everything is fine! (even if your kernel isn't caching as aggresively as much as you want it to) 12:33 < inthl> the workload is high I'd say. I just wonder what I am doing wrong here since I have the feeling that the metadata cache is almost not being used 12:34 < djph> if the HDDs (SSDs) are fast enough, the system may simply not need to store in RAM 12:34 < djph> I mean, yes, drives are still crazy slow in terms of storage ... but ... 12:34 < kerframil> inthl: free(1) will give you a quick overview 12:35 < djph> *but if the actual data movement isn't actually "to/from the disk"... 12:35 < inthl> the system has NVMe and HDDs for storage, where the bottleneck is. I would think that metadata caching would improve this, honestly I have other hosts with 128G of RAM, where over 100 is used for metadata caching 12:35 < kerframil> right 12:35 < inthl> and that one is like 99% idle 12:35 < inthl> the 2 I am talking about are about 30% idle, so lots of work to do, with gazillions of files, and almost no metadata caching, 40G or RAM free all the time 12:36 < inthl> 40 of 64 12:36 < kerframil> another scenario is that you run a primary application which uses O_DIRECT or another method to skip the cache; typically because they implement specialised caching stategies of their own. 12:36 < kerframil> but in most cases, it's simply because the workload doesn't demand it. "gazillions of files" is not a description of the workload. 12:37 < inthl> well I mean't that there are no 1000 files that are being operated with all the time, there are many out of a huge pool of files which are being accessed all the time. 12:37 < azarus> well, are there metadata changes? 12:37 < azarus> many metadata* 12:38 < inthl> changes..? well yes and no, not many 12:40 < inthl> there are users working on that files, also uploading and changing the structures, but I'd say this is really minor. 12:40 < inthl> on those files 12:40 < kerframil> inthl: in that case, I would expect the use of pagecache to gradually increase as time goes by. keep in mind though that access a file doesn't just shove the whole thing into cache. disk-backed pages (where a page is 4 KiB) are faulted, upon which they are incorporated into pagecache. 12:40 < kerframil> inthl: so it can be quite a granular thing, notwithstanding the readahead buffer. 12:40 < azarus> if you're that curious about how to control caching; I suggest reading up on https://www.kernel.org/doc/Documentation/sysctl/vm.txt 12:40 < azarus> but most likely, defaults are perfectly fine 12:41 < inthl> I already went over that txt 12:41 < azarus> cool 12:42 < inthl> my changes there were: dirty_ratio 90, dirty_background_ratio 80 and vfs_cache_pressure set to 1 12:42 < searedvandal> you could always try to optimize stuff with iosched and vm, but in most cases the defaults are good. 12:43 < inthl> well i tried noop, deadline and sticked to cfq, I see no significant results there. it surely simply might mean that the disks are too slow in general 12:45 < kr_217> kerframil: I want to avoid archived pathnames i.e. /a/b/c 12:46 < kerframil> kr_217: ok. just switch the working directory with -C and archive . then, as someone else mentioned. 12:48 < buoyantair> Whats a good sftp client for linux 12:49 < MrElendig> sftp 12:49 < Armand> Filezilla, for GUI 12:49 < ArlequInOut> without GUI, the command "sftp" is good, but with gui, filezilla is what you want 12:51 < MrElendig> your file manager 12:51 < MrElendig> most of them supports sftp 12:52 < kr_217> thanks kerframil and han-solo 12:52 < ArlequInOut> sshfs is good if you want to access file on sftp like file in file browser 12:55 < phinxy> Is scp and sftp related? 12:55 < MrElendig> partly 12:56 < MrElendig> they both is a part of openssh, but they work differenctly 12:56 < MrElendig> differently* 12:56 < MrElendig> uses the same transport but different protocols 12:56 < tempate> Hello. I've been working with Debian in a daily basis for a couple years now. I'm quite comfortable with it and have managed to get the hang of it quite well. Anyway, it's the only GNU/Linux distro I've really tried, so I'm kind of biased. I'm getting a new machine next week. Should I stay still with Debian or maybe try another distro? 12:57 < MrElendig> tempate: just fire up some virtual machines and play around 12:57 < tempate> alright 12:58 < tempate> sounds easy enough 12:58 < monotux> tempate, debian is nice but try Fedora if you'd like something rpm-based and more bleeding edge, or try NixOS if you want something that will really challenge you 12:58 < tempate> my current machine won't handle that most likely so I'll have to do so in the new one 12:58 < monotux> arch linux is nice if you don't have any other hobbies but fixing your computer after each upgrade 12:58 < MrElendig> mostly fud ^ 12:59 < azarus> arch is pretty stable, imo 12:59 < tempate> I honestly don't want to be banging my head each time I want to download something, monotux. I would like to learn more but not quite all the time, if you know what I mean. 12:59 < azarus> (while I was using it, which I don't anymore) 12:59 < tempate> what do you use now? 12:59 < azarus> not really relevant, and highly controversial 12:59 < monotux> tempate, in that case nixos might be a bit too much of a challenge, it is very different from most other stuff I've tried. 13:00 < MrElendig> tempate: try packaging some software not in the repos on debian, and then try it on arch 13:00 < tempate> I was thinking about Fedora but I'm not sure if the best thing to do is simply to jump into rp 13:00 < MrElendig> :p 13:00 < tempate> rpm* 13:00 < monotux> MrElendig, azarus, yes, FUD but I'm not all wrong :) 13:00 < tempate> monotux: yeah, that's what I was thinking 13:00 < monotux> AUR is really nice for simple packages 13:00 < oiaohm> monotux: when nix works it not that horrible. 13:00 < MrElendig> monotux: not talking about using aur, but making the packages yourself 13:00 < monotux> oiaohm, it just has a certain learning curve 13:01 < tempate> MrElendig: is it easier on arch? 13:01 < MrElendig> a hell of a lot easier 13:01 < monotux> oiaohm, I'm using nixos on most of my machines, servers, desktops and laptops, but it is a bit more cumbersome to handle non-packaged software 13:02 < azarus> I really like alpine linux, but I might be biased as I also am a maintainer of some packages there 13:02 < monotux> tempate, try FreeBSD if you'd like to learn something new, it is not as much of a learning curve as nixos/etc but it is a really nice piece of software 13:02 < monotux> unless you'd like to game 13:02 < azarus> Also, OpenBSD <3 13:02 < oiaohm> monotux: Linux from scratch I class as the worst end. 13:02 < monotux> oiaohm, I work with buildroot at work =) 13:02 < tempate> monotux: nah, not a gamer 13:03 < MrElendig> lfs is a book and some tools to help you create a new distro 13:03 < MrElendig> it is not a distro itself 13:03 < tempate> monotux: is it good for a day to day basis, though? 13:03 < oiaohm> monotux: so my define of not that horrible might be a little distorted 13:03 < azarus> Is there LFS, but with simple tools such as busybox and OpenRC? 13:03 < monotux> MrElendig, nixos is arguably the same as well 13:03 < MrElendig> re fbsd: portage could do with a big overhaul 13:04 < azarus> freebsd doesn't use portage 13:04 < oiaohm> MrElendig: when you pick up distribution someone has built from LFS and not documented what they have done its right up there in the horrible pile. 13:04 < MrElendig> monotux: nixos provides buildfiles/packages 13:04 < MrElendig> azarus: ports, whatever they call it 13:04 < monotux> tempate, yes, it is highly stable, most software runs there and if you could live without wayland you'll be very happy with fbsd 13:04 < azarus> MrElendig: yup. portage is strictly a gentoo thing 13:04 < MrElendig> azarus: cloned from ports :p 13:04 < oiaohm> MrElendig: particularly when they could not pick a package manager and is a managled beast of rpm,deb, tgz and something script nasty. 13:04 < azarus> FreeBSD has wayland, for the "me too" rights 13:04 * azarus sticks with X11 13:05 < MrElendig> oiaohm: even better, when they decided that package managers are bloat 13:06 < oiaohm> MrElendig: decided package manager is bloat is better than package managers with each other not depenacy solving properly having to use force and pray. 13:06 < MrElendig> make install is even more messy, because then you have nothing at all 13:07 < MrElendig> no idea what flags they used etc 13:07 < adsc> you can redirect the output of all of your commands to files and save those as documentation 13:07 < MrElendig> atleast with a package manager you hopefully have the specfiles etc 13:07 < oiaohm> MrElendig: worse thing about those without dependancy solving properly you by mistake run update on the package manager and automatically uninstalls a big block of the system attempting to fix dependancy problems. 13:07 < adsc> save it together with your bash history 13:08 < oiaohm> MrElendig: installed by make install is safer than mixed package managers a sneeze away from deleting most of the system. 13:08 < MrElendig> oiaohm: bumblebee issue 123 13:08 < MrElendig> *cough* 13:10 < oiaohm> MrElendig: that is a fairly funny bug https://github.com/Bumblebee-Project/Bumblebee/issues/123 You attempt to change graphics drivers and it loses the mouse. 13:12 < MrElendig> and it deletes /usr 13:12 < MrElendig> https://github.com/MrMEEE/bumblebee-Old-and-abbandoned/issues/123 13:14 < oiaohm> MrElendig: so both bumblebee 123 bugs are quite wacky just that one is kind of worse. 13:15 < MrElendig> patch thread is sort of fun to read too https://github.com/MrMEEE/bumblebee-Old-and-abbandoned/commit/a047be85247755cdbe0acce6f1dafc8beb84f2ac 13:16 < MrElendig> sadly the best image in that thread is gone due to silly hosting 13:16 < oiaohm> MrElendig: yes stuff like that in high custom distributions are quite normal. One of the worst was finding that by alias tar was set to --remove-files and add to achive kind of made it fun if you attempted to extract content. 13:16 < MrElendig> sidenote: linux kernel had a similar bug for years 13:17 < MrElendig> but no one had ran the cleanup script as root and therefor no one noticed it for ages 13:17 < oiaohm> MrElendig: there have been about 20 different bugs I know of in Linux kernel history that would nuke /usr 13:17 < oiaohm> MrElendig: mind you my time with Linux starts in 1995 13:17 < oiaohm> MrElendig: roughtly 1 every year for a while. 13:17 < MrElendig> one of the make clean scripts would really clean everything, including / 13:18 < oiaohm> Yep something did exactly what it said it would. 13:18 < oiaohm> Just no user ever wanted that. 13:18 < oiaohm> You hope. 13:19 < oiaohm> MrElendig: was that the old total_clean option. 13:19 < oiaohm> MrElendig: that seamed like it had been added as a really scary joke. 13:19 < MrElendig> can't remember which of the targets it was 13:20 < oiaohm> I cannot think of one in recent years. 13:21 < oiaohm> Hopefully no one can add another one. 13:22 < MrElendig> was ages ago, too bad the lkml is pretty much impossible to search 13:23 < w3b573r> hi 13:24 < Sveta> hi w3b573r 13:25 < w3b573r> hi sveta 13:27 < babloo> Hi 13:28 < babloo> will any be able to help me in centos 7 installation 13:28 < MrElendig> #centos might 13:29 < babloo> iam not allowed there not sure it shows banned 13:30 < root__> whois 13:31 < Sveta> oops 13:32 < Sveta> babloo: use a non-webchat irc client 13:32 < Sveta> babloo: such as quassel irc 13:32 < Sveta> babloo: centos channel banned webchat, because it was abused there, alas 13:32 < root__> hai 13:33 < searedvandal> hey root 13:33 < babloo> oh ok 13:33 < babloo> thanks 13:33 < MrElendig> babloo: just use a proper irc client instead of the web one 13:33 < babloo> i am reusing IRC after 5 years now 13:33 < MrElendig> babloo: hexchat or weechat 13:33 < Sveta> welcome back 13:34 < babloo> after i left linux i didnt use irc :) 13:34 < babloo> good to be back 13:34 < babloo> thanks MrElendig 13:34 < MrElendig> babloo: centos bans the web client probably due to all the misuse by trolls/spammers and the like 13:34 < babloo> thanks Sveta 13:37 < babloo> ok thanks bye 13:39 < zapotah> irssi 13:39 < zapotah> join us, we have perl 13:43 < mawk> no zapotah 13:43 < mawk> everyone knows that weechat is better 13:43 < mawk> stop trying to confuse the minds of our youth 13:44 < djph> irssi > weechat :P 13:44 < djph> vim > nano 13:44 < djph> uhhh 13:44 < djph> sysvinit > systemd 13:44 < mawk> my android teacher is making vim propaganda discretly alongside her class 13:44 * djph goes to look for more things to start a holy war 13:45 < mawk> in the final assignment git repository I have to submit a .vimrc tailored for android system programming 13:45 < mawk> but I use emacs 13:46 < monotux> that sounds like a strange assignment 13:46 < monotux> but if she says so just do it 13:46 < mawk> it's not the only thing to submit 13:46 < darsie> My midnight commander ssh link timed out during computer suspend and now I get Cannot chdir to "/sh://pi@192.168.0.199". Any other way other than restart mc? 13:47 < djph> it's strange that she'd ask for a vimrc 13:49 < AndroidKitKat> mawk: can you use a vim plugin manager? 13:49 < AndroidKitKat> Just grab a bunch of things from vim-awesome 13:49 < mawk> yeah probably 13:49 < AndroidKitKat> and I always include this in my .vimrc 13:49 < AndroidKitKat> let me get it 13:49 < AndroidKitKat> https://www.irccloud.com/pastebin/4pudWQWS/ 13:50 < AndroidKitKat> pretty handy imo 13:50 < AndroidKitKat> lets you pick up from where you left odd 13:50 < AndroidKitKat> off 13:51 < mawk> I see 13:51 < mawk> thanks 13:51 < AndroidKitKat> np 13:51 < Exterminador> hi guys! stupid question: folder /etc/letsencrypt/live/domain.com/ is owned by root. I'd like to give read permission only to another user too. what is the correct chmod? thanks in advance 13:51 < AndroidKitKat> other than that, what lang are you writing the apps in? 13:51 < AndroidKitKat> Java or Kotlin? 13:53 < AndroidKitKat> I think you need to add that folder with read access to a group, and then add that user to that group 13:53 < AndroidKitKat> I couldn't tell you how to properly do it 13:53 < mawk> we're not writing apps I think 13:53 < mawk> I never went to that class 13:53 < mawk> but it's system programming 13:53 < mawk> knowing that teacher it is in C 13:56 < djph> Exterminador: 755, files inside should be 644. 13:56 < AndroidKitKat> mawk: you in uni? 13:56 < Exterminador> djph: tks! 13:56 < djph> assuming they're PUBLIC certificates. Private keys should be 600 13:56 < mawk> no AndroidKitKat , engineer school 13:57 < AndroidKitKat> cpeg? 13:57 < AndroidKitKat> or cs? 13:57 < AndroidKitKat> im CS 13:57 < AndroidKitKat> was going to do cpeg but i hate learning EE 13:57 < mawk> lol 13:57 < mawk> I don't know these acronyms, I'm french 13:57 < AndroidKitKat> ah 13:57 < AndroidKitKat> Computer Engineering? 13:57 < mawk> but it's not computer science 13:57 < AndroidKitKat> ah 13:58 < mawk> well there is a bit of that, tho 13:58 < AndroidKitKat> Computer Engineering is Comp Sci + Electrical Engineering 13:58 < AndroidKitKat> bad stuff 13:59 < Exterminador> djph: well, for that specific user I need that privkey.pem be readable too. and as I'm the only one using the machine, I don't think that's a security problem 13:59 < mawk> but it's more differentiated in favor of programming 13:59 < mawk> I took the embedded computing major 14:00 < mawk> title is "real-time and embedded systems" 14:01 < azarus> "remember, real time is not about performance!!" 14:01 < xinobi> sudo rm -rf is not working I get Directory not empty any advise? 14:02 < JimBuntu> xinobi, are you IN the directory that you want to remove? 14:02 < xinobi> JimBuntu: yes, I'm using bash but this is osx 14:02 < xinobi> I guess it's relatively different 14:03 < JimBuntu> xinobi, then you want to move out of it... so, `cd ..` then try to rm -rf /path/to/dir` 14:03 < azarus> asking questions about mac OS in ##linux... genius 14:03 < xinobi> but sudo rm -rf it should work 14:04 < xinobi> JimBuntu: I'm out of the directory I'm not that noob ;) 14:05 < xinobi> I'm just curious why it doesn't work I've checked the permissions and ownership as well everything ok 14:05 < djph> Exterminador: then make privkey owned by that user (e.g. www-data) 14:06 < xinobi> I can delete other files, but this one is in .Trash and it's undeletable 14:06 < djph> Exterminador: ... or rather, make apache start as root to load the keys, then drop back to whatever user 14:06 < Exterminador> djph: the problem is that I also need it to be owned by root :) 14:06 < JimBuntu> xinobi, I am looking at the man page for it. I see -d to attempt to remove directories as well 14:06 < djph> Exterminador: yeah, brain-fart on the file ... need more coffee 14:07 < Exterminador> djph: probably the best approach would be to copy the content of the folder to said user and chown the copied files? 14:07 < xinobi> JimBuntu: sudo rm -d blender-2.78a-OSX_10.6-x86_64/ -> rm: blender-2.78a-OSX_10.6-x86_64/: Directory not empty 14:08 < xinobi> JimBuntu: even: sudo chflags noschg blender-2.78a-OSX_10.6-x86_64/ followed by sudo rm -rf ... it's completely crazy 14:09 < JimBuntu> xinobi, I had no issue removing a directory with a directory inside it via 'rm -rf NAME' just now, in Darwin/macOS. Agreed, something else is going on 14:09 < xinobi> JimBuntu: completetly 14:10 < Exterminador> could be the trailing slash? ¯\_(ツ)_/¯ 14:10 < xinobi> Exterminador: nope 14:11 < djph> Exterminador: no, keep as tight of a lid on your private keys as possible. 14:11 < Exterminador> djph: noted 14:12 < djph> Exterminador: why's "not root" need to read the file anyway? 14:13 < Exterminador> djph: it's an IRC server and I need it to read the fullchain.pem and privkey.pem to have valid TLS connection 14:15 < Exterminador> although I'm the only user with access to that machine, I was trying to avoid to change much parameters regarding reading sensitive files 14:15 < djph> normally, a server should fire up as root to read stuff like that, and then fall back to not-root once that's done. 14:16 < Exterminador> well, IRCd won't start as root even 14:16 < djph> should be, iirc, suid 14:17 < Exterminador> djph: I'll try the 755 to the folder and 644 to the files inside it and if that works, it's good enough 14:18 < Dan39> Exterminador: i've done exactly what you want with my ircd, let me look what i did 14:18 < Dan39> definitely doesnt need suid or start as root, i run as normal user 14:18 < Exterminador> Dan39: cool! lemme know as soon as you know what you've done exactly 14:18 < Exterminador> :) 14:19 < elichai2> Hey, I have something weird going on. I have a tcp socket with subscription. now if I can both server&client on my machine it works great but if the server is on AWS EC2 instance (the port is open), it get's timedout. using tcpdumped I noticed the client sends a SYN and recives ACK-RST. anyone have any idea why? 14:19 < elichai2> ( that's the python code: https://paste.pound-python.org/show/xUyUdeKUIWchvTnoPux5/) 14:20 < Dan39> Exterminador: i have the cert and key o+r... 14:21 < Dan39> but if you are on a shared system that probably isn't a good idea to have the key o+r 14:21 < unyu> Sorry for the noob question. Is it possible to install several versions of freetype and fontconfig side by side? 14:21 < Exterminador> well, it's a VPS, and I'm the only one with access to it via pub keys (no password login whatsoever) 14:22 < MrElendig> unyu: depends on the distro, but why? 14:22 < Dan39> Exterminador: what distro? ACLs turned on? 14:22 < Exterminador> Dan39: Ubuntu 18.04 LTS 14:22 < Dan39> Exterminador: thatll probably be a no by default then 14:23 < Dan39> Exterminador: i wonder if it would be fine to change the file's group to your user 14:23 < Dan39> not sure how picky certbot is 14:23 < unyu> MrElendig: The reason is that Mathematica doesn't work with the latest version of freetype and fontconfig on Arch's repositories. I know it has been quite a while since the freetype and fontconfig packages were updated, but that means I have had that problem for that long a while. 14:23 < Exterminador> Dan39: I suppose not, since the web chat will run under another user and will need to read the same files 14:23 < MrElendig> unyu: use a container 14:23 < unyu> Oh. 14:24 < Dan39> Exterminador: then create a new group 14:24 < MrElendig> unyu: instead of having to pollute your system with old insecure software 14:24 < Exterminador> I like to run each thing under a different user (IRCd, webchat, and so on) 14:24 < Dan39> Exterminador: create an "ircdkeys" group and add both the ircd and webchat users to it? 14:24 < unyu> MrElendig: That makes sense. I will read more about those. Thanks for the pointer. 14:24 < Dan39> then set the key's group to that...? 14:24 < Exterminador> Dan39: that could work, probably 14:25 < Dan39> the cert doesnt really matter, doesn't have to be secure 14:25 < MrElendig> unyu: or yell at them to get them to update, or switch to something free 14:25 < Exterminador> yeah, my concern is the privkey 14:25 < Dan39> Exterminador: the concern is a service running as another user, like httpd, getting access to the key 14:25 < unyu> MrElendig: I would love to use a free software alternative, actually. But my professors distribute Mathematica notebooks. 14:26 < azarus> gnu octave? 14:26 < unyu> Octave is more of an alternative to MATLAB. And I indeed use Octave whenever my professors would use MATLAB. 14:27 < azarus> Sorry, didn't know :( But cool that you use it! 14:29 < Exterminador> thanks for the help guys! very appreciated, as I'm a noob regarding VPS management. ;) 14:29 < Dan39> Exterminador: get it working? 14:30 < Dan39> Exterminador: ive found letsencrypt kind of a pain with ircd since it has to be renewed every 3 months 14:30 < cristian_c> hello 14:30 < Dan39> but you can automate it depending on the ircd 14:30 < Dan39> hi cristian_c 14:30 < cristian_c> I've built and installed easycap smi driver (kernel module) 14:30 * Dan39 gives cristian_c a gold star! 14:31 < pxfgod> Linux uses a 1-1 threading model, with (to the kernel) no distinction between processes and threads -- everything is simply a runnable task. IS THAT STATEMENT ABOVE TRUE IN NOWADAYS' KERNEL? 14:31 < azarus> yup 14:31 < cristian_c> I've got driver sources from github repository (via AUR on archlinux, and via github in ubuntu) 14:32 < cristian_c> I've also loaded firmware too (in addition to driver loading) 14:33 < cristian_c> so, I've started vlc and I've connected easycap dongle to input source and to usb port 14:34 < cristian_c> if I look at dmesg output I see a large amount of messages (all them almost the same message) 14:34 < Dan39> is this going anywhere cristian_c ... 14:36 < cristian_c> 'smi2021 Skip broken frame N line, but need 240 in current 480 height' 14:36 < cristian_c> where N is an integer number 14:37 < cristian_c> vlc screen is always blank, btw 14:39 < cristian_c> so, how could I solve this issue, in order to grab video from the device? 14:39 < cristian_c> any ideas? 14:52 < Sabotender> hey MrElendig are you around? 14:52 < MrElendig> sort of 15:00 < ALowther_> Hello, how can I prioritize network configurations. I am connected to ethernet and wifi, of which, only the ethernet is connected to WAN. But for some reason pages are failing to load, I assume because my machine is trying to route over wifi which is connected to nothing. How can I tell it to route everything over eth? I have already checked Metric in 'route -n' and eth is lower. 15:02 < GodOfsea> Hey 15:02 < tyzoid> ALowther_: I'd just delete the routes over wifi 15:02 < tyzoid> check your default route via `ip route`, then add a new one for eth 15:03 < tyzoid> The default one is probably over your wifi connection 15:03 < GodOfsea> Something weird is going on in my linux 15:06 < ALowther_> tyzoid: Cool, thanks 15:07 < GodOfsea> https://paste.linux.community/view/0f80dfc6 15:08 < GodOfsea> The last stanza is unexpected 15:08 < GodOfsea> I get the last stanza everytime I do it 15:09 < tyzoid> GodOfsea: Should still work if you put it into the bind config 15:09 < tyzoid> I get the same thing you do 15:09 < tyzoid> but since I use an external dns provider, I remove the newlines / quotes and put it into a textbox on a web interface 15:10 < GodOfsea> Well I am using linode , as DNS manager , it doesnt support double quotes in records 15:10 < GodOfsea> So I just , place the 2nd stanza right beside the 1st 1 ? 15:10 < GodOfsea> With just a space ? 15:11 < tyzoid> Well you need it to be a txt record 15:11 < tyzoid> so grab everything from v=DKIM1; 15:11 < tyzoid> and stop at the close paren 15:11 < GodOfsea> yes , its dkim , so its text record 15:11 < tyzoid> delete newlines and quotes 15:12 < GodOfsea> ohh ok and what about 5 spaces ? 15:12 < GodOfsea> in between them 15:12 < tyzoid> GodOfsea: Doesn't really matter, but I'd clean it up. 15:13 < tyzoid> for comparison, you can see my record: dig mail._domainkey.tyzoid.com TXT 15:14 < GodOfsea> where's that other double quoted stanza after p parameter ? 15:14 < GodOfsea> its not there 15:15 < Sabotender> MrElendig: do you know how to secure erase a nvme drive under linux? 15:16 < GodOfsea> tyzoid: you mixed them ? 15:16 < Sabotender> I was getting ready to pack this thing up and send it back for warranty replacement, when I discovered some sensitive data on the drive. I thought I had moved everything OFF this drive (stupid dropbox not going where I want it to) 15:17 < Sabotender> MrElendig: HDPRAM doesn't work with this type of SSD 15:17 < tyzoid> GodOfsea: No, it's there 15:17 < Sabotender> it says 'inappropriate ioctl for device' 15:17 < MrElendig> nvme do have a secure erase, how secure it is varies from device to device 15:17 < tyzoid> GodOfsea: Bind will concatenate those strings for you. Since you're not using bind, you need to concatenate the strings to put into your dns host 15:17 < Sabotender> I was trying to do some research and I came across the command 'fstrim' but I dont know 15:18 < Sabotender> MrElendig: the problem is that ADATA does have a vendor specific applictation for such erasures, but surprise surprise...windows only 15:18 < GodOfsea> tyzoid: https://paste.linux.community/view/c617d826 15:18 < GodOfsea> Ok 15:18 < GodOfsea> I will try that 15:18 < MrElendig> Sabotender: and how much do you trust adata.... 15:19 < Sabotender> not very much since this device I haven't owned not even six months and it has failed 15:19 < MrElendig> other than that nvme-format -S .... 15:19 < MrElendig> see the man page for details 15:20 < MrElendig> (part of nvme-cli) 15:21 < GodOfsea> tyzoid: all the "\" are new lines ? 15:21 < W3B573R> hai 15:22 < tyzoid> GodOfsea: Where are you seeing backslashes? 15:22 < Sabotender> no manual entry for nvme-cli 15:22 < GodOfsea> dig mail._domainkey.openapprentice.org TXT 15:22 < GodOfsea> Try this 15:23 < tyzoid> GodOfsea: Yeah, you needed to remove the quote characters 15:23 < MrElendig> Sabotender: man nvme-format 15:23 < GodOfsea> tyzoid: quote characters are " right ? 15:24 < MrElendig> W3B573R: you should not irc as root 15:24 < GodOfsea> I removed them , but looks like it seperated the key at "/" 15:24 < Sabotender> no manual entry for nvme-format 15:24 < MrElendig> Sabotender: helps if you actually install it 15:24 < MrElendig> it is a part of the nvme-cli project 15:25 < MrElendig> install nvme-cli 15:25 < MrElendig> man page is also available online 15:25 < tyzoid> GodOfsea: Not sure what you mean, you don't need any quote or newlines 15:26 < tyzoid> the key is base64 encoded 15:26 < GodOfsea> PM ? 15:26 < jim> AM? FM? 15:27 < MrElendig> LSB/USB? 15:27 < MrElendig> CWD? 15:27 < GodOfsea> Yo jim :D 15:28 < jim> those are TLAs, not TLAs... 15:28 < xoxo> https://www.hastebin.com/juxiricite.pas <- can someone tell me how to run this command ? 15:28 < Armand> So long as they aren't SLAs 15:28 < xoxo> line 1 is my attempt 15:28 < GodOfsea> what ? 15:28 < jim> GodOfsea, hi, how's it going 15:28 < qrvpzvb> so, systemd fails to boot when I have a disk connected 15:29 < GodOfsea> Good , no exams , working on a opensource project 15:29 < leftyfb> xoxo: "new" needs to be a folder 15:29 < GodOfsea> openapprentice.org 15:29 < GodOfsea> trying to setup email servers 15:29 < MrElendig> qrvpzvb: post errors 15:29 < leftyfb> xoxo: also, you need to specify the first and last page to extract. Just like the message tells you 15:29 < MrElendig> qrvpzvb: smells like you not using UUID/LABEL 15:30 < xoxo> leftyfb: new is a folder 15:30 < qrvpzvb> MrElendig: I can't it wents into emergency mode 15:30 < MrElendig> so? 15:30 < qrvpzvb> MrElendig: also, that disk is not in fstab 15:30 < MrElendig> you can post from there 15:30 < xoxo> leftyfb: java -jar pdf2table_win.jar pdf2table laa.pdf new 3 3 still doesnt work 15:30 < MrElendig> you can inspect which disk is which label 15:30 < jim> xoxo, looks like you did run the jar, but then the jar (not java itself) threw a usage message 15:30 < xoxo> oh 15:30 < MrElendig> s/label/device in \/dev/ 15:30 < MrElendig> are you using uuid in fstab/bootloader? 15:31 < qrvpzvb> it's not showing up 15:31 < qrvpzvb> at all 15:31 < xoxo> jim: how do i fix that 15:31 < jim> xoxo, is 'new' an existing dir? 15:31 < xoxo> yes 15:31 < MrElendig> what is the full error? 15:32 < qrvpzvb> I think it's just the fsck that fails 15:32 < jim> and laa.pdf is an existing pdf in the current dir? 15:32 < qrvpzvb> though, that disk is NTFS and I have no fsck for that 15:32 < xoxo> jim: yes 15:33 < MrElendig> normally it won't automount random disks unless you have configured to do so 15:33 * MrElendig suggests posting the error 15:33 < qrvpzvb> i'll have to reboot to see it 15:34 < jim> ok, maybe try specifying f and l 15:34 < xoxo> jim: fixed 15:34 < jim> the usage message say they're optional, but maybe not 15:34 < jim> xoxo, how did you fix it? 15:35 < Sabotender> nvme admin command error: invalid_format(10a) 15:35 < forgon> Does someone know a good link listing typical Linux file locations for various purposes in depth? 15:35 < xoxo> jim: the pdf wasnt in the folder :p 15:35 < forgon> for example /usr/bin /usr/local/share /opt /etc ... 15:36 < jim> oh ok :) (that's not your fault, the command should have said "can't open file" probably (would have been quicker for you to figure it out) 15:36 < qrvpzvb> it didn't happen this boot... 15:36 < jim> ) 15:37 < jim> forgon, maybe the filesystem hierarchy standard? 15:37 < qrvpzvb> right, I think it's the drive order 15:37 < qrvpzvb> I should use UUID I guess 15:37 < triceratux> forgon: https://wiki.debian.org/FilesystemHierarchyStandard ? 15:37 < Sabotender> MrElendig: i get that error above 15:38 < MrElendig> Sabotender: would have to talk to the hw maker then 15:38 < azarus> forgon: http://refspecs.linuxfoundation.org/FHS_3.0/fhs/index.html 15:38 < azarus> the "Filesystem Hierarchy Standard" 15:38 < MrElendig> if you used the correct flags for nvme-format 15:38 < Sabotender> i see 15:38 < MrElendig> even when it works it might not be quite as secure as you think 15:38 < Sabotender> I used this : nvme format /dev/nvme0 --namespace-id=1 --ses=1 --pi=1 15:39 < MrElendig> some devices claim they securely erased all the things and then didn't 15:39 < forgon> \jim \triceratux \azarus: Thanks. 15:39 < MrElendig> eg wd had some drives that would throw away the key, but there would still be a "backdoor" key hardcoded in the chip....... 15:39 < MrElendig> which ofcourse people discovered 15:40 < triceratux> https://www.linux.com/learn/look-filesystem-hierarchy-standard-30 15:40 < azarus> forgon: why the backslashes? :P 15:40 < MrElendig> forgon: also https://www.freedesktop.org/software/systemd/man/file-hierarchy.html 15:41 < Sabotender> this is very frusturating 15:41 < MrElendig> forgon: for a more modern approach 15:41 < MrElendig> Sabotender: just zero it the old way? 15:41 * Sabotender shrugs 15:41 < MrElendig> though some drives might optimise away all 0's 15:41 < Sabotender> dont know how 15:41 < djph> drillbit through the platter? 15:41 < MrElendig> dd if=/dev/zero of=.... 15:41 < Sabotender> also I was told that doesnt work for solid state discs 15:41 < djph> ohwait, nvme ... drillbit through the chip(s) 15:42 < MrElendig> djph: data have been recovered from disks that have been shot multiple times so..... 15:42 < MrElendig> more reliable on ssds 15:42 < MrElendig> than on hdds 15:42 < Sabotender> run the platters though a wood chipper 15:42 < Sabotender> :-P 15:42 < djph> MrElendig: oxy-acetylene? 15:43 < MrElendig> induction heater 15:43 < qrvpzvb> MrElendig: ok, so the problem was that the drive order was messed up... I could UUIDs, but for now, I just changed the SATA ports around 'cause I'm lazy 15:43 < MrElendig> qrvpzvb: use uuid, takes no time to convert 15:44 < MrElendig> you can actuall pipe output from findmnt to fstab and it would work 15:44 < MrElendig> (with the correct flags) 15:44 < qrvpzvb> actually, my fstab does use UUIDs 15:45 < qrvpzvb> it's probably my syslinux setup 15:45 < MrElendig> root=UUID=whatever 15:47 < MrElendig> https://www.youtube.com/watch?v=fUiPBTnSNuk 15:47 < kerframil> note, UUID is supported natively in /etc/fstab, but not the kernel 15:47 < MrElendig> any modern kernel accepts UUID just fine 15:47 < MrElendig> (unless you did silly things when building it yourself) 15:48 < MrElendig> can also use /dev/disk/by-id/..... 15:48 < MrElendig> s/id/uuid/ 15:48 < kerframil> that's not true. it requires functionality implemented in the initramfs. the kernel only understands PARTUUID. I realise that many mainstream distros build initramfs images that do exactly that, but this is still an important distinction. 15:49 < MrElendig> kerframil: statement still stands, there are very few reasons for not using a initramfs these days 15:49 < kerframil> sure, if it works in your distro, go ahead and use it. I've no qualm with that. it's just useful to understand how things work. 15:50 < jim> and some dists build an initramfs when you do things like install a kernel (and update that initramfs when you do certain other things) 15:50 < kerframil> formats understood by the kernel itself are detailed in a comment in init/do_mounts.c 15:51 < kerframil> (and /dev/disk/by-id/* isn't one of them either) 15:51 < w3b573r> hi 15:51 < jim> and, this is an argument about where something is implemented... :) 15:51 < jim> hi 15:52 < kerframil> it's not an argument. a "modern kernel" does not accept UUID. this is a fact. it doesn't even understand what UUID means. anyone can read do_mounts.c to verify that. the initramfs has to do the lifting in that case. that's all. 15:53 < w3b573r> any one know hacking ssh 15:53 < Iridos> I am building a program and icc complains that it cannot find unistd.h. That's in /usr/include… shouldn't that come from some default include path? And if so, how would one determine that default include path 15:53 < jim> w3b573r, like what? 15:53 < leftyfb> w3b573r: ttps://youtu.be/KEkrWRHCDQU 15:53 < leftyfb> w3b573r: https://youtu.be/KEkrWRHCDQU 15:55 < Gnjurac> can i use ls 15:55 < Gnjurac> and grep as input parementer 15:56 < Gnjurac> sudo dpkg -i asInput ls | grep *.deb 15:56 < leftyfb> Gnjurac: sudo dpkg -i *.deb 15:57 < Gnjurac> i figured out zilion time when i do stuff in terminal i would be able to do it faster whit grep 15:57 < Gnjurac> leftyfb: does *. work only for dpkg ? 15:57 < w3b573r> leftyfb> https://www.youtube.com/watch?v=KEkrWRHCDQU great 15:57 < Gnjurac> can i example mousepad *.txt and it will open evry txt in folder or what? 15:58 < Gnjurac> will try ti 15:58 < kerframil> Gnjurac: yes, it's called pathname expansion. it's exactly that property that means that a command like `ls | grep *.deb` makes no sense. 15:59 < kerframil> Gnjurac: firstly, *.deb is a bogus regex. secondly, if you have any files ending in .deb in your curent working directory, the shell would expand them all to a list of words. those words would then all be passed as parameters to grep. 16:00 < Gnjurac> ty for explanation 16:00 < kerframil> in short, if you specify a glob, and the shell is able to match that glob to one or more files/paths, then that expansion is performed before your command is launched 16:00 < kerframil> be it mousepad, grep or anything else 16:00 < kerframil> if the glob doesn't match anything then it stays literally as it is 16:01 < kerframil> anyways, globs are useful. learn then, and use them well (and leave grep for the things it is good at). 16:07 < Sabotender> MrElendig: I am trying the shred command 16:07 < Sabotender> because everything else that ive tried just wouldnt work 16:15 < kerframil> Sabotender: if it's an SSD, use the ATA Secure Erase method 16:16 < Sabotender> dont know how, and I am tired of crap simply not working for reasons that I can't understand and that people are unable to explain and that I am unable to research on the net 16:16 < kerframil> Sabotender: https://www.thomas-krenn.com/en/wiki/SSD_Secure_Erase 16:17 < MrElendig> kerframil: problem is that the hardware makers sucks and lies 16:17 < MrElendig> kerframil: why implement the standard secure erase when you can make it a prorpietary feature 16:17 < MrElendig> that only works using a magic driver which only runs on windows 16:18 < Sabotender> right and the offending nvme drive IS my windows drive so I am stuck using an Ubuntu Live USB session 16:18 < kerframil> MrElendig: not sure that really applies in this case. standard operating procedure for mass-market SSDs is to employ encryption (transparently) at all times. so a secure erase just spins a new encryption key, rendering the existing data useless. hence, it is also fast. 16:18 < MrElendig> kerframil: he tried, it failed 16:19 < MrElendig> kerframil: btw, see what I wrote about eg WD failing 16:19 < kerframil> MrElendig: I missed that. is it an NVMe drive, by any chance? 16:19 < MrElendig> yes 16:19 < Sabotender> I tried hdpram (said something about invalid ioctl device) tried the one that MrElendig recommended and that didnt really work either :-( 16:19 < kerframil> ah, yes 16:19 < MrElendig> kerframil: WD had a nice feature where it had a key hardcoded in the hardware 16:20 < MrElendig> kerframil: which would always decrypt the data 16:20 < Dan39> ugh 16:21 < kerframil> Sabotender: in that case, try the blkdiscard command. it will simply instruct the drive to discard (trim) all of its sectors. it may take some minutes to completely take effect, but with deterministic-zero-at-trim, the result should be that the drive is effectively zeroed and bereft of any garbage that needs subsequent collection. 16:21 < MrElendig> there was also the bruteforce attack 16:21 < Sabotender> ill research the command 16:21 < MrElendig> fstrim will tell it do do that 16:22 < kerframil> sure, but why build a filesystem to do that when you can just do it at the block level (if a full wipe is the goal)? 16:22 < kerframil> also, I concede that malicious firmware is a potential threat 16:22 < MrElendig> +too 16:24 < MrElendig> wasn't even intentionally malicious 16:24 < kerframil> that's very bad 16:24 < Sabotender> blkdiscard /dev/nvme0n1 : BLKDISCARD ioctl failed : Operation not supported 16:25 < kerframil> hmm, that's a surprise 16:25 < MrElendig> not with adata :p 16:25 < kerframil> ah 16:25 < Sabotender> I used "sudo blkdiscard -s -v /dev/nvme0n1" 16:25 < kerframil> oh 16:25 < kerframil> drop -s 16:25 < kerframil> a regular discard should do the job 16:25 < kerframil> if it's accepted ... 16:26 < Sabotender> blkdiscard: /dev/nvme01: BLKDISCARD ioctl failed: Input/Output error 16:26 < kerframil> just checking but are you sure you're using the right device node? 16:26 < pennTeller> Guys, when I compile a Linux Kernel into a bzImage, is a bzImage something I can put into a USB and boot from it? 16:26 < kerframil> you seem to be alternating a little there 16:27 < MrElendig> you forgot a n there 16:27 < Sabotender> its a typo 16:27 < Sabotender> I have to type out all this stuff manually as I am working from a different computer 16:27 < MrElendig> is this a disk you are throwing away due to it being broken? 16:27 < MrElendig> no network? 16:27 < Sabotender> no, its a disc I am trying to erase so I can return and get a warranty replacement 16:28 < MrElendig> Sabotender: so what I said 16:28 < Sabotender> MrElendig: its linux, why does that surprise you? it hardly ever works out of the box with wireless 16:28 < MrElendig> eh, wireless generally works 16:28 < MrElendig> some distroes who doesn't ship certain firmware due to politics can be a bit of a trouble though 16:28 < Sabotender> generally my behind. I have never had wireless 'just work' on any of my various computers, but that is neither here nor there 16:29 < jim> pennTeller, I believe so... when I was compiling kernels at first, I booted from floppy, and all I had to do was dd the bzimage to the floppy and then set its root fs 16:29 < Psi-Jack> triceratux: Heh, Now that I have my MBUSBs in hand again... I managed to get the SwagArch stuff working totally and completely off just the ISO loopback method with actually less trouble.. 16:30 < pennTeller> jim thanks, how does one set the root fs now a days? 16:30 < Sabotender> MrElendig: https://paste.ubuntu.com/p/GjWBn8w5hs/ 16:30 < Sabotender> I dont think I can erase this disc even if I Wanted to 16:30 < Sabotender> take a look at that data 16:30 < Sabotender> I was hoping I would still be able to erase it 16:30 < MrElendig> with IO errors trim is unlikely to work 16:30 < Dan39> just look at it 16:31 < Psi-Jack> Buuut, then the installation just failed. Le-sigh.. 16:31 < kerframil> pennTeller: usually not directly. in principle, you could if (a) your kernel was built with CONFIG_EFI_STUB=y (b) you are booting with UEFI (c) your USB drive is formatted as FAT32 and the kernel image is situated appopriately beneath the EFI/ directory (d) your UEFI firmware allows you to specify and pass the relevant options and/or you registered an entry with efibootmgr. 16:31 < jim> nowadays? you could install grub to the usb, and then put a root=/dev/somenode on its kernel command line 16:31 < Psi-Jack> Oh, duh. forgot to connect it to the network. 16:31 * MrElendig would use a initramfs too 16:31 < jim> but, I don't remember the command I used to set the root in the kernel 16:32 < kerframil> pennTeller: the stub means that the kernel is a valid UEFI executable and that you can thus forgo a traditional bootloader. in practice, it wouldn't necessarily be so easy and you're probably best off using a traditional bootloader. 16:32 < Sabotender> MrElendig: its not just that, look at 'available spare' 16:32 < pennTeller> kerframil, thank you for the info 16:32 < pennTeller> jim thanks Jim, I appreciate the help 16:32 < jim> right... if you don't use (something like) an initrd, then you have to make sure everything needed to mount / is compiled into the kernel (i.e., not as a module) 16:33 < MrElendig> I wouldn't really worry about the data on it unless a pointy haired boss told me to do so 16:33 < triceratux> Psi-Jack: welcome to swagarch. what does your loopback grub2 stanza look like ? id like to stop extracting that squashfs every month 16:33 < MrElendig> in which case I would physically destroy the thing and have him pay for the replacement 16:33 < Sabotender> MrElendig: it has some of my private data on it 16:33 < Sabotender> I thought it did not but it does 16:33 < Psi-Jack> triceratux: I'll paste it in a few. In the middle of trying to install off it. 16:33 < MrElendig> unlikely that they will go digging after it 16:34 < Sabotender> I think this drive is stuck in 'read only' mode 16:34 < MrElendig> specially since it might involve having to unsolder the chips and possibly decap them 16:34 < Sabotender> thats the thing. who knows what Adata is going to do with it 16:34 < Sabotender> no no, the drive is frozen in read only mode 16:34 < Sabotender> you can read from the drive just fine 16:34 < MrElendig> if you really care, consider the 200usd lost and just get a new one 16:34 < Sabotender> you just cant do anything else to it 16:34 * Sabotender sighs 16:34 < Sabotender> I am going to call and complain to newegg 16:34 < kerframil> Sabotender: that's definitely possible. that would be bad as it would rule out shred also (did you try yet?) 16:34 < Sabotender> maybe they will send me a replacement 16:35 < Sabotender> kerframil: I tried shred and I got a line full of io errors 16:35 < kerframil> Sabotender: ah. not good. 16:35 < Sabotender> and then the drive got stuck 16:35 * iflema jim thanks Jim 16:35 < Sabotender> like..it disappeared from linux (wasnt able to see it) and I couldn't shut the computer down normally because the HDD busy light was stuck on 16:36 < jim> welcome... what did I do? :) 16:36 < kerframil> Sabotender: what MrElendig said. if you want the confidence of knowing that the data on it is irrecoverable, you appear to have little choice but to write it off. 16:36 < pLr> finally there are more users in #archlinux than #ubuntu YOTLD IS UPON US 16:36 < Sabotender> when I get a new drive (not adata) I am going to install windows 10 in a secure encrypted fashion. I really hate doing that because it forces me to remember long, complicated passwords, and I am not even sure if windows can do that 16:37 < kerframil> it can (either with bitlocker or veracrypt) 16:37 < Sabotender> I know that linux can, which I have it set up that way, but that's linux 16:37 < MrElendig> pLr: have happened several times in the past 16:37 < MrElendig> pLr: #archlinux used to be the largest channel on the network for a while 16:37 < pLr> MrElendig: i was away for a long while i guess things have changed 16:37 < MrElendig> but for various reasons about 800 nicks dropped off the network 16:38 < Psi-Jack> triceratux: Curious though.. Do you use UEFI? And did SwagArch work with installing EFI stuff? I'm checking out my /boot now, not even an efi or EFI directory at all. 16:39 < MrElendig> Psi-Jack: findmnt /mnt 16:40 < triceratux> Psi-Jack: nope its just a 2G bios optiplex. havent needed the uefi stuff yet. hopefully itll be issue free when i decide to bite it off 16:40 < MrElendig> er.. /boot 16:40 < Psi-Jack> MrElendig: Heh, I'm looking at it inside the BIOS with it's built-in filesystem browser. heh 16:41 < Psi-Jack> Hmm 17:09 < Psi-Jack> triceratux: https://paste.linux-help.org/view/f6aef168 BTW 17:09 < Psi-Jack> Err. Correction 17:10 < Psi-Jack> triceratux: https://paste.linux-help.org/view/b2bd67d2 -- fixed, missed the x86_64 directory. 17:14 < triceratux> Psi-Jack: really appreciate it. ill try it out shortly 17:15 < Psi-Jack> Hehe triceratux That actually detects the media of which booted, plus finds the volume names of the media and ISO, and keeps everything localized. :) 17:16 < triceratux> Psi-Jack: i kinda deduced that from reading it. never seen anything like that 17:16 < Psi-Jack> hehe 17:16 < Psi-Jack> That's the way I roll. :) 17:22 < Psi-Jack> triceratux: You should see the whole setup. Pretty extensive. :) 17:23 < system16> hi. i need to install "git" in order to use this command git clone https://github.com/Juve45/batstat.git but its 26 MBs and this is concerning for me. is it safe ? 17:23 < system16> i never had to install such a big app 17:24 < Psi-Jack> wut? 17:24 < rumpel> w00t??1?1 17:24 < balletjebal> ^^ 17:24 < Reventlov> system16: most probably ok 17:25 < Reventlov> but, here, the binary is only 55Ko 17:25 < rumpel> system16, you don't need to install git. you can also download it as .zip 17:25 < Reventlov> KB* 17:25 < system16> most apps that i installed were around 150 KBs. thats y im asking 17:25 < system16> i want this app : batstat 17:25 < Reventlov> and the full git history is 832KB 17:25 < rumpel> system16, look for the friendly green "Clone or download" button 17:25 < Reventlov> yeah, what I say, it's only 55KB compiled. 17:25 < system16> there is none 17:25 < system16> since i have no DE 17:26 < Psi-Jack> system16: "y" does not supplicate the replacement for the real word "why". 17:26 < Reventlov> just do a "git clone https://github.com/Juve45/batstat.git", a "g++ main.cpp -lncurses -pthread -std=c++11 -o batstat" and you're done. 17:26 < system16> my server isnt powerful enough for a DE 17:26 < Psi-Jack> server? 17:26 < system16> sftp server 17:26 * Psi-Jack scratches his head. 17:27 < system16> file server. 17:27 < akd> Does anybody has a live stream link in asia for watching Russia Saoudia Arabia? 17:27 < SuperSeriousCat> >sftp server >want DE 17:27 < Psi-Jack> akd: Google.com. Got a Linux question? 17:28 < system16> yeah clone and that stuff is complicated. i dont wanna break stuff 17:28 < fron> do we still use automounts or its a old technology? 17:29 < Psi-Jack> fron: What are you really trying to do/ask? 17:29 < fron> I am just reading around docs for RHCSA cert 17:29 < fron> so came across automounts 17:29 < system16> well that was quick 17:29 < system16> thanks guys 17:30 < Psi-Jack> fron: Heh. automounting is doable in systemd these days with an automount and mount service. 17:31 < fron> oh ok.. any other alternative for automount? 17:31 < system16> bash: /usr/local/bin/batstat: cannot execute binary file: Exec format error 17:31 < system16> why ???? 17:31 < fron> automount is a system that makes the connection to the server thats is offering access to the home directory.. is my understanding correct? Psi-Jack 17:31 < Psi-Jack> fron: Incorrect. 17:31 < system16> is it because of having no GUI / DE ? 17:32 < fron> Psi-Jack: could you brief me then about what its used for 17:32 < Psi-Jack> fron: Automatically... mounting.... On access.. 17:32 < Psi-Jack> Just what it sounds like. :p 17:33 < system16> Psi-Jack, do i need a DE for batstat ? 17:35 < system16> batstat: ELF 64-bit LSB executable < when i type file batstat i get this 17:36 < system16> batstat: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, BuildID[sha1]=a198958fbf4d75807187b963a2341706c6fe2a01, not stripped 17:37 < system16> article : https://www.ostechnix.com/how-to-check-laptop-battery-status-in-terminal-in-linux/ 17:37 < system16> scroll to the bottom of the page 17:37 < Psi-Jack> 32-bit? 1995 is calling. 17:38 < system16> so i installed a 64 bit app ? 17:38 < Psi-Jack> No. you COMPILED it as 64-bit. :p 17:38 < system16> oh 17:38 < system16> how can i fix that 17:39 < system16> chmod +x /usr/local/bin/batstat 17:39 < system16> Psi-Jack, do i need a gui for that program ? 17:42 < Psi-Jack> ... 17:43 < Psi-Jack> For starters. I haven't even LOOKED at the github URL you pointed out earlier. And don't plan to. 17:43 < system16> apt remove batstat will remove it right ? 17:44 < Psi-Jack> and now I'm pretty sure you're just trolling. Good luck. 17:44 < system16> no 17:44 < system16> why do you think that im trolling ? 17:44 < system16> i use devuan. devuan cant understand apt-get 17:46 < system16> yeah im not a pro in linux but im not a troll guy 17:46 < azarus> Hm, I'm trying to compile vice, the commodore emulator. Upon runnning ./configure I get: http://ix.io/1dlF 17:46 < azarus> Any clue to what `xa` is? 17:46 < toothe> sigh...why does my company use Kali? 17:46 < toothe> :( 17:46 < Armand> O_o 17:46 < MikeFromIT> wat 17:46 < Armand> Why would ANY company use Kali ? 17:47 < Armand> That's just.... wrong 17:47 < toothe> Armand: its recognized by the DoD 17:47 < Psi-Jack> Ignorance. Stupidity. Etc. 17:47 < toothe> yeah, i realyl dislike it. 17:47 < Armand> HAHAHAHAAA 17:47 < system16> Armand, maybe a pen testing company ? 17:47 < Psi-Jack> toothe: Pardon? 17:47 < toothe> system16: we are that. 17:47 < ayecee> well good 17:47 < toothe> Psi-Jack: which part? 17:47 < toothe> Its just so annoying to maintain. 17:47 < Armand> Ahh.... "Tech Support" in India ? 17:47 * Armand ducks 17:47 < Psi-Jack> toothe: DoD? 17:47 < toothe> Kali should be run in a VM, not on the hardware. 17:47 < toothe> Psi-Jack: department of defense. 17:48 < Psi-Jack> toothe: I know. I was military. DoD actually approved Kali? 17:48 < toothe> yes. 17:48 < Armand> toothe: Incorrect... they wouldn't approve Kali. 17:48 < toothe> its recognized government-wide 17:48 < Psi-Jack> Since.. When? 17:48 < Armand> EVER 17:48 < toothe> Armand: I dunno what to tell you, it is. 17:48 < Armand> Dude... My aunt works at the Pentagon. 17:48 < Armand> Seriously. 17:48 < ayecee> so? my dad is the president of xbox. 17:48 * Psi-Jack quickly kidnaps Armand 17:49 < rumpel> my uncle owns mars 17:49 < Armand> lol 17:49 * system16 ignores Psi-Jack 17:49 < SuperSeriousCat> Your uncle wash my car 17:49 < Psi-Jack> Oh big loss there. LOL 17:49 < Armand> You'd think that OffSec would brag about that on their website.. 17:49 < ayecee> Armand: what are you saying? "my aunt works at the pentagon, therefore i am an authority on government software procurement" ? 17:50 < Armand> ayecee.... I'ma just blank your rhetorical noise for now. M'kay. 17:50 < Psi-Jack> I don't actually see any DoD approvals for Kali. 17:51 < Armand> Psi-Jack: Absolutely not. 17:51 < ayecee> what do you see approvals for? 17:51 < Armand> Especially considering how many non-US citizens are involved. 17:52 < Psi-Jack> ayecee: Well, Fedora Linux, for one. :) 17:52 < ayecee> nice 17:53 < Armand> OffSec aren't even listed under the DodIN list. 17:54 < ayecee> could be because it's such a rapidly evolving field 17:54 < Armand> Let's check the PCL... 17:54 < Sabotender> looks like I am SOL when it comes to this drive 17:54 < Sabotender> I am out 54USD 17:54 < Armand> NOPE 17:54 < Psi-Jack> Forge (custom government distro), Fedora, openSUSE, Debian, Ubuntu, Solis and Linux Mint, I see. 17:55 < ayecee> solus? 17:55 < Sabotender> because I am *not* sending this drive to ADATA in its current state. There's passwords on here, photos that I dont want people seeing, tax documents, all unable to be deleted becuase the drive is stuck in 'read only' mode 17:55 < Psi-Jack> heh yes. but they listed it as Solis. LOL 17:56 < ayecee> Sabotender: could try erasing it via the drive's internal secure erase thingy 17:56 < ayecee> https://ata.wiki.kernel.org/index.php/ATA_Secure_Erase 17:56 < Sabotender> ayecee: is that windows only? 17:56 < ayecee> no 17:57 < ayecee> what a weird question 17:57 < Sabotender> well a lot of proprietary software that ive seen is windows only 17:57 < ayecee> it's not software 17:58 < Sabotender> okay ill look at it 18:01 < Sabotender> ayecee: ah, hdparm doesn't work 18:02 < ALowther_> What tool would you recommend using to modify a directory of files, searching for a specific phrase and inserting a phrase after?...Specific to my circumstance. I have hundreds, maybe thousands of .htm & .html files. I want to modify them all by adding something underneath the current banner. Luckily, it seems they all use a consistent tag to load the banner. I want to search every htm/html file for that tag, when it finds it, find the 18:02 < ALowther_> closing '>' for that tag and then insert an other tag to insert a different image just beneath it. 18:02 < Sabotender> I got an error something along the lines of 'ioctrl incompatible blah blah' 18:02 < ALowther_> another* 18:02 < prussian> lol, what SSD is this Sabotender 18:03 < ayecee> an adata 128gb ssd. hdparm should work with it. 18:03 < Sabotender> perhaps I was trying to use an old version of hdparm? 18:03 < ayecee> oh, a m.2 ssd 18:03 < ayecee> i guess that might not 18:03 < prussian> it's probably not Serial ATA then 18:04 < ayecee> ¯\_(ツ)_/¯ 18:04 < Sabotender> I dont think it is, because linux refers to it as 'nvme0' 18:05 < Sabotender> this makes me mad because I have to spend MORE money 18:05 < Sabotender> I am going to install windows in a secure manner from now on 18:06 < ayecee> #firstworldproblems 18:06 < Psi-Jack> Or just rid yourself of Windows forever. 18:07 < Sabotender> no, because most of the games I play are windows only 18:09 < ayecee> oh neat, there's a "nvme-format" command that might work for this. 18:10 < ayecee> as described here: https://www.systutorials.com/docs/linux/man/1-nvme-format/ 18:16 < BluesKaj> Sabotender: buty an xbox :-) 18:16 < BluesKaj> buy even 18:17 < ayecee> o_O 18:17 < ayecee> pc gamer master race, come on man 18:17 < Sabotender> I already own an xbox, but its graphical capability can't even touch the power of my 1080Ti that I have in my pc 18:19 < BluesKaj> bet you play on a 65" 4K TV too :-) 18:19 < Sabotender> pretty much 18:20 < BluesKaj> gamers.... 18:38 < squarebracket> i'm trying to do an out-of-tree build of i915 with intel's patches for doing hardware transcoding, but the driver's including a file from drivers/platform. is there a way to reference that file from outside the tree, or am i stuck with an in-tree build? 18:39 < fron> hi there, i get this error mount.nfs: access denied by server while mounting 35.154.87.157:/home/centos 18:40 < Raed> fron: Looks like you dont have permission to mount the share. 18:41 < fron> how to have the permission? 18:41 < Raed> fron: You need to change the permissions on the server side. check server logs to see why it's denying it. 18:41 < fron> Raed: server logs? /var/log/messages? 18:42 < Raed> fron: It depends on where it defaults to in whichever distribution you are using. 18:42 < BluesKaj> squarebracket: what makes you think that building the driver from source is going to give a better driver? ? 18:43 < fron> Raed: i see this Jun 14 16:38:15 ip-172-31-22-161 rpc.mountd[10118]: refused mount request from 103.215.170.1 for /home/centos (/home/centos): illegal port 64404 18:43 < BluesKaj> especially the i915 which is the most comon inbtel audio driver out there 18:43 < Raed> fron: Ok, so it doesnt like the port for whatever reason. 18:44 < fron> how do i change the port? 18:44 < revel> Illegal, you asy? 18:45 < Raed> fron: Quick google shows that error is because you dont have the insecure options on the share in /etc/exports, so if the port is above 1024 it will deny the mount. 18:45 < revel> I think I'll have to go report fron to Interpol. 18:45 < Raed> revel: Lol 18:45 < revel> s/asy/say/ 18:45 < squarebracket> BluesKaj: intel makes what they call "intel media server" which requires changes to the i915 driver for better/faster video gpu transcoding 18:46 < squarebracket> BluesKaj: i can't say i know the specifics of it, i'm just tasked with building the driver. which i know isn't a great excuse. maybe it's just that intel's changes haven't been upstreamed yet? i'm honestly not sure. 18:47 < BluesKaj> oops my mistake I meant video driver squarebracket 18:47 < fron> Raed: one more question 18:47 < Raed> fron: Ok.. 18:47 < fron> how to know what are all the options nfs has 18:48 < BluesKaj> squarebracket: , I wasn't aware the i915 had better versions available 18:48 < fron> like insecure,ro,rw,sync.,et 18:48 < fron> etc* 18:48 < Raed> fron: The man pages / documentation 18:49 < fron> also instead of /etc/exports can we use /etc/fstab? is there any difference if we use the latter for nfs 18:50 * BluesKaj searches intel media server 18:52 < Raed> fron: etc/exports defines the actual partitions or directories you want to mount, fstab is for the system, you can use fstab on the system, to auto mount the nfs share, but you still need /etc/exports to be set up. 18:55 < fron> Raed: so if a restart happens, unless we define in fstab, nfs wont sync up? 18:56 < Raed> fron: What do you mean? You need an exports file to tell the nfs server which directories to serve out. fstab is for telling the system which file systems to mount. 18:56 < Raed> fron: So, you can use fstab, to tell the client system to automatically try to mount the nfs share when it boots up 18:57 < Raed> fron: Otherwise you will have to do it everytime you want to mount it by hand. 18:58 < squarebracket> BluesKaj: i'm not really sure if it's "better" per se.... just that the patched module is needed to make use of features from the intel media server sdk, which we use. which maybe means it's not a good candidate for inclusion in the upstream driver /shrug 19:01 < BluesKaj> squarebracket, I suppose it depends on the intel hardware ...thjere's also the i965 video driver on newer onboard gpus 19:03 < squarebracket> anyway, it would be nicer if i could build out-of-tree, but the i915 drivers includes `drivers/platform/x86/intel_ips.h` and i don't know if that's something i can include from outside the tree, or where the actual code in `intel_ips.c` ends up 19:05 < squarebracket> i imagine it's linked against whatever `intel_ips.c` compiles to, and the header is just a single function. :\ 19:06 < revel> BluesKaj: Unless I'm mistaken, the kernel just has CONFIG_DRM_I915, there's no seperate i965 or above driver. 19:07 < Reventlov> Hi 19:07 < Reventlov> do you know where I could get statistics on hardware used on linux desktop? 19:08 < Reventlov> For example, which percentage of ubuntu users run an intel processor, or an specific intel wifi card 19:08 < BluesKaj> squarebracket: ok, then the HW is the regular intel video chip 19:09 < lnnb> Reventlov: what would you do with this information? 19:09 < revel> Reventlov: If you can find similar statistics for Windows users, then it shouldn't be any different. 19:10 < Reventlov> lnnb: well, science. 19:10 < Reventlov> revel: well I don't know anything about windows. 19:11 < Reventlov> Also, the linux ecosystem is way more scattered than windows, so, yeah, it should be different. 19:11 < Reventlov> So, do you know of any hardware statistics initiative on Linux? 19:11 < koala_man> Valve does one for Steam users 19:12 < Reventlov> koala_man: neat, i'll check that out. 19:12 < lnnb> "science" is practically dodging the question 19:12 < Reventlov> lnnb: well, yes, because this is not your business, mainly. 19:12 < lnnb> right, so i won't help, thx 19:13 < Li> I'm using "ps aux" trying to follow how an automatically running program "V0399" was configured to run on rPi ubuntu mate. I found one line "112 root ******** SCREEN -dmS Ultra /home/pi/V0399" .. Checked crontab nothing different from another normal default distro!! can anyone suggest where to look for how this programming starts automatically? 19:13 < Reventlov> koala_man: seems there is nothing about network chipset, but, anyway, thanks for the hint :) 19:14 < koala_man> nope, but it does show intel cpus 19:14 < Reventlov> yep :) 19:15 < lnnb> sounds more like market research, not actual science 19:15 < Reventlov> lnnb: aaaand you're wrong. But, anyway, you said you were not helping 19:15 < dgurney> I'm not sure there's any such statistics easily available even for Windows machines 19:15 < lnnb> i'm not helping, i'm trying to convince myself you are doing actual science but i'm failing 19:16 < Sitri> Li: grep -R the etc directory 19:18 < YADW> Hello, I don't really know if this is the right place to ask, but I'm having a hard time making a program from source. Can I ask here for help? 19:19 < koala_man> yes, though if the program has its own channel, they'll probably be more helpful 19:22 < YADW> Thanks. I'm almost sure it doesn't, since it's a library to render swf files to png... Anyway, whenever I "make install" it gives me an error 1 and error 2 19:24 < Reventlov> YADW: paste the whole output on the pastebin linked in the topic 19:24 < Reventlov> from the beginning, when you write make, to the end, after you write make install. 19:25 < Dagmar> ...or just use termbin 19:26 < YADW> Well, I'll have to make you wait a couple of minutes, since the output is in italian (I heard that's a collateral effect of being an Italian guy, in Italy, on an Italian translated system). I'll translate it in a moment. 19:27 < alexandre9099> hi, which are the modules that take care of usb? can i reload them all? 19:27 < Dagmar> YAOW: There's always prefixing your commands with LANG=C 19:28 < YADW> Dagmar: didn't know about it, it's quite clever actually. 19:28 < alexandre9099> Dagmar, btw, what "C" stands for? 19:28 < Dagmar> alexandre9099: It's just the default 19:29 < alexandre9099> yea, but i mean, why "C" :D maybe "C"ommon? 19:29 < jim> I think it's the C locale (and dunno what it stands for, maybe C prog lang?) 19:29 < Dagmar> Basically, it tells any gettext using application to emit the untranslated versions of the strings 19:29 < Dagmar> i.e., what's in the C code 19:29 < tbsn> good afternoon all 19:29 < jim> hi 19:30 < alexandre9099> hmm, maybe that :) 19:31 < Psi-Jack> triceratux: Hmmmm... What's the "fix" for the swagarch pacman-key issue, do you know off-hand? 19:31 < tbsn> I'm setting up a simple btrfs backup/media server. can I put the 2.5" drive in my laptop, install on that, and simply switch it over to the server for ssh access, or should the install happen on the computer itself? 19:32 < well_laid_lawn> you can swap the hdd over ok 19:32 < koala_man> tbsn: yes, you can typically do that 19:32 < Dagmar> You can do whatever you want. however, you've not told anyone a *reason* why you're wanting to do this 19:33 < YADW> This is make --> https://pastebin.com/bdLKPysH 19:33 < Psi-Jack> YADW: pastebin.com is frowned upon due to many issues they themselves have caused. Pastes being reformatted, malvertising, adblock blocking, being blocked due to many reasons. See /topic for the channel's official pastebin. 19:33 < tbsn> just rather not have to put a monitor and keyboard on the computer im using as a server 19:34 < jim> tbsn, which dist? 19:34 < koala_man> YADW: seems fine 19:35 < Psi-Jack> tbsn: Even a server could use a KVM switch to access the video, keyboard and optionally mouse as-needed. 19:35 < tbsn> well, haven't chosen yet. probably debian. 19:35 < triceratux> Psi-Jack: it taught me to say SigLevel = Never in the /etc/pacman.conf. havent had to say it for awhile. if the first neofetch hangs when the first terminal opens his key is hung. but its been ok for a few days 19:35 < Psi-Jack> Oh... That's just the wrong way. :) 19:35 < razwelles> Hey, I'm trying to dig around and figure out what the most common software based suspend is for linux, and how it's architected 19:35 < jim> Psi-Jack, that, plus lots of javascript, can crash a low-ram machine 19:36 < Psi-Jack> jim: What? 19:36 < YADW> This is make install --> https://paste.linux.community/view/80da858e 19:37 < jim> Psi-Jack, another reason to not prefer pastebin.com 19:37 < Dagmar> YAOW: Looks like the problem is that it's version 0.9.2 19:38 < Psi-Jack> Ahhhh 19:38 < Psi-Jack> True though 19:38 < Dagmar> There's nothing going terribly wrong during the build, but clearly the author has screwed up the Makefile's install stanza 19:38 < YADW> Dagmar I suspected it was something in the makefile, or on the dev side in general 19:38 < electrosys> Hi 19:38 < YADW> Funny, it's supposed to be the stable branch lol 19:39 < jim> hi 19:39 < electrosys> i'm all excited becuase i got offlineimap/mutt/gmail working with oauth2 and i got the labels working so i can delete e-mails to the trash folder. 19:40 < electrosys> and im useing weechat for the first time which is pretty cool. 19:40 < jim> YADW, it looks like the make added -o to an rm invocation, and according to the man page there's no -o (there is --one-file-system however)_ 19:41 < YADW> jim yes, I noticed that, too. I don't know where to fix that, though. Is it in the makefile? 19:41 < triceratux> Psi-Jack: there was a bug report noting something similar before the project moved from github this week. im hesitant to talk up that project useful as it is. youll need to keep a close eye on it 19:42 < buoyantair__> open source alternative to gmail? 19:43 < revel> Dovecot. 19:43 < hexnewbie> buoyantair__: That's an oxymoron, gmail's a cloud service. Meaning, it's not only closed source, it operates on somebody else's computer. 19:43 < Artemis3> Maybe he wants software to set up your own cloud mail? 19:44 < hexnewbie> buoyantair__: You wouldn't be in any better position even if they ran Dovecot/Roundcube/Zimbra 19:44 < revel> What's zimbra? Spam detection? 19:44 < buoyantair__> hexnewbie: I was talking about an alternative to gmail o.o 19:44 < buoyantair__> something like mailtrain (which is like an alternative to mailchimp) 19:45 < revel> A service rather than software then? Since I don't personally see how an "open-source service" makes sense. 19:45 < buoyantair__> protonmail works but it doesnt seem to accept all emails 19:45 < buoyantair__> (in free tier) 19:45 < revel> cock.li 19:45 < buoyantair__> revel: No I mean software 19:46 < revel> Last I checked, Mailchimp was a service. 19:47 < hexnewbie> buoyantair__: Then Postfix with Dovecot or Courier and a proper email client ought to work. Roundcube is popular web mail client that doesn't suck too much. My university uses Zimbra, which has a sexy web frontend and IMAP/POP3/SMTP, but I don't know if it uses theirs or third-party 19:48 < hexnewbie> The FSF and OSI have approved Zimbra as free and open source (respectively), but Debian hasn't. 19:48 < buoyantair__> Oh, where can I know if they approved or not? :O 19:49 < hexnewbie> buoyantair__: https://en.wikipedia.org/wiki/Common_Public_Attribution_License has links 19:50 < hexnewbie> The OSI link is lacking, though. 19:51 < Psi-Jack> triceratux: Yeah, I just had to delete the /etc/pacman.d/gnupg/* contents, and re-init and populate 19:51 < jim> hexnewbie, why does debian not support zimbra? 19:52 < Psi-Jack> It's not that Debian doesn't support Zimbra. It's Zimbra that doesn't support Debian. 19:52 < Psi-Jack> But, actually, Zimbra does in fact support Debian. 19:54 < Psi-Jack> Though Zimbra is a beast. 19:54 < hexnewbie> jim: The frontend uses the CPAL license, which according to Debian is a badgeware license. Requiring attribution that Debian couldn't comply with. 19:57 < hexnewbie> I.e. they require a logo to be displayed when the program is started/web interface displayed, i.e. a splash screen or button on the web page. 20:01 < jim> hexnewbie, ok, thanks 20:04 < Ajven> Hello, Im trying to make my 1080Ti PCI passthrue by qemu but taking error - https://pastebin.com/ceF9dMeE maybe you could help with that ? Its Arch linux host 20:05 < buoyantair__> Why are other linux distros not listed on gnu's webpage? :O 20:05 < hiya> Anyone here using shadowsocks client VPN style? 20:05 < hiya> I am trying to forward all my UDP+TCP traffic to UDP socks5 proxy that I have setup using shadowsocks client 20:05 < hiya> https://hastebin.com/tirukotiza.bash 20:06 < hiya> This is how I am trying it ^ but it just doesn't work 20:06 < hiya> It shows my home IP instead 20:15 < electrosteve> hi 20:19 < yskapell> hi all 20:20 < yskapell> I have kali linux on an Acer E15 laptop 20:20 < yskapell> and I want to upgrade BIOS 20:20 < yskapell> How can I do it without installting windows? 20:20 < ayecee> consult your user manual or the manufacturer's website for other options 20:21 < ayecee> it's not normally practical to upgrade the bios from within linux. 20:21 < yskapell> any suggestions? 20:22 < ayecee> yes. consult your user manual or the manufacturer's website for other options 20:22 < ayecee> two suggestions, both for you. 20:35 < seven-eleven> hi 20:36 < seven-eleven> what software should I use to create a windows VM in linux, virtualbox or kvm? the windows vm has to be access from the WAN. the network is behind a WAN router, so I think i'd have to give my VM a virtual LAN IP like 192.168.2.10 20:37 < koala_man> if you have to ask, virtualbox 20:37 < seven-eleven> alright, going for virtualbox then, thanks! 20:41 < electrosteve> @yskapell generally that needs to be done with a program provided by the manufacturer, do they provide a linux script? 20:44 < well_laid_lawn> I don't like nics that are brand names 20:55 < NoirX> hola 21:04 < Nawab> can anyone tell me how to output the output of tar to a directory outside the directory the archive is in? 21:05 < hexnewbie> Nawab: You want to specify where to extract the archive? 21:07 < phogg> Nawab: the dir the tarball is in does not matter, the dir you are in matters (your working dir). Use -C to specify an alternate 21:07 < Nawab> ok 21:24 < NCS_One> hi 21:25 < dunpeal> Not sure if it's a bash question, but I'm running less with -Q (highest level of silence, bell should never be rung), yet the bell is rung every time I try to scroll past the EOF. 21:25 < NCS_One> how can I remove a folder that has some error, with "ls -al" I get "d????????? ? ? ? ? ? folder" 21:26 < DLange> NCS_One: start midnight commander (mc), press F8 on the folder 21:27 < hexnewbie> NCS_One: What error does rmdir give you? Or attempts to list it directly? 21:27 < ayecee> NCS_One: you can't. 21:27 < NCS_One> hexnewbie: says its is a read only file 21:28 < ayecee> filesystem, probably 21:28 < hexnewbie> NCS_One: That's not an actual error message 21:28 < chey> heyy 21:28 < NCS_One> hexnewbie: "rm -R folder/" gives "rm: cannot remove 'folder': Input/output error" 21:28 < ananke> NCS_One: this is likely a result of a damaged filesystem 21:29 < davidej> Does anyone know if Kent Overstreet is in this chan - random chance? 21:29 < hexnewbie> NCS_One: If you get question marks for directory entry of a directory your user has +x permissions to, *and* I/O error, *and* read-only filesystem, either filesystem corruption or dying hardware. Run dmesg to tell 21:29 < ananke> NCS_One: the error you just pasted only further confirms that 21:29 < davidej> I'm going to wind up emailing him a question but I was curious if he was on IRC 21:31 < Pentode> davidej, although some developers do, the "unofficial" linux channel, otherwise known as ##linux isn't the best place to find those kinds of developers. :( 21:31 < ayecee> most people prefer not to be readily identified by their real name on irc too 21:32 < Pentode> yeah that too 21:32 < nchambers> oh 21:33 < Foxhoundz> Are there any NTP servers DIRECTLY connected to an actual atomic clock? 21:33 < ayecee> yes 21:33 < Foxhoundz> And if so, are they reachable 21:33 < ayecee> probably yes 21:33 < ayecee> wouldn't be much use if they weren't 21:33 < Foxhoundz> I thought they simply handed off time to secondary NTP servers 21:33 < ayecee> but you almost certainly shouldn't be using those as your ntp servers. 21:33 < Foxhoundz> which then served it to the world 21:33 < bls> it's unlikely they'd let random end users connect to them 21:34 * Pentode looks for his ribidium standard 21:35 < ayecee> ribit 21:35 < Pentode> -i+u 21:35 < Pentode> oops '[ 21:35 < Pentode> oops ;p even 21:36 < ayecee> :D 21:36 < Pentode> bah! 21:36 < Pentode> whats wrong with a regular ntp anyway? not accurate enough? 21:36 < ayecee> we may be assuming too much 21:36 < nchambers> people use NTP for accuracy? 21:36 < Foxhoundz> well 21:36 < Foxhoundz> I was just reading into high frequency trading 21:36 < Foxhoundz> and how they would sync time 21:37 < Foxhoundz> I'd imagine the tolerances are very, very tight 21:37 < bls> and you believe that a lower strata time source is less accurate than a higher one? 21:38 < ayecee> the implication being that that is not always the case 21:38 < prussian> Microchip cesium clocks soon 21:38 < electrosys> hi 21:38 < ayecee> i've always wanted a radioactive computer 21:39 < electrosys> I guess you have to have an registered nick to talk? doesn't relly matter which one though once you have at least one? 21:39 < Foxhoundz> maybe these trading systems have onboard atomic clocks themselves 21:39 < ayecee> electrosys: yup 21:39 < ayecee> maybe these trading system have their own timing signal that is not specifically related to the clock 21:40 < bls> or maybe time from a lower strata time source is accurate enough 21:40 < xcm> is accurate _real_ time of any use for trading? 21:40 < electrosys> i think the trick to high frequency trading is to make a trade on two different timestamps, and take advantage of the gap 21:40 < ayecee> such a broken system 21:40 < searedvandal> the regulatory nightmare of trading. utc via gps is probably what many do 21:41 < electrosys> its all a scham, when was the last time you said, hey, im going to buy that stock I like thr product. You generally make trades based on the chart. its all a game. 21:41 < ayecee> high frequency traders do that, not everyone does that 21:41 < SpeakerToMeat> can you guys think of any way to copy a single file from two machines to a third one over the net? 21:42 < ayecee> SpeakerToMeat: maybe, if you can explain how a single file is on two machines 21:42 < ayecee> quantum superposition? 21:42 < SpeakerToMeat> Other than split 21:42 < SpeakerToMeat> ayecee: hmmm no... simple cp... 21:42 < ayecee> could simple cp solve this new problem too? 21:42 < SpeakerToMeat> I'm trying to upload from two slowe internet locations at once 21:42 < ayecee> ok, there's the missing piece 21:42 < electrosys> SpeakerToMeat: can you just copy the files over with wget/rsync and then cat them together? 21:43 < SpeakerToMeat> Yeah split. sigh I guess split is the best way 21:43 < dotcomboom> the third machine is downloading one file from two mirrors basically? 21:43 < ayecee> yes 21:43 < ayecee> i mean, bittorrent could do this, but it probably takes more effort to set that up than it's worth. 21:43 < ayecee> this is what torrents are made for. 21:44 < dotcomboom> I think wget/uget might support it, somehow 21:44 < ayecee> i don't think so 21:44 < ayecee> but who knows, eh 21:45 < Pentode> curl can do that, can't it? 21:46 < ayecee> can it? 21:46 < Pentode> i think so 21:46 < ayecee> i don't think so 21:46 < searedvandal> aria2 21:46 < SpeakerToMeat> Well you can specify blocks with curl but I'm not sure about multiple sources at once 21:47 < dotcomboom> just found someone saying that on superuser 21:47 < dotcomboom> sample: "aria2c -s5 host1/bigfile host2/bigfile ... host5/bigfile" 21:47 < dotcomboom> or just "aria2c -s5 host1/bigfile host2/bigfile" because two mirrors 21:47 < SpeakerToMeat> Hmm interesting. 21:47 < searedvandal> yeah, aria2 should do what you want 21:48 < dotcomboom> https://aria2.github.io/ 21:48 < SpeakerToMeat> Now I only need to fix the ssh forward not working in th erouter of one of the two locations 21:49 < Pentode> ayecee, well you would have to run two instances i guess then stick them together manualy. 21:49 < SpeakerToMeat> thanks searedvandal 21:49 < mcj> https://weechat.org/about/screenshots/weechat/weechat_2013-04-27_phlux.png/ Anyone know what font that is? 21:49 < SpeakerToMeat> And doing a pull rather than push will be easier to monitor 21:50 < Pentode> +l what is it with me i think this virus is eating my brain 21:51 < dotcomboom> mcj I think that's... the xterm default? can't be sure 21:52 < dotcomboom> no, actually not quite. but they look really similar 21:52 < mcj> That's what I thought but it isn't 21:53 < dotcomboom> put a bounty on it :p 21:53 < searedvandal> It's not Terminus? 21:53 < searedvandal> to me it looks pretty close to Terminus, but I don't have my glasses on 21:53 < SpeakerToMeat> There's font identification sites.... 21:53 < Pentode> kinda close but it's not terminus 21:53 < mcj> Deff not http://terminus-font.sourceforge.net/shots.html 21:53 < mcj> SpeakerToMeat: LINK 21:54 < mcj> link me now~ 21:54 < Pentode> i use it extensively 21:54 < SpeakerToMeat> mcj: Man you're such a sweet person 21:54 < dotcomboom> there's WhatTheFont 21:54 < Pentode> the l's are different 21:54 < electrosys> how do i access the openvpncli? 21:54 < electrosys> dpesm 21:54 < dotcomboom> but I doubt it'd work for terminal fonts like that 21:55 < electrosys> doesn't seem to be with openvpn its self, man says /usr/local/sbin something but it doesn't exist in my machine. locate openvpn hmm... 21:55 < Pentode> its awfully similar though 21:55 < Pentode> the only difference i can even find so far is the lowercase l 21:56 < Pentode> maybe it _is_ terminus and the lowercase l looks different for some reason? 21:56 < Pentode> could be a different version 21:56 < SpeakerToMeat> mcj: Give me a second I'm trying to do you the favor, many mo9dern sites ocr the font ut your sample is pretty small 21:56 < dotcomboom> Terminus better edition (tm) 21:57 < dotcomboom> I just set my terminal font to comic sans to see how 1337 it looks and it looks weird for even comic sans standards 21:57 < mcj> SpeakerToMeat: Yeah I'm trying as well and I'm getting nothing helpful 21:57 < dotcomboom> things are squashed together xd 21:58 < mcj> Lol I think my sample has every letter in it :p 21:58 < mcj> Which is even more annoying that it can't find it 21:58 < dotcomboom> just remake the font 21:58 < dotcomboom> for your own personal use of course xd 21:58 < dotcomboom> last resort 22:01 < SpeakerToMeat> mcj: the sample lacks many of the identifying characters 22:03 < dotcomboom> it's a small bitmap font, OCR sites are used to large type fonts used by "graphic designers"(tm) 22:03 < dotcomboom> your best bet is probably searching around 22:03 < SpeakerToMeat> dotcomboom: I'm doing a manual identification 22:04 < hexnewbie> Bitmap fonts are easy and possibly legal to replicate from a screenshot 22:04 < SpeakerToMeat> But lacking many identifying characters, mainly uppercase characters (R, M, G, J) 22:04 < mcj> I'm just ognna find the guy and ask him hahah 22:04 < dotcomboom> oh, okay 22:04 < dotcomboom> icy 22:04 < spreeuw> the eternal quest for the best font, and rendering reproducability over stacks 22:04 < Maddix> not sure why you need a vm but you can always put windows onb a usb stick and boot that 22:05 < dotcomboom> Maddix adding on to that if you have another windows machine to use Rufus, and the ISO image for microsoft's website Windows To Go works albeit very slowly 22:05 < hexnewbie> mcj: I would be very surprised if it isn't one of the fonts found in apt-cache search bitmap font 22:05 < dotcomboom> I tried it with a 64GB drive 22:06 < dotcomboom> (usb 2.0) 22:06 < SpeakerToMeat> Fail, the closest identification is 21st but it's not 21st 22:07 < searedvandal> mcj, he shouldn't be too hard to find here on freenode 22:08 < dotcomboom> ah yeah 22:08 < dotcomboom> I mean, it is a weechat screenshot after all 22:08 < searedvandal> when you get the answer, let us know :) 22:08 < dotcomboom> #weechat 22:09 < wizzi> Hi, what do you recommend about cracking wifi? any ideas, tools .. 22:09 < Dominian> We don't discuss that here 22:09 < spreeuw> airkrak 22:10 < wonderer> hi, can anyone can help with this post? https://community.spiceworks.com/topic/2140751-apache-rewrite-rule-not-working-when-trying-to-block-site-ru-spam-referrers?page=1#entry-7776068 22:10 < Dominian> wonderer: Did you try #httpd 22:10 < wonderer> trying Dominian 22:11 < zapotah> wizzi: #security , though expect to be barred with homework 22:11 < wizzi> Dominian: i'm just learning and testing on my devices not on people 22:11 < momomo> I want to replace this wifi laptop card with a faster one, but what chipset, what connections is that matches? 22:11 < momomo> https://www.intel.com/content/dam/www/public/us/en/documents/product-briefs/centrino-wireless-n-135-brief.pdf 22:11 < momomo> Here is a search I made: https://www.prisjakt.nu/kategori.php?m=s364858195&o=produkt_pris_inkmoms 22:12 < DrunkRhino> So, anyone have any other ideas on how to get libc++(abi) to build properly? I tried modifying the makepkg to pass the -j 2 flag to ninja, but I think I might've put it in the wrong place. Also I have llvm-svn bins installed instead of llvm main since I use mesa-git as well. https://ptpb.pw/txK7 is a paste of "vmstat 3" that ran until I had to REISUB 22:12 < Dominian> wonderer: #httpd would be the best place to ask. 22:12 < wizzi> zapotah, thank you i'll see there 22:13 < hexnewbie> DrunkRhino: Get way more RAM. 22:13 < DrunkRhino> hexnewbie, would if I could, but I'm stuck for now with the 4GB I've got. 22:14 < zapotah> youre hosed then 22:14 < spreeuw> dont use j2 if you're short on ram 22:14 < hexnewbie> DrunkRhino: Is ninja something libc++ forces you to use? If not, use something saner that let's you pass -j2 or -j1, or read the docs 22:15 < zapotah> momomo: its a pcie half-mini chip 22:16 < DrunkRhino> hexnewbie, more or less. Even then I'm stuck trying to pass it via Arch's pkgbuild 22:16 < zapotah> momomo: as for whatever your antenna connector is, good luck finding a compatible chip 22:16 < spreeuw> put swap on ssd? 22:16 < spreeuw> put /tmp on disk to free up ram 22:16 < spreeuw> close all other programs 22:17 < spreeuw> etc 22:17 < zapotah> momomo: and you will only have a single antenna most certainly so you cant really stick anything in that requires two or more 22:17 < dotcomboom> I'd just go for a usb one.. or expresscard 22:17 < DLange> DrunkRhino: why do you think you run out of RAM? 22:17 < zapotah> momomo: if you do, you will burn out the chip 22:17 < DrunkRhino> Also, I get that 4 gigs isn't exactly spectacular these days, but I'd have thought it would be enough. spreeuw, that was my next thought. I've got an SSD in an old netbook that I've got set up as a Netflix 22:17 < zapotah> DrunkRhino: just dont install a gui and you will probly be fine 22:17 < DrunkRhino> zapotah, what? 22:17 < spreeuw> I'm not sure if its possible with 4GB 22:18 < zapotah> DrunkRhino: if youre building libc++, you dont need any kind of gui 22:18 < hexnewbie> Or don't build libc++ and get a binary package, if Arch has sane things like that. 22:18 < spreeuw> also there are compiling accelerators that tape up more ram by avoiding disk 22:18 < momomo> zapotah: hmm ..... so upgrading to this: https://www.prisjakt.nu/produkt.php?e=2980218 ... will cause issues? 22:18 < spreeuw> *take up 22:18 < zapotah> momomo: im not clicking that 22:18 < DrunkRhino> DLange, htop shows the RAM usage going up before freezing, and the vmstat looks like it's going nuts with the swap once it does. 22:18 < momomo> or this: https://www.prisjakt.nu/produkt.php?e=2980218 ... its just a product link 22:18 < spreeuw> you may want to check if they are in use but.. 22:18 < DLange> so let it swap, that's what it is there for 22:18 < zapotah> DrunkRhino: youre not using a gui on this box youre trying to build on right? 22:18 < DLange> just takes longer then 22:18 < spreeuw> another thing may be distributed gcc 22:19 < spreeuw> so you can use resources of another host 22:19 < spreeuw> proabbly complex to set up 22:19 < DrunkRhino> zapotah, considering it's a general purpose tower, yes. But I suppose I could fire it up, stop GDM and give it another go. 22:19 < spreeuw> or just ocmpile it elsewhere on a vm copy of the machine with more ram 22:19 < DLange> simple solution: use Debian, runs fine with 4GB RAM :) 22:19 < spreeuw> then put the binaries back 22:19 < DrunkRhino> spreeuw, don't have DistCC or CCache set up on this. 22:20 < zapotah> DrunkRhino: just for building that, just dont start the ui elements 22:20 < zapotah> DrunkRhino: that shit uses a fuckton of ram 22:20 < zapotah> in this 4GB scenario anyway 22:20 * zapotah lols at that 22:20 < dgurney> momomo, that's an m.2 card, so it wouldn't even fit in a mini-pci slot 22:22 < DrunkRhino> zapotah, well, looks like I'll just have to do it later on then. Not like I have another machine to switch to and use as it's doing that. 22:22 < momomo> dgurney: i see ... this is tricky .. im feeling this card is too slow .. must upgrade ... laptop is a monster .. so should be able to 22:22 < DrunkRhino> Hell, most of my set up for most of my machines is pretty much shoestrings and duck tape at this point. 22:22 < zapotah> momomo: thats funny 22:23 < dgurney> momomo, you will be able to upgrade to a better mini-pcie card, but even then you need to be careful, because some laptops have whitelists 22:23 < wizzi> zapotah, No one is serious there ...i think this is the best channel i see! 22:23 < zapotah> momomo: i heard a customer say "theres tons of resources available" 22:23 < momomo> dgurney: https://www.prisjakt.nu/produkt.php?p=2300705 this one looks the same on the pics 22:23 < zapotah> wizzi: keep trying, this is not the place for your question either 22:24 < zapotah> momomo: what that fellow meant was that theres an ARM dual core with 2gb of ram 22:24 < momomo> dgurney: i think these matches: https://www.prisjakt.nu/kategori.php?o=produkt_pris_inkmoms&m=s364859687 22:24 < zapotah> we terminated that client 22:24 < momomo> zapotah: ARM dual core? 22:24 < dotcomboom> momomo I'd use an expresscard (if your laptop has support for it) or an external usb adapter 22:24 < dgurney> momomo, yes, those are mini-pcie cards, but if your laptop has a bios whitelist, they will not work even if they fit in the slot 22:24 < wizzi> zapotach, yeah i'll google it or asking on another place 22:25 < dotcomboom> momomo I 22:25 < dotcomboom> I'd recommend if you're going to use an internal card to look up your laptop model for compatibility 22:25 < dgurney> ^ 22:26 < dgurney> all modern Lenovo laptops that I know of have whitelists (due to regulations) for instance 22:26 < momomo> dotcomboom: it's a custom built laptop from utopia computers ... 03:00.0 Network controller [0280]: Intel Corporation Centrino Wireless-N 135 [8086:0892] (rev c4) 22:26 < dgurney> hmm, in that case it probably doesn't have a whitelist 22:26 < DrunkRhino> zapotah, do you think spreeuw's suggestion of putting /tmp on an SSD would speed it up significantly? If it'd just be a slight boost I'd rather not bother fussing with getting it out of the netbook and mounted in the case 22:26 < momomo> dgurney: "modern" -> since a russian company took over ibm 22:26 < momomo> 's 22:27 < dgurney> *chinese 22:27 < momomo> chinese 22:27 < momomo> yes, sorry 22:27 < Reventlov> shitty lenovo 22:27 < zapotah> DrunkRhino: depends, you can pretend you have "ram" by expanding on swap 22:27 < spreeuw> its not about speed 22:27 < zapotah> DrunkRhino: wether that is fast or not, who knows 22:27 < spreeuw> its about having enough ram at the peak ram use 22:27 < zapotah> yeah 22:27 < spreeuw> to complete it at all 22:28 < dotcomboom> momomo, if it's a custom build, did they give you faster wifi card options when ordering? if so then try looking them up 22:28 < momomo> dotcomboom: don't remember .. it was a while ago, but worth looking up .. i think they still sell this chassi 22:28 < DrunkRhino> Sorry, yeah. That's more or less what I'd meant. The progress bars for the tests it was running prior to build were giving me estimates of 25-30 mins, but with the freezing and all that was definitely a lowball estimate 22:29 < zapotah> spreeuw: sometimes its funny shit tbh when theres timeouts in the build scripts :D 22:30 < momomo> dotcomboom: https://www.utopiacomputers.co.uk/haze-15 ... it says m.2 interface 22:30 < dgurney> momomo, when did you buy the laptop? it's possible they updated the model 22:31 < dotcomboom> they might not have used m.2 for the wifi, just ssds but I'm not sure, not an expert on these things 22:31 < momomo> dgurney: like 4 year ago :s 22:31 < AndroidKitKat> whats a program i can use to manage mutliple screens in i3 22:31 < ntd> https://j.ludost.net/blog/archives/2018/06/13/ancient_su_-_hostile_vulnerability_in_debian_8_and_9/ 22:32 < dgurney> momomo, so it's outdated then 22:32 < momomo> probably .. i will call them tomorrow .. thanks for the help :) 22:32 < dgurney> in any case, since those utopia machines seem to be just rebadged Clevo machines, it's possible that a mini-pcie ac card will work in it 22:32 < dgurney> but I suppose it's best you ask them directly 22:32 < dotcomboom> good idea to contact them 22:33 < NGC3982> hi guys 22:33 < dotcomboom> hio 22:34 < Lope> gcc -v shows 7.x.y I've installed gcc-5 22:34 < Lope> how can I select it as the default? 22:34 < NGC3982> im on a coorporate network and in an old unix machine. on this network, i can login to a mainfram machine with a hostname. can i in some fashion resolve what kind of machine is running the mainfram from simply the hostname? 22:35 < zapotah> dotcomboom: dgurney: or just, open the thing and see whats what 22:35 < NGC3982> the user i have on the mainframe login itself is locked and cannot be reached. 22:35 < RayTracer> Lope: maybe with update-alternatives 22:35 < zapotah> but, i guess thats a hazard considering the obvious level of the guy 22:35 < NGC3982> type of host is "zSeries" 22:36 < zapotah> NGC3982: no 22:36 < Lope> raytracer it didn't work. I tried running update-alternatives --remove-all gcc; update-alternatives --remove-all g++ 22:36 < Elladan> NGC3982, a hostname is just a name in some database, it doesn't have any relationship to the type of machine. 22:36 < Lope> I tried to use a similar syntax as I used for java but I couldn't find it at /usr/lib/gcc/5.5.0/gcc or bin/gcc etc 22:36 < RayTracer> Lope: maybe add your desired gcc with update-alternatives instead of removing the others 22:36 < NGC3982> Elladan, zapotah: i see. thank you. 22:37 < jprjr> NGC3982: hostnames are entirely arbritrary. *maybe* you could use something like nmap to attempt detecting what kind of host it is based on how it responds to IP traffic 22:37 < NGC3982> i was thinking about something like a ctcp version in irc. 22:37 < NGC3982> all i know is that it is a local machine, is probably very old and is of type zSeries. 22:37 < jprjr> But nmap's output will just be a guess really 22:37 < NGC3982> the unix machine doesnt have nmap, sadly. 22:37 < Elladan> I wouldn't advise using nmap probing on a corporate network unless you have permission from IT security. 22:37 < NGC3982> and im i no power to install things. 22:37 < NGC3982> yeah. 22:37 < dgurney> NGC3982, I suppose the easiest thing to do is simply ask the right person(s) 22:37 < SpringSprocket> aren't there dhcp-based hostnames too? 22:38 < SpringSprocket> ie host-label rather 22:38 < NGC3982> dgurney: of course, but i was interested in how to solve it myself, just for fun. 22:38 < jprjr> NGC3982: question: what's the actual problem you're trying to solve? Like why do you need to know what kind of machine it is? 22:38 < RayTracer> NGC3982: cat /proc/cpuinfo? 22:39 < NGC3982> jprjr: it is not a problem. im just interested in what kind of machine it is, and what tools i can use in a tight enviroment to reach it. 22:40 < NGC3982> RayTracer: the machine itself does not run *nix commands. my guess is that its a zSeries IBM mainframe running something like AS400. 22:41 < NGC3982> as i said, its just for fun and learning. simply not getting there is totally ok. 22:41 < NGC3982> the thing is, i have access as a user to the mainframe, but the user enviroment is so locked out i simply cant enter commands that arent exactly work related (so absolutely no system information can be found whatever i try). 22:42 < Elladan> SpringSprocket, there are a variety bits of information dhcp can hand out, and often the dhcp server and the dns server/database are connected. 22:42 < jprjr> NGC3982: yeah best bet is to just ask somebody. My only experience with AS/400 is using a pre-made menu system, I don't think I've ever had shell access on one of those 22:42 < NGC3982> they only thing pointing to IBM zSeries is the fact that the configuration file on the windows client specifies it, but that might just be default. 22:42 < NGC3982> jprjr: yeah 22:42 < SpringSprocket> nice 22:43 < Elladan> In corporate settings dhcp, dns, and switch port configuration may all be connected for security reasons on the back end. 22:45 < NGC3982> i see. 22:46 < Lope> RayTracer: i just had to run `HOST_COMPILER=g++-5 make` thanks 22:47 < Psi-Jack> triceratux: Wow. That's actually more serious of a problem than I expected. Every time I reboot, the swagarchrepo breaks things again... and again.... 22:48 < RayTracer> Lope: thanks as well :) 22:49 < triceratux> Psi-Jack: well i finally got that grub2 stanza of yours working. i know the volume labels involved so i dont need it to be quite so fancy http://pastebin.centos.org/840346/raw/ 22:49 < ggVGc> is there any terminal emulator or similar that would let me run it y-inverted? 22:49 < ggVGc> e.g input line at top and lines moving downwards 22:50 < triceratux> Psi-Jack: i had no idea thats still possible. its horridly documented in the wiki :P thank you much for calling it to my attention 22:50 < Psi-Jack> Hehe 22:52 < Psi-Jack> I'm beginning to wonder why you think SwagArch is so awesome, when the pacman-keyring itself is so utterly broken. 22:52 < fnDross> how well does nix run on 32bit processors? 22:53 < Psi-Jack> fnDross: nix? 22:53 < revel> NixOS, presumably. 22:54 < Psi-Jack> I do not presume/assume so quickly. ;p 22:54 < revel> Might as well ask #nixos instead, not sure. 22:54 < fnDross> got an old pc server that has XP32bit on it, and now needing GPT for a 3tb drive 22:54 < revel> I would, considering I've never seen anyone call anything else "nix" here. 22:54 < SpringSprocket> fnDross: I personally think it runs better than 64bit but it's just philosophical at this point 22:54 < electrosys> ahh , i need nmcli .. 22:54 < triceratux> Psi-Jack: mainly because it wasnt broken until the most recent iso & he looks like hes working on it. im not seeing any issues today either so its something real minor. & thats the only thing wrong with the distro 22:55 < Psi-Jack> triceratux: Well, I just installed it today.... And every time I reboot, pacman-key is completely fubar. 22:55 < NGC3982> revel: but, *nix? 22:55 < SpringSprocket> fnDross: ie sometimes some process don't use multiple threading which comes out using only 50% CPU 22:55 < NGC3982> i thought we was talking about general unix based systems. 22:55 < revel> "nix", not "*nix" 22:55 < Psi-Jack> fnDross: Well, the 3TB HDD may be an issue. 22:55 < NGC3982> i did not assume he cared about the asterisk :-p. 22:56 < SpringSprocket> designing program with multithreading hasn't followed splitting cores equally 22:56 < NGC3982> but then again, i never knew about nixos. 22:56 < triceratux> Psi-Jack: believe me youre not the only one whos noticing that. its just that his bugtracking is disrupted because he scurried to gitlab so fast :D 22:56 < fnDross> it is.... need a 64bit os to read the 3tb drive properly 22:56 < fnDross> from what ive read 22:56 < Psi-Jack> fnDross: nix? -- please clarify, too. 22:56 < Psi-Jack> triceratux: Oh? He ran from github to gitlab? 22:56 < fnDross> meaning the many flavors of linux 22:57 < NGC3982> oh lord i was actually right for once. 22:57 < triceratux> Psi-Jack: yep all the bugreports are 4 days old & are 404ing on github 22:57 < revel> wut 22:57 < triceratux> Psi-Jack: distrowatch is updated already 22:57 < fnDross> freeenas,openas,ubuntu,puppylinux etc 22:57 < SpringSprocket> fnDross: most distributions cover nicely the 32/64bit nicely, it's all done throughout the libs and program rather 22:58 < SpringSprocket> I'd say nowaday 64bit is best covered 22:58 < SpringSprocket> although 32bit is very well established, just not updating as much as before 22:58 < fnDross> hoping to setup a dualraid clone that backsup to the 3tb usb drive 22:59 < Penguin> The term "nix" has historically meant UNIX and UNIX-like stuff. 22:59 < fnDross> oh 22:59 < fnDross> lo 22:59 < fnDross> l 22:59 < Psi-Jack> Aha! 22:59 < Psi-Jack> I found t he issue. 22:59 < Psi-Jack> Heh. /etc/pacman.d/gnupg is freaking tmpfs... 22:59 < triceratux> https://github.com/SwagArch/swagarch-build/issues/81 23:00 < Psi-Jack> Why swagarch why!? 23:00 < triceratux> Psi-Jack: they want to make sure you can fix *all* the issues, not just most of them ;) 23:00 < SpringSprocket> 64bit has better/faster addressing through HDD/sdd from what I read, due to adressing translation 23:00 < Psi-Jack> But, /etc/pacman.d/gnupg as tmpfs ruins custom repos too. 23:01 < revel> Penguin: And then a package manager/OS stole that. 23:01 < fnDross> and can i use Microsoft's exFAT, read something about issues on it 23:01 < Penguin> fnDross: "UNIX-like" includes the flavors of Linux. 23:01 < Penguin> revel: I guess they're allowed to do that... to make things confusing. 23:01 < Psi-Jack> triceratux: Heh, and that github URL doesn't work. 23:02 < triceratux> Psi-Jack: that was the report that --populate fails at installtime but is ok after a reboot. so his keyserver is borking out but if pacman is configured with proper persistence a swagarch user doesnt notice it 23:03 < triceratux> but theres no such thing as a swagarch user even tho the distro is between 50-70 on dw 23:04 < Psi-Jack> Ugh, I see the problem is deeper. 23:04 < Psi-Jack> Freacking /etc/systemd/system/ installed pacman-init.service and etc-pacman.d-gnupg.mount. 23:04 * Psi-Jack wants to slap the hell out of SwagArch's maintainer. 23:06 < bls> welcome to a fork of a fringe amateur distro 23:06 < dgurney> swagarch is such a dumb name 23:07 < triceratux> i love it tho. i get bored unless my linux is crumbling before my eyes 23:07 < Psi-Jack> Heh 23:07 < plexigras> running `sudo nmcli dev wifi connect ... password ...` returns `Error: Connection activation failed: (7) Secrets were required, but not provided.` 23:07 < bls> nah nah man, it's swaggy 23:07 < Psi-Jack> Welp, Antergos was better, minus the conflicting packages. 23:07 < Psi-Jack> But.. I figured I'd give it a try regardless. 23:07 < bls> is Swaggy-P their mascot? 23:08 < plexigras> has anybody an idea what i could be doing wrong? 23:10 < electrosys> :plexigras it looks like you may need to pass along a password or something for the connection 23:10 < plexigras> what? but i am??? 23:11 < electrosys> plexigras: for now try -a and see what you get 23:11 < plexigras> is that --ask? 23:12 < plexigras> same error :/ 23:13 < Psi-Jack> There, disabled the freaking tmpfs pacman.d/gnupg nonsense. 23:13 < electrosys> yeah ask 23:13 < plexigras> with -a or --ask and entering the password manually that is 23:13 < electrosys> dont pass a password and use -a or --ask 23:13 < plexigras> thats what i have done and i still get the same error 23:13 < Psi-Jack> Yeah, NOW everything works. :) 23:14 < electrosys> plexigras: look at secret agent as well 23:14 < triceratux> Psi-Jack: how do we mere mortals do such a thing ? 23:14 < plexigras> electrosys: what is that? 23:14 < ayecee> petition the gods to intervene, presumably 23:15 < Psi-Jack> triceratux: systemctl disable etc-pacman.d-gnupg.mount; systemctl disable pacman-init.service; reboot; pacman-key --init; pacman-key --populate archlinux swagarch; profit. 23:16 < electrosys> plexigras: have you -> man nmcli ? 23:16 < plexigras> yes there is nothing about secret or agents on there 23:16 < electrosys> *cough bs cough* 23:16 < djph> well, perhaps not "agents" 23:17 < plexigras> oh wait i was on networkmanagers man page 23:17 < electrosys> your looking specifically for "secret agent" ;) 23:18 < plexigras> i'm currently reading the man page 23:18 < Psi-Jack> triceratux: It's VERY annoying too that they installed those systemd units to /etc/systemd, not where they belong in /lib/systemd 23:18 < plexigras> i also tried connecting using nmtui and nm-applet 23:18 < SpringSprocket> Psi-Jack: they just want to mimic sysvinit 23:19 < Psi-Jack> SpringSprocket: Well, it's wrong. 23:19 < SpringSprocket> it's always been wrong 23:19 < electrosys> yeah, agents sounds like some server thing, maybe that isn't what you need. you just need to pass creds 23:20 < SpringSprocket> personally I think only sysvinit could have it right this way 23:20 < electrosys> im not 100% sure though, im trying to learn nmcli myself. 23:20 < MarkusDBX> is there some new/beta version of glx that handles opengl forwarding slightly better? 23:20 < MarkusDBX> playing around with iglx 23:20 < electrosys> anyone know how to get dosbox going in the framebuffer? I think maybe you just need to get the SDL subsystem launched? 23:21 < plexigras> maybe i should just restard my pc 23:21 < electrosys> plexigras: that sounds like a windoze solution 23:21 < plexigras> well i'm all out of ideas 23:21 < electrosys> do you have the command params in the right order? 23:22 < plexigras> thats an idea i should try to mix it up 23:22 < bls> if only these things were documented so you didn't have to guess at random stuff 23:22 < plexigras> but right now i'm using the same order as in the arch wiki article 23:23 < electrosys> bls: they are documented, its just that theres a little bit of a learning curve when you first approach a program 23:26 < electrosys> plexigras: nmcli connection ? 23:26 < electrosys> plex-> nmcli connection help 23:27 < ntd> https://betanews.com/2018/06/14/the-firefox-powered-cliqz-web-browser-puts-your-security-first/ 23:27 < ntd> lol 23:28 < ntd> i remember "cliqz" 23:28 < lnnb> is it just me or do other people also getting OCSP failures all over the interweb the last few weeks? 23:28 < ntd> and i don't exactly associate it with "security" 23:28 < ayecee> lnnb: probably just you 23:29 < electrosys> plexigras: but wait, are you using arch? if you are your probably not using connection manager, maybe connman? 23:29 < electrosys> i mean not using network manager 23:29 < zapotah> lnnb: someone counting on that not verifying shit just passes as OK 23:29 < electrosys> i think network manager is a gnome thing, no? 23:30 < bls> no, it's lower level than that 23:30 < lnnb> yeah in my browser you have to specifically say "treat ocsp connections failure as error" 23:31 < plexigras> yes im using arch but i have set up networkmanager not connman 23:31 < Psi-Jack> SpringSprocket: Eh, nothing was ever "good" about sysvinit. lsb-init had some benefit, but it still was problematic. 23:31 < electrosys> ok 23:32 < SpringSprocket> Psi-Jack: sure, but it was linear and well ordered 23:32 < Psi-Jack> ... 23:33 < SpringSprocket> Psi-Jack: there were only few scripts wich were long to read through 23:33 < Psi-Jack> Linear, yes... Well ordered, however... It was easily prone to race conditions. 23:33 < Psi-Jack> Problem is, stupid people kept putting set -e in their init scripts. 23:36 < revel> What's so bad about `set -e`? 23:38 < SpringSprocket> it disciminates error handling 23:38 < SpringSprocket> which should be implemented anyway 23:38 < ayecee> disciminates 23:38 < revel> Disciminates? 23:38 < Psi-Jack> revel: Per LSB-init standards, it violated everything. 23:39 < Psi-Jack> LSB-init expected certain well-defined return codes based on different statuses. 23:41 < revel> You could `set -eE` and `trap $somefunction ERR` 23:41 < DrunkRhino> Anybody happen to know if there's a quick way to force all applications under GNOME to use a title bar? Some of my windows have them and the window controls are smaller there than in apps like Nautilus, which stick the title only up in the top bar AFAICT 23:42 < DrunkRhino> I figured it'd be under GNOME tweaks if there was but I don't see anything like that there. 23:42 < tymczenko> is `/etc/rc.local` a good place to set screen lightness on startup? 23:43 < ayecee> what are the other options? 23:43 < tymczenko> that's what I'm asking you, guys) 23:43 < electrosys> tymczenko: maybe ~/.rc.local ? 23:44 < electrosys> i think /etc is where the sampes or system wide go. 23:44 < revel> ... That's a thing? 23:44 < electrosys> but im still learning. 23:44 < ayecee> it's not a thing 23:44 < ayecee> tymczenko: set how? 23:44 < revel> Whew. 23:44 < tymczenko> ayecee, something like `echo X > /sys/class/backlight/intel_backlight/brightness` where X = some number 23:45 < revel> Some number below /sys/class/backlight/intel_backlight/max_brightness 23:45 < tymczenko> revel, yeah, sure 23:46 < ayecee> *shrug* 23:46 < ayecee> guess that can't be done with sysctl 23:46 < milpool> of all the stuff that i put into /etc/rc.local, setting brightness is one of the more reasonable ;) 23:47 < bls> could it be specified as an argument to a kernel module or is backlight support compiled in? 23:48 < tymczenko> milpool, seriously? so it's a decent way of setting brightness? 23:48 < Aleric> Hi... what is the name of this pop-up thing that I get while I alt-tab over all open windows? 23:48 < revel> Steve. 23:49 < searedvandal> tymczenko, there are a couple of ways to set brightness 23:49 < birdbolt1> this channel is a lot quitter than i expected 23:49 < milpool> tymczenko: i don't know, tbh. i put stuff like that in rc.local if it works, but i wouldn't say it's "the right way". i just don't care. 23:49 < birdbolt1> quieter * 23:53 < tymczenko> milpool, ok, thanks for your help 23:54 < tymczenko> searedvandal, writing a number to `/sys/class/backlight/intel_backlight/brightnes` is ok, or should I consider some other way? 23:56 < searedvandal> tymczenko, it should be fine doing it that way. or you can add a udev rule and do it that way. I think the arch wiki has a decent page about backlight 23:57 < tymczenko> searedvandal, thanks --- Log closed Fri Jun 15 00:00:58 2018