--- Log opened Tue Jun 19 00:00:04 2018 00:00 < phogg> Tostane: were all of the files on those disks owned by a single user? 00:01 < mawk> degenerate, you forgot --append no ? 00:01 < Dexx1_> Hey guys. When I put in a sdcard into my sdcard reader, the drive doesn't show up. How can I get it working? 00:01 < guest32> hi 00:01 < phogg> yeah, I had just come to the same conclusion 00:02 < phogg> Dexx1_: check dmesg, it will have noticed the drive if anything did 00:02 < phogg> Dexx1_: or, rather: doesn't show up *where*? 00:03 < Tostane> Ill check the owner on the files 00:04 < useless-eater> hi guys, I got a silly question probably, but all of a sudden my computer does not allow me to mv a directory with content. wth?! why? 00:04 < MrElendig> Dexx1_: what does dmesg say and what hardware is this? 00:04 < MrElendig> and the distro 00:04 < Tostane> I did try to use gparted and disks utilities to mount the drives I remade the uid I tried mounting them in mnt and media no matter how I did it they would not show on those 2 drives 00:04 < phogg> useless-eater: "does not allow" in what way? 00:05 < phogg> Tostane: if you can see the contents of the drives as root then the data is there. 00:05 < useless-eater> phogg: "Directory not empty" 00:05 < phogg> Tostane: Do you know how file ownership and permissions work on Unix filesystems? 00:06 < phogg> useless-eater: that should only happen if you try to rmdir a directory. Are you sure you do not have a bad alias like: alias mv=rmdir? 00:06 < Dexx1_> phogg: MrElendig this is my dmesg: https://pastebin.com/raw/UQWUdVds 00:06 < MrElendig> Tostane: which filesystem, and are these internal disks? 00:06 < useless-eater> seems like my mv command was lobotomized D: 00:06 < MrElendig> Dexx1_: please use a sane pastebin 00:06 < MrElendig> useless-eater: `type mv` 00:06 < useless-eater> mv is /bin/mv 00:07 < Dexx1_> MrElendig: like what? 00:07 < phogg> Dexx1_: try termbin, ptpb, sprunge 00:07 < phogg> paste.debian.net is also fine 00:07 < phogg> anything with a raw mode and not requiring JS 00:07 < MrElendig> ptpb, gist, bpaste 00:08 < MrElendig> phogg: and doesn't mangle what you post, and doesn't have malware installing ads etc etc 00:08 < phogg> MrElendig: indeed 00:08 < Tostane> the original system and this I use the same username so the owner is same 00:08 < Tostane> only change I made was the password 00:08 < bls> same username does not mean the same uid 00:09 < MrElendig> Tostane: username means nothing, the uid is what matters 00:09 < Dexx1_> phogg: isn't this raw mode?? https://pastebin.com/raw/UQWUdVds 00:09 < MrElendig> the username is just for convenience for the stupid fleshbags between the keyboard and chair, the computer only cares about the uid/gid 00:09 < MrElendig> you can even have multiple users with the same uid, or a uid with no username 00:10 < bls> Dexx1_: pastebin.com is blacklisted in here 00:10 < MrElendig> Dexx1_: /msg phrik badpastebin 00:10 < Tostane> I did not use uid for my drives 00:11 < Tostane> I put a volume label on each then mounted them with that lable 00:11 < bls> Tostane: yes you did. it's now POSIX filesystems work 00:11 < bls> s/now/how/ 00:11 < balletjebal> gents ,on windows theres a key to copy paste text like ctrl c or ctrl v does it work like the same on linux ? 00:11 < bls> balletjebal: depends on the application 00:11 < Dexx1_> phogg: ok, now I have a different problem. How do I install these pastebin programs? (like termbin ptpb sprunge etc) because sudo apt-get install doesn't work 00:12 < MrElendig> you don't 00:12 < koala_man> copypaste in X is complicated 00:12 < MrElendig> you just need curl 00:12 < bls> balletjebal: most people just use highlight-to-copy and middle-button-to-paste 00:12 < phogg> Dexx1_: dmesg | nc termbin.com 9999 00:12 < MrElendig> 2>&1 | curl -F c=@- https://ptpb.pw 00:12 < MrElendig> curl -F c=@path/to/a/file https://ptpb.pw 00:12 < Dexx1_> phogg: MrElendig http://termbin.com/7ebi 00:12 < Dexx1_> ok, that's easy 00:12 < bls> balletjebal: some applications support Shift+Insert for paste 00:12 < MrElendig> ptpb also does images etc 00:13 < phogg> Tostane: the UIDs will still be in the filesystem from your previous install. If they do not match exactly you will have no permission to access the files, but root will be able to because root has all permissions. 00:13 < Dexx1_> (although, I see zero difference between that and my original paste) 00:13 < MrElendig> proper syntax highlighting etc 00:13 < MrElendig> Dexx1_: except that we can actually open it 00:13 < MrElendig> Dexx1_: and there isn't \r\n line endings 00:13 < phogg> Tostane: if you can find out the UID being used for the files and you know which user you *want* to be able to access the files I can help you automatically change the ownership. 00:13 < Dexx1_> ah 00:13 < MrElendig> Dexx1_: and no malware installing ads 00:13 < bls> and there's no chance of getting served malware with this link 00:13 < MrElendig> nor bitcoin mining 00:13 < balletjebal> loads of options there then :P 00:14 < MrElendig> no capcha 00:14 < phogg> Dexx1_: dmesg | curl -F c=@- https://ptpb.pw/ # to make MrElendig happier 00:14 < phogg> but I think not everybody has curl 00:14 < MrElendig> nice oops 00:14 < Tostane> I do a fstab /dev/disk/by-label/vol1 /mnt/vol1 auto ……. 00:14 < MrElendig> also, that is not the entire dmesg 00:14 < phogg> Tostane: it doesn't matter 00:14 < bls> Tostane: not relevant to your current problem 00:14 < Tostane> I 00:14 < phogg> Tostane: owner IDs are stored *in the filesystem*. 00:15 < phogg> Tostane: open up a terminal and run the 'stat' command on a file. 00:15 < Dexx1_> phogg: can you condense that command down to 'pastebin'? So I can dmesg | pastebin 00:15 < MrElendig> looks like you have some hardware issues with that machine 00:16 < MrElendig> of the kind that might warant some replacement 00:16 < Dexx1_> MrElendig: like what? 00:16 < Dexx1_> The sd card reader worked last week btw 00:16 < Tostane> if i open it from the gui and do a properties on a file it says unreadable 00:17 < MrElendig> laptop? 00:17 < Dexx1_> yes 00:18 < MrElendig> built in card reader? 00:18 < Dexx1_> yes 00:19 < Tostane> in terminal if i try to open that drive it says permission denied 00:19 < MrElendig> post lspci 00:19 < bls> Tostane: do these things as root 00:19 < MrElendig> Tostane: which fs? 00:19 < MrElendig> Tostane: findmnt /wherever/you/mounted/it 00:19 < Dexx1_> MrElendig: http://termbin.com/r3db 00:20 < Dan39> dmesg...? 00:20 < Dexx1_> Dan39: talking to me? 00:20 < Dan39> nvm see above 00:20 < Tostane> im seeing some message system-product-name-invalid-length-Id-Fiixed-up-to-11 00:20 < Dan39> found it :P 00:21 < Tostane> I changed to the server/media folder the drive is in there 00:21 < MrElendig> hm so pci connected card reader 00:21 < Tostane> i did a stat on the drive itself and got that error 00:22 < phogg> Dexx1_: yes, but so can you. Just wrap it in a shell function 00:22 < Tostane> ill try doing a stat on a drive that works 00:22 < phogg> Dexx1_: or drop it into a shell script. If you don't know how this is a great time to learn. It's easy. 00:22 < MrElendig> a known buggy one too 00:22 < bls> stat is for getting information about files, it's not going to provide any useful information about the device entry 00:23 < Dan39> Dexx1_: what is up with all the traces and segfaults in dmesg? 00:23 < MrElendig> crappy intel wifi 00:23 < phogg> Tostane: the purpose of stat was so that you could find the numeric ID of the owner of a file. If you're running it against a device file you're not going to get anything useful. 00:23 < MrElendig> Dan39: lspci -vnn 00:24 < Tostane> ok another possible problem in termial is were im seeing that system product name invalid entery length could this Linux have not installed something correctly 00:24 < phogg> Tostane: where are you seeing that? 00:25 < bls> there's a definite chance something is screwed up, but right now it looks like you're chasing after ghosts instead of providing useful debug information 00:25 < MrElendig> er... Dexx1_ 00:25 < Dexx1_> Dan39: I dunno, you are the expert here :P 00:26 < Tostane> it shows up in termial as soon as termal loads 00:27 < Tostane> I googled it but there is just questions from other people having same problem it seems to be from ubuntu os 00:27 < phogg> I chased a ghost once. Ate it, too. Scored 200 points. 00:27 < phogg> Or maybe it was just a dream. 00:27 < bls> prolly just had pacman fever 00:27 < phogg> Tostane: when you say "terminal" do you mean "virtual console"? Sounds like your syslog is dumping some kinds of errors there. 00:28 < Tostane> i think when it installed linus auto generates a computer name and the name is messed up 00:28 < MrElendig> linus did no such thing 00:29 < phogg> Tostane: First of all, no. It doesn't do that, and if it did nothing would be messed up. 00:29 < bls> and none of this is relevant to whether or not your UIDs are wrong 00:30 < phogg> Tostane: you should focus on one problem at a time 00:30 < Dexx1_> MrElendig: yeah? 00:31 < MrElendig> Dexx1_: post lspci -vnn 00:31 < MrElendig> or just the vid/pid from it, 00:31 < MrElendig> for the card reader 00:31 < Tostane> ok i found out it is the lnux host name now im seeing how i change it 00:32 < Dexx1_> MrElendig: http://termbin.com/7neu 00:32 < phogg> Tostane: try the command 'hostname' for that 00:32 < bls> does that still work if systemd has taken control of the hostname? 00:33 < phogg> I neither know nor care. 00:33 < MrElendig> Dexx1_: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/717435 00:33 < Dagmar> It's always the host name 00:33 < MrElendig> Dexx1_: solution: buy some decent usb cardreader 00:35 < Dagmar> Apparently hostnamectl is the "new" tool for systemd for changing the host name 00:35 < Tostane> I changed the host file and rebooted it now 00:35 < Dagmar> The /etc/hosts file? 00:36 < Dagmar> That's not really where the hostname is set, but whatevs 00:36 < MrElendig> https://bugs.launchpad.net/linux/+bug/790754 too 00:36 < Tostane> ok it did not change for some reason 00:36 < Dagmar> The machine will only act marginally psychotic if it's IP address doesn't resolve to the same hostname 00:36 < bls> this is turning into some fine performance art... 00:36 < MrElendig> seems quite troublesome 00:36 < Dagmar> Tostane: Because /etc/hosts is not the place to set the hostname 00:37 < Dagmar> Toastane: `hostnamectl set-hostname darkstar` 00:37 < MrElendig> Dexx1_: https://github.com/aur-archive/ricoh-e823-fix/blob/master/e823fix.wrapper might work 00:37 < phogg> I thought even systemd still used /etc/hostname for the host name 00:37 < Dagmar> It *should* be 00:38 < MrElendig> it does 00:38 < phogg> they just pointlessly rewrote the tool you use to interact with it because mumble mumble nih muble progress mumble 00:38 < Dexx1_> MrElendig: any idea why it worked fine for the past year+? 00:38 < MrElendig> man 5 hostname 00:38 < Dagmar> I've yet to see any hipsters arguing there's something wrong with the way init handles the machine's own hostname 00:38 < Tostane> ok i changed the hostname file now 00:38 < MrElendig> Dexx1_: maybe related to all the kernel crashes? 00:38 < MrElendig> hardware going bad? 00:38 < veridiam> you'll have to file for a legal name change, it seems you live in an area where artificial life has power of attorney, you can't just change it's name 00:38 < phogg> by which I mean "The existing tool is completely broken. Doesn't work at all. Had to be replaced." 00:38 < MrElendig> #justubuntuthings 00:38 < Dagmar> heh 00:39 * phogg is just guessing based on the general trend 00:39 < Tostane> that fixed it i changed both hosts and hostname 00:39 < MrElendig> phogg: except that it is not pointless 00:39 < Dagmar> Again, that's not what /etc/hosts does, but whatevre 00:39 < phogg> Tostane: the hosts part will not matter. Very little software cares if it can reverse the name into an accurate IP 00:39 < Tostane> it got rid of the weird message in terminal so im happy 00:40 < Dagmar> One generally doesn't even _need_ the host machine's name in /etc/hosts 00:40 < phogg> Tostane: To be clear the thing that fixed it was the *second* thing you did. The first one had as much effect as waving a dead chicken. 00:40 < phogg> Probably less. 00:40 < Dagmar> The only thing that really needs to be there is the localhost entry 00:40 < bls> that's consistent with this. complain about something, ignore help, do something completely unrelated 00:40 < Tostane> the host used to show up from my windows boxes i am not sure they would like that long name 00:40 < phogg> Tostane: who cares what Windows likes? 00:40 < Dagmar> Good thing we don't have to care about what your windows machines think 00:41 < phogg> do what you WANT and drag them kicking and screaming into the 20th century 00:41 < Tostane> windows just for game 00:42 < Dagmar> The dead are easier to drag 00:43 < Psi-Jack> True statement ^^ 00:43 < Tostane> this drive problem is frustrating im tempted to run and buy a 8 terabyte drive and copy the files off and format it 00:44 < MrElendig> do that, and use the drive for backups afterwards 00:44 < Dagmar> Good luck with that 00:44 < phogg> Tostane: I assure you that doing so is not necessary. Your drive is probably fine. 00:44 < MrElendig> those backups you should have set up years ago :p 00:44 < phogg> Real men don't use backups. They just upload their work to public ftp and let the rest of the world mirror it. 00:44 < Dagmar> I think I'm going to be sticking with the 3Gb size for awhile 00:44 < Dagmar> er 3Tb 00:45 < Dagmar> It's all fun and games until SMART takes 14+ hours to complete 00:45 < MrElendig> 4tb is generally the best price/performance/size/reliability 00:45 < Dagmar> _Fourteen bloody hours_ 00:45 < MrElendig> though it is slowly creeping upwards 00:45 < phogg> 3T is about as big as god intended storage devices to get. 00:45 < phogg> just like 3 stories for buildings 00:46 < phogg> anything bigger is suspicious 00:46 < MrElendig> sidenote: the helium filled drives have turned out to be much more reliable than the naysayers predicted 00:47 < Dagmar> It was sound engineering 00:47 < Tostane> strange i opened the /media/drive as admin and changed the permission from root to the server name now i can see the files 00:47 < Dagmar> Those are some good drugs 00:48 < bls> heh, and the performance continues 00:49 < MrElendig> server name happens to be the same as your user name? :p 00:50 < phogg> MrElendig: Oh did they? I am pleased to hear it. Got stats on how much more? 00:50 < phogg> Tostane: that's all you ever had to do: change the owner of the files to one that exists on your system 00:51 < MrElendig> https://www.backblaze.com/blog/hard-drive-stats-for-q1-2018/ 00:52 < Tostane> the drives that were working are set as root also i wonder why it just did not like these 2 drives 00:52 < Psi-Jack> MrElendig: Nice. 00:52 < Psi-Jack> Ooh, they have HGST's up there. I've been wondering about those. 00:52 < bls> Tostane: some mysteries are best left to the universe 00:52 < Dagmar> Probably because they were formatted with ntfs or something equally insane 00:53 < bls> the more likely reason is that he had UID mismatches, but at least changing the drive's hostname fixed it :P 00:53 < MrElendig> the 12tb seagates are doing well 00:53 < Psi-Jack> They are.. Insanely well. 00:54 < Dagmar> HGST has been bringing their A-game 00:54 < MrElendig> still just a year old though 00:54 < phogg> a thousand blessings on backblaze 00:54 < Dagmar> Enough to make me shortlist them for spending of consulting money 00:54 < MrElendig> but some claimed they would leak in less than 6 months :p 00:55 < Tostane> i used ext4 on them i read there is another file system that is better for systems that sleep a lot that does not cache writes i wanted to try 00:55 < Psi-Jack> Seagate, in this, has been doing overall pretty dang good. 00:55 < Dagmar> MrElendig: They've been rocking it for ages, including when GOogle was publishing their drive failure stats 00:55 < MrElendig> the smaller seagate drives have had issues 00:55 < Psi-Jack> Yeah, the 4TB ones. 00:55 < Dagmar> Random Seagate models have had issues 00:55 < bls> doubts were valid given the history of the industry and their less than up-front marketing material 00:56 < Tostane> I was wanting to get some 12 segates the feedback on them is good that i found. 00:56 < Dagmar> ...because they didn't just burn everyone from Maxtor alive when the absorbed them 00:56 < Dagmar> ...or at least just give them a decent severance package 00:56 < MrElendig> https://www.backblaze.com/blog/helium-filled-hard-drive-failure-rates/ 00:57 < MrElendig> going to be interesting to see how it developes over a couple of more years 00:57 < Dagmar> When you're buying another company that's failing, and there's no IP transfer happening, you have to wonder about the people who think it's a good idea to keep the staff who failed in the first place 00:57 < Dagmar> Like, sure, maybe once our name is on the building they'll magically stop sucking 00:57 < bls> those things are usually, but not always, down to the suits, not the engineers 00:58 < Dagmar> bls: Except all the MAxtor bufu was clearly dumb engineering 00:58 < Tostane> sadly i bought a lot of maxstor drives to see them all fail 00:59 < bls> maybe, I've just been in the situation of sitting on a bad product, reporting that it was a bad product, and getting overrulled and have it get shipped in the hopes it won't be *that* bad, or we'll do damage control later 01:00 < Tostane> i mostly run segate now same drives for over 3 years flawless knock on wood its time to replace them soon 01:00 < Dagmar> That line of Seagates that they tried to hide the fact that they were _eating their own p- and g-list_... All the firmware upgrade tools had Maxtor showing 01:00 < Dagmar> Clearly that was Maxtor engineering at work... designing drives that would overwrite their own sector maps with SMART logs 01:00 < phogg> I still have some maxtor drives that are working. 01:01 < Dagmar> The Seagate drives I had a customer actually decide to *change my parts list* for over $2 a drive... The ones that would not reliably initialize on a SATA 2 controller... Maxtor again 01:01 < phogg> I bought a lot of maxtor drives. Once you have the jump configuration for a brand memorized it's annoying to change 01:02 < Dagmar> Entertainingly, even the tool to upgrade the firmware to fix it that I was given (which was for IDE disks LOL) also said Maxtor 01:02 < Dagmar> Should have said "MAXTOR BRICKING TOOL" 01:02 < Dagmar> It said "IDE PATA" right there on the screen, I told the CS rep, they swore it was the right tool, and... bricked 01:02 < phogg> I sense a certain amount of hostility here. Maxtor kicked Dagmar's puppy something fierce. 01:03 < Dagmar> It was a completely wasted afternoon on hold with idiots 01:03 < MrElendig> worse, nuked his .pron 01:03 < Dagmar> Bricked both the damn drives and had to ship them back 01:03 < Dagmar> ...and they wouldn't even cross ship 01:03 < Tostane> back in the day maxstor ran low on drives and was shipping real bricks instead before they went under 01:05 < Tostane> we got one of the first 40 gig maxstors and had to replace a transistor on it right away 01:05 < Dagmar> They were even more amusing when they were using the name Quantum 01:05 < Dagmar> The Quantum Failball 01:05 < Tostane> oh that is scary name quantum 01:06 < Dagmar> No one ever apparently thought to do any QA on the units, so _most_ of that line had bad cache RAM 01:06 < Dagmar> ...and that was the thing that was supposed to make the drive _faster_ than everyone else. 01:07 < Dagmar> If you bought them refurbbed where someone had probably already replaced that chip and tested them, they were great 01:07 < Tostane> when i close my store down i had to hall a large bed pickup truck full of hard drive to the city dump 01:07 < Dagmar> Otherwise, they were just thin 5 inch drives which failed early with massive data corruption 01:07 < Johnjay> why does anything in star trek work? Quantum. 01:07 < MrElendig> should go to recycle, not the dump 01:07 < phogg> Tostane: you wasted a great opportunity to experiment with thermite 01:08 < Dagmar> Or just some nerds who want a lot of magnets, and know where to take the chassis as scrap metal 01:08 < MrElendig> could build a giant hdd speaker wall 01:08 < Tostane> i wish i had them now it be fun to tear them apart and sell to the makers 01:08 < MrElendig> or armor up your cubicle 01:09 < MrElendig> armour* 01:09 < phogg> armor is fine, says the USA 01:09 < Dagmar> I would be fine with some light armor in my cubicle 01:09 < MrElendig> hdds at 30° would offer good protection against nerf projectiles 01:09 < Dagmar> Actually I was thinking small caliber bullets 01:09 < Dagmar> They'll certainly stop a 22 01:10 < Dagmar> ...then again, a thick leather jacket can stop a 22 01:10 < MrElendig> tell that to all the kids that died at utøya 01:10 < phogg> they don't need to completely stop a round, just divert its trajectory away from lethal areas 01:10 < MrElendig> .22 is quite deadly 01:10 < Tostane> we used to tear the drive down and replace platters way back in a make shift clean room and do component replacement on the board. that was in the 20 to 40 gig drive days 01:10 < phogg> MrElendig: they should have been wearing used hard disk armor 01:11 < Tostane> when a 40 gig drive costed 1500 lol 01:12 < Assimilater> this may be a stretch, I just don't know where else to turn because the nxp community isn't proving very helpful. I'm developing router software for my senior design, and I have an imx6ullevk I'm working on. It comes with a linux os but I'm trying to disable networking features on it so my software does all the networking 01:13 < phogg> Assimilater: it's hard to disable all of it and still have access to the hardware 01:13 < Cache_Money> When I created my id_rsa key I added a password. When I attempt to ssh into a server using my private key it always forces me to input my password. Is this because when I created it I used a password? If I created a new id_rsa key without a password would I be able to ssh without having to type my password every time? 01:13 < bls> Cache_Money: yes. yes. 01:13 < Assimilater> phogg: would I need a complete disable in order to prevent it from talking? 01:14 < bls> Cache_Money: or you can run ssh-agent or gpg-agent to cache the password for you 01:14 < Tostane> What is NXP 01:14 < Assimilater> nxp is a microcontroller manufacturer 01:14 < Cache_Money> bls: okay. I’ll create a new private key and not type a password this time 01:16 < Tostane> oh nxp is like arm 01:16 < Assimilater> kind of, they make arm stuff 01:22 < Cache_Money> bls: Here’s the start of my ssh config https://gist.github.com/Abundnce10/adbdcbb5d306981b7b623ad8ee802692 Doesn’t `UseKeychain yes` mean I wouldn’t have to type my password every time I try to ssh into a server? 01:24 < bls> Cache_Money: only if the keychain app is actually running, has the key added to it, and has had a valid password entered 01:25 < Cache_Money> bls: How do I test that? 01:25 < bls> and not sure what version of ssh you've got. don't see that option documented anywhere on my systems 01:25 < Tostane> Thanks for the ideas this seems to have got me going again. Im going to poke it with a stick a few times and see if any more bugs appear I was using Linux mint 17.2 64 bit on it. usualy it works pretty good. 01:25 < bls> Cache_Money: `pgrep -afl ssh-agent` `printenv | grep -i ssh` 01:25 < bls> also `ssh-add -l` 01:27 < Cache_Money> bls: okay. I’ll just re-create a new key and get it setup that way. Thanks for your help 02:00 < ansraliant> morning 02:05 < jim> ansraliant, hi 03:46 < johnc--> hi, i scheduled a posts on WordPress, but my site has no hits. so i try setting crontab from cpanel interface. after doing some google, i finally set to this. can anyone please tell me whether the syntax is correct or not? 03:46 < johnc--> 0 6 * * * wget -o /dev/null --timeout=120 http://mysite.com/wp-cron.php?doing_wp_cron=true >/dev/null 2>&1 03:46 < johnc--> once everyday at 6AM 03:47 < meingtsla> Surely WordPress has its own internal way of scheduling posts? 03:49 < meingtsla> That aside, you should fully qualify the path to wget. Do not assume that cron will have the same environment variables that you do. 03:51 < johnc--> meingtsla, i think the WordPress scheduled posts/cronjob activates only when a user visit a site. 03:53 < thatpythonguy> if I'm not using a login manager and I start X using startx, what is the 'correct' way to stop the x server? 03:54 < fryguy> thatpythonguy: close whatever process was the last to run in .xinitrc 03:54 < lnnb> sigterm? 03:55 < thatpythonguy> @fryguy, so basically, stop the WM 03:55 < meingtsla> johnc--: Ahh I see 03:55 < fryguy> thatpythonguy: usually, yes 03:55 < thatpythonguy> @lnnb I'm not familiar with that 03:55 < lnnb> i think its the default for kill command 03:55 < meingtsla> johnc--: So that would run once daily at 0600. 03:56 < johnc--> so, the syntax is correct? because i googled and modify as far as i can understand. 03:56 < lnnb> if killing wm takes down the server it's not set up right 03:56 < thatpythonguy> i was hoping there was a graceful way to stop x haha, oh well. if I did stop the WM, would the x server be 'closed'/inactive? 03:56 < lnnb> sigterm is graceful 03:57 < lnnb> well, it's supposed to be at least 03:57 < lnnb> sigkill is the nasty one 03:57 < thatpythonguy> @lnnb how do I invoke sigkill? 03:57 < thatpythonguy> oops meant sigterm 03:57 < meingtsla> johnc--: See my previous comment about fully qualifying the path to wget. Also, I might do -O instead of -o. 03:58 < lnnb> kill 03:58 < thatpythonguy> does x have a process id tho? i mean can you just kill x like that 03:58 < lnnb> every process has an ID 03:59 < johnc--> meingtsla, i got disconnetecd maybe i miss it. is that zero or capital O 03:59 < thatpythonguy> what about ctrl-alt-backspace? 03:59 < lnnb> might or might not work 03:59 < thatpythonguy> depeding on distro I guess? 03:59 < meingtsla> johnc--: 184931 < meingtsla> That aside, you should fully qualify the path to wget. Do not assume that cron will have the same environment variables that you do. 04:00 < domhnall> johnc--: it's a capital 'O' 04:00 < lnnb> yeah i think XOrg terminology calls that "zap" which is configurable at compile time 04:00 < johnc--> ok 04:01 < thatpythonguy> @lnnb so basically, ctrl-alt-backspace is gonna kill x and go to a tty1-6 unless the user has specifically messed with that shortcut in some way 04:01 < thatpythonguy> correct? 04:01 < johnc--> thanks meingtsla, domhnall. 04:02 < lnnb> *if* the feature is supported, i think it is reasonable to expect that to happen if x was started from a TTY 04:04 < thatpythonguy> i wish there was a stopx command lol, i just figured out that 'zap' only restarts, i was looking for kill 04:07 < bb36e> Is there any Ctrl+alt+del equivalent in Linux? I guess alt+sysrq is kind of similar…just wondering 04:08 < ansraliant> funny thing, ctrl+alt+backspace never worked for me for some reason. I just kill the Display Manager, and it kills the X... worked most of times. 04:08 < thatpythonguy> @ansraliant that would work except I'm migrating to using no Display Manager and instead booting into console and running startx 04:09 < bb36e> I think many distros disable ^alt+del by default, so now you have to manually re-enable it 04:09 < bb36e> **disable ^alt+backspace 04:10 < thatpythonguy> I'm doing some research and it seems like there's basically no simple command to stop x 04:10 < domhnall> ctrl+alt+del does nothing on macbook keyboard. (running linux) 04:11 < domhnall> thatpythonguy: usually just doing 'exit' works for me. But not sure what startx brings up on other's distro. For me it's 3 terminals. 04:11 < thatpythonguy> ctrl-alt-delete only restarts it, killing all x processes is ugly and not graceful, stopping the login manager doesn't work if one isn't running... i'm not sure there's a simple, clean way 04:12 < thatpythonguy> @domhnall, startx runs ~/.xinitrc which most people use to start their Window Manager (like gnome, unity, etc) 04:13 < thatpythonguy> using 'exec gnome-session' or the like inside .xinitrc 04:13 < domhnall> think I'd be starting xfwm or whatever it is... 04:14 < thatpythonguy> running 'exit' in a tty1-6 might actually work because it would close all the current user's processes, including x server 04:14 < kittonian> hi all. has anyone here successfully got calendarserver 9.2 up and running? I have been struggling all day long with this and cannot get it to work (and there's no docs) 04:19 < Tech_8> suup 04:20 < bb36e> thatpythonguy: apparently alt+systq+k kills all processes on your current vtty, might help? 04:20 < thatpythonguy> @bb36e I can see how that would also work, but ideally the method would only stop x server and not all other processes 04:20 < thatpythonguy> I appreciate it tho 04:26 < ascii8bom> anyone have issues with channging screen brightness on KDE Plasma? the only way i've been able to control the brightness is via 'echo (VALUE HERE) > /sys/class/backlight/intel_backlight/brightness' 04:27 < ansraliant> I used to have problems changing the screen brightness on an old Dell. It was because it tried to use dell_backlight instead of intel_backlight 04:27 < domhnall> ascii8bom: atm, using kde on obsd only but no issues. 04:28 < ansraliant> ascii8bom: adding the parameter acpi_backlight=vendor to the kernel boot parameters worked for me. 04:29 < ansraliant> after that it used intel_backlight 04:29 < ascii8bom> ansraliant, i'll give it a shot 04:32 < lnnb> thathpythonguy, if you don't want the process killed by x shutting down, start it outside of x process tree. but it might close anyway if it's toolkit freaks out when the socket closes 04:32 < ascii8bom> looks like i already have that loaded: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash pci=noaer acpi_backlight=vendor" 04:32 < nekoseam> test2 04:33 < lnnb> also read the comment in the startx script lol 04:34 < ansraliant> ascii8bom: and it does not work? check what folders you have here /sys/class/backlight/ 04:35 < ascii8bom> ls /sys/class/backlight/ 04:35 < ascii8bom> dell_backlight intel_backlight 04:36 < ascii8bom> ansraliant, correct, it doesn't work. only way i could find to change the backlight brightness is echo-ing to intel_backlight 04:38 < ascii8bom> i guess at this point i'm trying to figure out where i can configure it to write to intel_backlight when i press the key combination 04:40 < mjsmith> ascii8bom: I've got the location for my macbook 04:41 < ascii8bom> mjsmith, is it in a file somehwere? 04:41 < mjsmith> Ya it's in sysfs. I can send you my script 04:41 < ansraliant> ascii8bom: it certainly looks like the same problem I had with my dell. On my notes I have 3 options for the brightness. acpi_backlight=video, acpi_backlight=vendor and acpi_backlight=native 04:42 < dunnousernamefn> Hey, so I have this program that opens a pseudo-terminal (PTY) thingy, and I'm trying to read/write to it. It's emulating a micro which I programmed to just send back the characters it gets, but invert each bit. 04:42 < ansraliant> ascii8bom: the one that worked for me on opensuse is acpi_backlight=vendor 04:42 < ansraliant> ascii8bom: but on other distros it worked with native 04:42 < dunnousernamefn> cat doesn't seem to show a response when I echo to it, is that my fault or am I doing something wrong? 04:42 < ansraliant> ascii8bom: you might want to try them. 04:42 < ascii8bom> ansraliant, thanks for those options, i'll try them out now and see what happens 04:43 < mjsmith> https://pastebin.com/KwYD3qjh 04:43 < ascii8bom> mjsmith, thanks 04:43 < mjsmith> bindsym XF86MonBrightnessUp exec sudo /usr/local/bin/brightness.sh +100 # increase screen brightness 04:43 < mjsmith> bindsym XF86MonBrightnessDown exec sudo /usr/local/bin/brightness.sh -100 # decrease screen brightness 04:43 < mjsmith> If you're using i3wm or something 04:44 < mjsmith> For the macbook keyboard backlight it's this file and 0-255. BACKLIGHT="/sys/class/leds/smc::kbd_backlight/brightness" 04:45 < Evidlo> is it common for sites to disallow spaces in passwords? 04:46 < dunnousernamefn> Evidlo, are you like, spelling words? 04:46 < dunnousernamefn> I'm not sure, but it might be to prevent that 04:46 < dunnousernamefn> I've never tried it though 04:47 < Evidlo> I'm working on a password generator 04:48 < dunnousernamefn> o i c 04:48 < dunnousernamefn> I mean, you could always just stick to [A-Za-z0-9] and a few symbols 04:48 < Evidlo> and one of the options is "correct horse battery staple" style passwords 04:48 < dunnousernamefn> Ohhh, xkcd, I love it 04:49 < dunnousernamefn> I think someone already make an xkcd generator o_O 04:49 < dunnousernamefn> I doubt they'd exclude spaces 04:49 < dunnousernamefn> Unless they have some sort of whitelist 04:50 < dunnousernamefn> I honestly have no idea, some sites just suck at passwords 04:50 < dunnousernamefn> aha, https://security.stackexchange.com/questions/32691/why-not-allow-spaces-in-a-password 04:53 < Evidlo> think I'll do it anyways. the user can handle special cases 04:54 < bb36e> Survey: Wayland: good or bad? 04:56 < DynV> On windows, with Cygwin installed, is there something I can do to get the kernel version of a *nix ISO? 04:57 < Bashing-om> bb36e: I find wayland faster and smoother than Xorg - but it presently has limited application support . 04:57 < Dan39> DynV: is it a custom ISO or something? just checksum it and google the checksum 05:02 < DynV> it's antiX-17.1_386-full.iso 05:02 < DynV> sorry for the delay 05:03 < Dan39> just google that? -_- 05:03 < DynV> I thought I could just run a command in cygwin 05:03 < DynV> to make sure 05:04 < DynV> or use a text software to look in a file 05:04 < bb36e> Unzip it 05:05 < DynV> in which file is it mentioned? 05:05 < Dan39> uh no, its an iso, not a zip 05:05 < Dan39> -_- 05:06 < DynV> I can use 7zip to look into it 05:06 < Dan39> 7z prolly will, yea 05:07 < Dan39> Customised 4.10.5 kernel 05:07 < Dan39> there, a quick google search 05:07 < DynV> I did a search and couldn't find it 05:07 < DynV> ty 05:07 < Dan39> google antix 17.1? 05:07 < Dan39> ffs 05:08 < DynV> I did a searched https://duckduckgo.com/?q=antiX+17.1+386+full+kernel&t=ffsb&ia=web and https://duckduckgo.com/?q=md5+be122001d80e59977b4ad45075d8444c+kernel&t=ffsb&ia=answer 05:08 < DynV> s/searched/searches/ 05:09 < Dan39> DynV: but of course since it's customized, who the hell knows what it is 05:16 < bls> they should have a source repo with their patch list and .config 05:20 < erg_> hi if i do showmount -e on the client 05:20 < erg_> it dont see the nfs partition 05:20 < erg_> what could be wrong here 05:20 < pingfloyd> nfs is a filesystem 05:21 < pingfloyd> partition type never seems to matter with linux 05:22 < erg_> it says port mapper failure-timed out 05:33 < deww> anyone else got msged by ngravity look for "shells" 05:33 < dgs> yup 05:34 < deww> alright. 05:40 < pingfloyd> no 05:41 < esselfe> it's my bithday today, I'm proud to be a Linux user 05:41 < bls> hapy bithday 05:41 < esselfe> although I know this channel doesn't really celebrate birthdays 05:41 < esselfe> thanks 05:49 < zenix_2k2> not sure if this is the right place to ask but... when i create a file but HAVEN'T written anything to it, does it costs my memory to be fuller ? ( even a little bit ) 05:49 < zenix_2k2> like when i try "touch file.txt" but no texts were written to it 05:49 < esselfe> zenix_2k2: perhaps some cache, but it's minimal 05:50 < esselfe> now the question is... how many of these files? 05:50 < zenix_2k2> how about in range from 1 to 100 ? 05:50 < esselfe> one file is a nail in the ocean 05:51 < esselfe> still a few nails in the ocean 05:51 < zenix_2k2> but in the cache ? not even in the primary storage ? 05:51 < zenix_2k2> not sure if that was the term but i meant permenant storage like HDD or SSD 05:51 < esselfe> yes, but the hardware reference are loaded into memory for faster access 05:51 < zenix_2k2> well ok then 05:51 < zenix_2k2> thk 05:52 < hoang_khoi> paraxial: 05:57 < Tahlwyn> c 05:57 < Tahlwyn> whoops 05:57 < jimm> swhoop? 05:58 < esselfe> Whoopie Goldberg 05:58 < bls> Sheryl Swoopes? 05:59 < jimm> let's introduce her to Yahoo Serious! 06:00 * esselfe is compiling eog since firefox and 'links -g' are the only options to view images here 06:00 < esselfe> rust took a while to compile and my flightgear adventure stopped at a simgear failure 06:01 < esselfe> what is a good book about UTF-8/iso-8859-1 font management? 06:02 < esselfe> I can't seem to find any valuable resources out there 06:02 < esselfe> I have some X11 fonts, but they don't seem to cover most symbols 06:02 < esselfe> (they're limited) 06:02 < bls> doubt you're going to find a book on such an obscure topic 06:03 < bls> and DejaVu is pretty much the best unicode coverage you're going to get 06:03 < esselfe> right, I was even thinking about my own freetype-based font, but it's weird 06:03 < bls> at least for something free/libre 06:04 < esselfe> the best representation I've seen is from firefox reading a page made with &DIGIT; chars 06:04 < esselfe> up to 10000 06:05 < esselfe> (bash scripted) 06:05 < aaro> for X font covering unicode check "efont-unicode" 06:05 < esselfe> alright 06:06 < birdbolt1> hello, how can i ignore redicrection with curl 06:06 < birdbolt1> redirection* 06:07 < bls> pretty sure curl doesn't follow 3XXs unless you tell it to (via -L) 06:08 < defkult> ^ is correct 06:09 < nekoseam> https://github.com/dylanaraps/neofetch/releases/tag/5.0.0 Neofetch 5.0 released 06:11 < birdbolt1> bls, defkult thanks! 06:11 < birdbolt1> is 3XXs the techincal term for redirection or something? 06:12 < bls> no, it's what people use for the general classes when they can't remember the individual codes heh 06:13 < bls> birdbolt1: like the topic headings on https://en.wikipedia.org/wiki/List_of_HTTP_status_codes#3xx_Redirection 06:14 < nekoseam> wew a mac pro 06:14 < nekoseam> those things are beasts 06:26 < nekoseam> Anybody here try the acme text editor? Is it really all its cracked up to be? 06:26 < bls> I used it for a couple of years 06:27 < nekoseam> I've standardized on JOE 06:27 < bls> it had some cool features I haven't seen reproduced to this day, but also had some really awkward stuff 06:29 < bls> closest I've seen to that interface was using org-babel in emacs 06:30 * darkmeson also has a slight preference for joe for some reason 06:30 < nekoseam> its amazingly fast 06:31 < darkmeson> actually, I guess the one reason is that it's not as huge as the vim packages tend to be 06:32 < darkmeson> the other is that I developed an irrational hatred for all things vi early on since I was constantly dumped to it and it told neither what it was, nor how to exit it 06:32 < nekoseam> I don't really bother using any CLI programs I can't figure out by reading the manpages 06:34 < darkmeson> I eventually figured out what it was by trial and error and learned to switch to another console to kill -9 it whenever I got dropped to it, but it wasn't until a couple years later, after the damage was already done, that I learned about $EDITOR 06:34 < darkmeson> nekoseam: that's just the thing. back then, it was hard enough to get the main packages, let alone the manpage sets 06:34 < bls> don't really understand why we stick with EDITOR=VISUAL=vi 06:35 < bls> when the alternative was ed, sure, but that's decades ago 06:35 < darkmeson> the modem maxed out at 14.4K, the connection was spotty at best, and there was a hard four hour disconnect if you somehow miraculously were able to stay connected that long to begin with 06:36 < darkmeson> and of course, nothing supported range requests, so all failed transfers had to start over from 0 06:37 < Dagmar> Because ed is the _default_, and vi (usually vim) is available everywhere 06:37 < Dagmar> ...or would you rather have systems defaulting to using... 06:37 < Dagmar> EMACS 06:38 < Dagmar> That would be a great way to introduce users to Linux... 06:38 < Dagmar> Simply tell them they will spend the next two weeks learning about the text editor. 06:38 < darkmeson> it couldn't have been any worse than early vi :) 06:38 < jimm> it has a builtin tutorial 06:38 < bls> not arguing that. but between having too listen to noob whinge about vi and neckbeards whinge about getting dumped into traingwheelseditor2.0, I choose the latter 06:39 < darkmeson> jimm EARLY vi, 90s 06:39 < Dagmar> I don't whinge. I just freakin' use vim and get on with my day 06:39 < Triffid_Hunter> I like nano, it's pretty powerful when you turn on all the features 06:39 < bls> they're much better at the whinging 06:39 < Dagmar> Vim is the editor used by people who just quietly get sh*t done. 06:39 < Triffid_Hunter> syntax highlighting, smart home, smart indent, block indenting, line numbers, etc etc 06:40 < Dagmar> I use IntelliJ for stuff like that. 06:40 < bls> I heard they even added regexps eventually! 06:40 < darkmeson> we used nano in college, but I forget why I ended up favoring joe. more screen real estate freed for content maybe? 06:41 < Dagmar> the emacs people might have been able to get a foothold, but they're usually spending all their time trying to figure out how to integrate calendaring functionality into Emacs or somesuch 06:41 < jimm> I kinda like this, you're sharing what you use, and no one else is commenting on your choice... that's one way to avoid a war 06:41 < Triffid_Hunter> bfs: oh yeah, regex search/replace too, it's awesome 06:41 < darkmeson> Dagmar: what? no sed? 06:41 < Dagmar> vim has those 06:41 < baiguai> +1 vim 06:42 < Dagmar> Maybe once in a blue moon I'll bother to use more than the basic 3-4 commands you need to get around in vim 06:43 < Triffid_Hunter> heh whenever I find myself in vim I just quietly swear to myself and immediately hit I.. amusingly I don't even have vi or vim installed at the moment 06:43 < bls> that was a bit of a breakthrough moment with vi...realizing the stuff you did with sed/awk/cut/grep could be done in vi, live, on screen 06:43 < Dagmar> I usually hit a instead of i... sometimes C 06:43 < Dagmar> Often o 06:44 < Dagmar> So much slash-searching it's not even funny tho 06:44 < binTrl> Can anyone recommend a file manager which allows adding custom protocols as plugins ? 06:45 < Dagmar> Nautilus, if you feel like writing them 06:45 < k_sze[work]> So I have a folder that is chmodded as 2775 (drwxrwsr-x), and I have a user that belongs to the group. What can cause the user to be unable to even `touch` a file in that folder? 06:45 < revel> Like sftp:// access? 06:45 < Dagmar> If it doesn't already exist, then it probably doesn't follow a very amenable usage mode. 06:45 < k_sze[work]> The user has already re-logged in since being added to the group. 06:45 < k_sze[work]> Other users in the same group are fine. 06:46 < baiguai> I use macros all the time - just wrote a couple, chained em - to convert ~20K records into 1K chunks of SQL union inserts. 06:46 < Dagmar> k_sze[work]: The file itself not being in their group and them not owning it 06:46 < Triffid_Hunter> binTrl: dolphin :P 06:46 < k_sze[work]> Dagmar: no, I mean creating a new empty file using `touch`. 06:46 < Dagmar> k_sze[work]: It's not without reason that new files created in a group-sticky dir take on the group membership by default 06:46 < Dagmar> k_sze[work]: ACLs, SELinux, or you just plain broke the filesystem 06:46 < bls> k_sze[work]: what're the intermediate permissions? 06:47 < Dagmar> ...or... one of your assumptions is not true 06:47 < darkmeson> baiguai: I mostly use sed, grep, and shell scripting for things like that 06:47 < k_sze[work]> bls: intermediate? 06:47 < Dagmar> A small amount of the time the issue is that someone found a way to actually shed their extra group permissions 06:47 < Dagmar> k_sze[work]: They do still have to be able to traverse all the parent directories. Sometimes people forget that 06:47 < baiguai> darkmeson: sweet 06:48 < bls> k_sze[work]: yes, as if /a/b/c/d the user has permissions on d but not a, b, or c 06:48 < k_sze[work]> the user *can* traverse, he's *in* the directory right now. 06:48 < Dagmar> OKay then that's not it. Have them run `id` 06:48 < k_sze[work]> I did 06:48 < k_sze[work]> he's in the group 06:48 < Dagmar> Okay then it's going to be ACLs or SElinux or the filesystem has become unwritable for some reason 06:49 < darkmeson> sed by itself is actually extremely powerful, but sadly, I'm nowhere near master level yet 06:49 < Dagmar> There's not a lot of moving parts to this, but those two are somewhat esoteric and rather problematic In The Wrong Hands 06:49 < k_sze[work]> Dagmar: I have another user in the group who *can* touch-create a new file. So it's not the fs becoming unwritable, I think. 06:49 < bls> doesn't help that it's so terse. I can occasionaly get a working loop in it 06:49 < Dagmar> k_sze[work]: okay, so again... ACLs or SElinux is in the way 06:51 < k_sze[work]> Dagmar: I use `getfacl` to check for ACLs on the folder? 06:51 < bls> also, what's the user's umask (can't recall if you can have 700 and only group write perms) 06:51 < binTrl> Triffid_Hunter: not K* stuff please. Requires a lot of baggage 06:52 < k_sze[work]> bls: how do I check? 06:52 < bls> that'd be a google or read the source code thing 06:52 < bls> looking 06:52 < Triffid_Hunter> binTrl: heh so does gnome and anything else with application-level protocol plugins.. perhaps you want to play with fuse? 06:54 < bls> yeah, you're better off mounting the FSs under the hood and picking a file manager based on appearance/workflow instead of whether or not it can mount dumb-protocol-no-one-else-uses 06:55 < k_sze[work]> bls: `UMASK 022` and `USERGROUPS_ENAB yes` in /etc/login.defs 06:55 < Dagmar> k_sze[work]: yes, and you'd use ls -Z to see any ACLs 06:56 < k_sze[work]> Dagmar: nothing interesting there, all folders preceded by a single question mark 06:56 < bls> right, those are the login defaults. people can change them though (and I'm still not getting any hits on the umask vs perm bits) 06:56 < Dagmar> Okay so no ACls then, which is good because it's easy to rule out 06:57 < binTrl> bls: I intend to use adb to get to my phone fs. Enlighten me. 06:57 < Dagmar> binTrl: `adb shell` doesn't generally require a lot of explanation 06:57 < binTrl> Dagmar: a gui based would be great 06:58 < Dagmar> Note that Android makes _extensive_ use of SELinux so the shell generally doesn't have a lot of access 06:58 < binTrl> One with nice thumbnails, multi select stuff 06:58 < Dagmar> Any file manager capable of MTP 06:58 < revel> Does it have to specifically be over ADB? 06:58 < bls> binTrl: as Triffid_Hunter mentioned, there's a lot of support for FUSE filesystems for weird things like that that your kernel maintainer or file browser author don't want to bother to support 06:58 < Dagmar> ...but again, it's not going to give you very much access to anything but the microSD card 06:58 < binTrl> Dagmar: MTP is freaking slow 06:58 < nothos> binTrl You can mount your android file system using tools to access it the same as any other disk 06:58 < Dagmar> USB isn't exactly super-fast 06:58 < binTrl> revel: Not if there's any better option 06:59 < nothos> Or like Dagmar said, a file manager that does MTP 06:59 < Dagmar> MTP is your only serious option 06:59 < revel> binTrl: There's SFTP support on quite a few file managers. 06:59 < veridiam> solid explorer + sftp is a dream 06:59 < Dagmar> The mass-storage device support never worked entirely correctly (it was actually unsafe) and is no longer available from like Android 5 onward 06:59 < bls> yeah, that was super freaking annoying 07:00 < Dagmar> You could be as lazy as I am about it 07:00 < binTrl> nothos: No you can't 07:00 < Dagmar> Pick the file in the phone's file manager, share to Google Drive 07:00 < Dagmar> ...then it's just magically available on my desktop somewhere a few minutes later. 07:00 < binTrl> Dagmar: We are talking about a few gigs of files 07:00 < Dagmar> Put them on an SDCard and remove the SDcard 07:00 < binTrl> Respect bandwidth 07:00 < revel> Better compress them first then :D 07:00 < bls> well I was around for the cutover, when my workflows were based on it being a mountable file system, mtp sucked, and gdrive wasn't a viable thing 07:01 < Dagmar> I don't really have any speed problems with MTP 07:01 < binTrl> Aaarrghhhh.... ADB *still* sounds better that gdrive, compress, SDcard ... 07:01 < revel> I do. And it can't do "walk and chew bubble gum" at the same time. 07:01 < revel> i.e literally any two different actions at the same time. 07:01 < Dagmar> I can basically guarantee you transplanting an sdcard is going to be faster 07:02 < nothos> binTrl Erm...yes you can? I've done :P 07:02 < iodev> Dagmar: SD card transplant .. no thanks 07:02 < binTrl> nothos: Enlighten me. 07:02 < nothos> binTrl https://wiki.archlinux.org/index.php/Media_Transfer_Protocol#FUSE_filesystems Try installing mtpfs 07:03 < binTrl> actually there's one https://github.com/sole/aafm but the gui is damn rudimentary 07:03 < iodev> nothos: the problem is the file managers are using mounters 07:03 < iodev> instead of direct MTP, SFTP accesss 07:03 < bls> binTrl: the point is to decouple the mounting from the GUI 07:03 < iodev> thorugh plugins 07:04 < veridiam> use a browser with ftp:// ? 07:04 < binTrl> nothos: anything with the word mtp attached to it is probably gonna suck 07:04 < nothos> binTrl True, but being able to do it != user friendly and enjoyable 07:05 < nothos> Basically linux in a nutshell sometimes :D 07:05 < binTrl> veridiam: Better, install ssh server on device and ssh over it. But I need it to work out of the box 07:06 < veridiam> well all that you get 'out of the box' is MTP or ADB if enabled 07:07 < binTrl> That's why I'm looking for a explorer with customizabe protocols (where I'd use ADB) 07:11 < binTrl> FUSE looks pretty neat. Gotta look at it 07:13 < darkmeson> oh, it is. sshfs in particular is a life saver 07:13 < Mo> Hi, on grep, is it possible to preserve the colors of the primary command? 07:14 < bls> if you went for FUSE, you could mount all kinds of stuff and you wouldn't have to worry about switching file managers for a single device and the authors of your preferred file manager wouldn't have to worry about them writing that code for themselves 07:14 < darkmeson> actually, just about anything fuse-based is better than native kernel modules if you can afford the slower access because they can be unwedged just by killing the backing daemon rather than having to reboot 07:15 < bls> Mo: are you piping multiple greps? do you really need to? 07:15 < bls> or is this want the grep output colors to show up in less 07:16 < Mo> bls: I have a command returning commands that I like to grep, whithout loosing colors. I got it working by disabling greps own coloring like --color=never. 07:17 < bls> ok, then you need to figure out how to tell the command in front of grep to not drop its colors 07:18 < bls> because individual commands can/do decide whether or not to output color based on whether or not their output is to a screen or to a pipe 07:18 < lnnb> Mo: it entered commands or you pressed return? 07:18 < Mo> I got it working by --color=never 07:19 < Mo> So that disables grep colors and instead does not strip the incoming colors in front of the pipe. 07:23 < lnnb> Mo: i see what you mean now 07:25 < lnnb> color=always ? 07:25 < lnnb> instead of never 07:26 < bls> color=always means grep outputs codes to clear all color, then apply its own color to matches 07:26 < bls> which gets rid of whatever color was set in its input 07:27 < lnnb> bls: it doesn't clear the first grep match if you put always on the second 07:27 < Mo> lnnb: always does the opposite and does not work. 07:27 < Mo> like bls said.. 07:27 < lnnb> works here 07:28 < lnnb> LC=C 07:28 < elf> deww: yes i got a msg from him too 07:28 < Mo> lnnb: 'ls -al ~' does that output colors for you? 07:29 < lnnb> not unless i use --color option 07:29 < lnnb> or am in a shell with it aliased 07:30 < Mo> lnnb: 'ls -al --color=always ~ |grep --color=always .' Does that still output the same colors for you? 07:30 < Mo> For me not, only for color=never on grep. 07:30 < lnnb> lol it changes the colors 07:31 < Mo> lnnb: Exactly, like bls said. 07:31 < lnnb> if i use color=never with grep the colors are still there from ls 07:31 < Mo> You got it. 07:31 < lnnb> why would it change the colors 07:32 < Mo> Like bls said. 07:32 < lnnb> so what happens if you don't terminate the color 07:33 < lnnb> i thought he meant strip the color when he said clear all codes 07:34 < bls> no, I meant that grep want to make sure you can tell what it matched, so it drops the color from its input, than adds its own color back on top 07:35 * Mo likes to get some grep --color=merged 07:44 < smallville7123> 15:41 smallville7123: How do i mount an iso 07:44 < smallville7123> 15:42 smallville7123: I keep getting /dev/loop0: permission denied 07:44 < smallville7123> 15:43 smallville7123: https://i.imgur.com/fHyZq8o.jpg 07:46 < binTrl> smallville7123: I see you are using codeanywhere. 07:47 < binTrl> Maybe you are not allowed to mess up with loop devices 07:48 < smallville7123> How tf would that makesence when i hv sudo rights 07:59 < RayTracer> smallville7123: try mount -o ro,loop 07:59 < Psi-Jack> smallville7123: Tip : There is a language policy here. One, try not to use vulgarity, even shorthanded vulgarity. Second, speaking of shorthanding, "hv" is not the same as "have", we prefer English here. 08:01 < smallville7123> https://i.imgur.com/M23qrzS.jpg 08:04 < sauvin> smallville7123, you can't directly mount an ISO that way if it's an ISO 9660 image (most data CDs and DVDs). 08:04 < smallville7123> lsmod shows no modules loaded 08:05 < sauvin> Try using -t iso9660, I think it is. 08:05 < Psi-Jack> That's an unusual screenshot. 08:06 < sauvin> Looks like it's on a phone. 08:06 < Psi-Jack> Yeah. 08:06 < Psi-Jack> With no sim card. heh 08:06 < ansraliant> a phone with arch 08:07 < sauvin> And still probably more capable than Agent Smart's shoe phone. 08:07 < Psi-Jack> Well, more like a phone connected to ssh somewhere with the connected wifi. 08:07 < ansraliant> https://www.reactiongifs.us/wp-content/uploads/2013/12/today_I_saw_weird_shit_billy_madison.gif 08:08 < ansraliant> haha 08:08 < smallville7123> i keep geep getting /dev/loop0: operation not permitted 08:08 < smallville7123> keep* 08:08 < ansraliant> I should not laugh at my own jokes... damn. 08:08 < Psi-Jack> Repeating that without information is not useful. 08:09 < smallville7123> What more do you need, you have a screenshot 08:09 < Psi-Jack> ... 08:10 < sauvin> smallville7123, mind you, I'm assuming "real computer" and not phone, tablet or things running goofy libc: try mount -t iso9660 isofile.iso mountpoint 08:10 < smallville7123> It a ssh container 08:10 < smallville7123> opemvz 08:10 < smallville7123> openvz* 08:11 < smallville7123> www.codeanywhere.org 08:12 < smallville7123> https://www.codeanywhere.org 08:12 < DynV> Dan39: it's a customized 4.9.87 08:13 < sauvin> "codeanywhere.org" won't resolve for me. Hrm. 08:13 < smallville7123> https://codeanywhere.com/ 08:15 < ansraliant> is this a web IDE? 08:16 < smallville7123> i guess 08:16 < jim> I did get it to resolve 08:16 < sauvin> "you guess"? 08:16 < Psi-Jack> And what does that have to do with mounting an ISO file? 08:16 < sauvin> smallville7123, why are you trying to mount an arch iso in this environment? 08:17 < smallville7123> Because it doesnt have a arch container and e kernel is too old to run a modern arch chroot 08:18 < smallville7123> the* 08:18 < smallville7123> 2.6.32 08:19 < Psi-Jack> ... 08:19 < Psi-Jack> Ooookkaay.. It's making even less sense now. :p 08:20 < smallville7123> Rip 08:27 < xormor> there was a command to look at the "ELF" magic etc. on an executable file. What command is that? 08:29 < birdbolt1> just double checking 08:29 < birdbolt1> rm -rf *.git 08:29 < birdbolt1> removes every folder all the way down 08:30 < birdbolt1> which ends in .git right> 08:30 < birdbolt1> well every folder and file 08:30 < birdbolt1> right? 08:30 < iflema> yes and no 08:30 < iflema> y the star? 08:31 < birdbolt1> .git files existed in multiple directories below the current path 08:31 < darkmeson> xormor: something other than 'file'? 08:31 < birdbolt1> .git folders too 08:32 < birdbolt1> I wanted them all gone and hoped that command should do 08:33 < birdbolt1> My buddy did git add * on his mac 08:33 < birdbolt1> and it added parent directories as well 08:33 < birdbolt1> whats up with that 08:33 < xormor> darkmeson, I think it was something other than that. 08:34 < darkmeson> birdbolt1: 'find . -type d -name .git' might be what you want 08:35 < birdbolt1> but i wanna delete not simply find them darkmeson 08:35 < darkmeson> run it that way to verify what would be acted on, then tack on '-exec rm -rf "{}" ;' to have it delete those directories 08:36 < darkmeson> ugh. that ; should be a \; 08:36 * darkmeson forgot that his client requires \'s to be escaped 08:38 < darkmeson> xormor: there's ldd to show linked libraries, or maybe something from elfutils? 08:39 < dnanib> find . -type d -name .git -print0 | xargs -0r rm -rf 08:39 < darkmeson> oh, wait, readelf 08:39 < ansraliant> what do you guys think of spacemacs? is it some kind of mutant? 08:39 < ansraliant> maybe you heard things 08:40 < ansraliant> I knew about an emacs die hard fan, he told me about it. I myself am a vim user. I know, a vim user friends with an emacs user? well.. it's the new millenium. 08:41 < ansraliant> Times change 08:42 * kurahaupo doesn't mention in polite company that he uses both vim and emacs keybindings 08:43 < Psi-Jack> OKay, something serious is going on here. I've got chrome open with a gazillions tabs, Steam is open, rambox (electron app) is open, evolution, hexchat, pidgin, synergy... And I'm only using 3.6GB RAM. 08:43 < Psi-Jack> That.... Is impressive. 08:44 < ansraliant> 3.6GB RAM left 08:44 < auronandace> Psi-Jack: if you have hardware acceleration on wouldn't some of those use your graphics ram? 08:45 < Psi-Jack> auronandace: What? No... 08:45 < Cleron> hello 08:45 < Psi-Jack> ansraliant: No, 3.6GB /used/ 08:46 < Psi-Jack> I used Arch before Solus, openSUSE and Fedora before that, and just this would've easily been consuming 8-16GB RAM. 08:46 < ansraliant> what are you using now? 08:46 < Psi-Jack> Solus. 08:46 < ansraliant> interesting 08:47 < Cleron> I realize its not recommended to put a gui on a linux server. Is it possible on a server to install Cinnamon or Gnome and start it up when i need to make some changes and shut it down after but let the linux server to keep running? 08:47 < auronandace> Psi-Jack: does your ram measurement include cacheing 08:48 < Psi-Jack> auronandace: Well, considering caches, that's 12GB of cache. 08:48 < darkmeson> compressed ram or swap or something? 08:48 < xormor> darkmeson, "readelf" 08:48 < Psi-Jack> Nope. 08:49 < darkmeson> page deduplication? 08:49 < Psi-Jack> No. 08:49 < hendrix> I'm on xfce, few apps and 6 pages open, 5,44GB used. 08:49 < Psi-Jack> hendrix: Exactly! 08:49 < Psi-Jack> When I say that before I would be using 8~16GB that was when apps used memory, not caches. 08:50 < hendrix> I long for days when 8MB ram was enough 08:50 < Psi-Jack> And now, only /using/ 3.6GB for applications, and the rest for caches.. 08:51 < darkmeson> that IS just RES right? 08:51 < darkmeson> oh wait, nevermind. I was thinking the usual linuxatemyram case 08:52 < darkmeson> *along the lines of 08:52 < darkmeson> also already covered, albeit a different way 08:53 < Psi-Jack> Heh 08:53 < Psi-Jack> Yeah, I'm just shocked. 08:54 < Psi-Jack> I don't think I've ever seen this kind of memory utilization in Linux ever before now. Now I want to know why. LOL 08:54 < birdbolt1> https://downloadmoreram.com/ 08:54 < luke-jr> browsers 08:55 < mavorsa> its illuminati 08:55 < luke-jr> OS: GNU/Linux 4.17.0-gentoo/ppc64le - CPU: 64x IBM POWER9 - Processes: 746 - Uptime: 2d 2h 28m - Users: 15 - Load Average: 61.07 - Memory Usage: 5203MB/65218MB (7%) 08:55 < luke-jr> ^ no browsers 08:55 < Psi-Jack> Someone hasn't been paying attention. :) 08:56 < luke-jr> dropped to 4.6 GB by killing krunner O.o 08:56 < Psi-Jack> luke-jr: 3.6GB /used/ (12GB in caches), Chrome open with ~30~40 tabs. Rambox (electron app, memory hungry), hexchat, evolution, steam, etc) 08:56 < luke-jr> Psi-Jack: how? 08:56 < Psi-Jack> Exactly! 08:56 < luke-jr> X seems to be 1.3 GB 08:56 < mavorsa> the power of jesus christ? :P 08:57 < ansraliant> how much memory do you have luke-jr? 65GB? 08:57 < luke-jr> ansraliant: 64 GB 08:57 < birdbolt1> i will one day be like you 08:57 < birdbolt1> as soon as msft starts making 64 GB ram surface books 08:57 < ansraliant> hahah 08:57 < luke-jr> I'm only using 2 of my 16 RAM slots for that 08:57 < luke-jr> this system supports up to 2 TB RAM :p 08:57 < birdbolt1> holy shit 32GB ram sticks exist? 08:57 < luke-jr> yes.. 08:58 < luke-jr> as do 128 GB RAM sticks 08:58 < birdbolt1> oh wow 08:58 < mavorsa> is this a server we're talking about? 08:58 < luke-jr> no 08:58 < ansraliant> no? 08:58 < Psi-Jack> No, this is a desktop. 08:58 < luke-jr> new desktop PC 08:58 < birdbolt1> are those ram sticks suitable for gaming? 08:58 < luke-jr> dunno 08:58 < birdbolt1> i know the speed fo the ram is usuuuualy negligible but 08:58 < luke-jr> I play Jedi Academy occasionally 08:58 < birdbolt1> nah man. I roll with AAA games 08:58 < luke-jr> kinda slow, but I'm pretty sure that's because I use CPU rendering 08:58 < birdbolt1> gpu intensive shooters 08:59 < Disconsented> 32GB UDIMM's arnt a thing (yet) but there are 32GB RDIMM's 08:59 < mavorsa> sounds like you could play solitaire on thing or something.... :P 08:59 < birdbolt1> lol 08:59 < morf> i think they have 256GB ram modules already btw 08:59 < mavorsa> but i could be overstepping that claim 08:59 < birdbolt1> you mean minesweeper 08:59 < birdbolt1> solitaire would run out of mem 08:59 < darkmeson> Cleron: yes, you can, but ssh forwarding to the local X11 server would probably be easier for very occasional use 08:59 < Psi-Jack> Huh... My work desktop is about the same. 2.3GB, Chrome open with about 20 tabs, rambox, hexchat, pidgin, keepassxc, evolution, etc.. 08:59 < luke-jr> mavorsa: with 64 threads, CPU rendering is playable for Jedi Academy :p 08:59 < hendrix> my computer supports whopping 8GB RAM! 08:59 < birdbolt1> what cpu is that? 08:59 < luke-jr> birdbolt1: dual 8-core POWER9 CPUs 09:00 < birdbolt1> holy shit what do you do with that 09:00 < birdbolt1> ive never seena dual cpu mobo irl 09:00 < luke-jr> not much 09:00 < smallville7123> How do i update elf.h 09:00 < ansraliant> with a load like that, it must me something heavy 09:00 < ansraliant> load average 61 09:00 < luke-jr> birdbolt1: initially, they only made dual socket motherboards for these 09:00 < Psi-Jack> smallville7123: You make little sence, still. 09:00 < luke-jr> birdbolt1: and if you only install 1 CPU, you only get 2 PCI-e slots :/ 09:00 < Cleron> just installed Cinnamon on debien when i reboot it goes back to commandline login. how to i start cinnamon? 09:00 < birdbolt1> I had a TR processor for a bit 32 threads *sob*. Had to sell the parts for a surface book cuz the thing was too big to carry to and from school very sememster 09:01 < smallville7123> My elf.h doesnt have the required defines i need 09:01 < luke-jr> ansraliant: well, I keep it busy with some idle tasks 09:01 < lnnb> luke-jr: if only the renderer was optimized for the cpu 09:01 < Psi-Jack> Man.. I /really/ want to know what the magic Solus is doing to accomplish this memory utilization! 09:01 < ansraliant> is there a solus channel? we could ask there 09:01 < Psi-Jack> Sure, there is. But most people are sleeping. :) 09:01 < ansraliant> kick the door and ask "what are you doing with the memory management?" all in caps 09:01 < Psi-Jack> And me, well, I'm about to sleep. 09:02 < luke-jr> what is Solus? 09:02 < Psi-Jack> luke-jr: A Linux distribution designed for desktop use. 09:02 < luke-jr> does it support POWER? 09:02 < Psi-Jack> It has the POWER. 09:02 < ansraliant> lel 09:02 < Psi-Jack> ... of greyskull? 09:02 < mavorsa> can it run a fully operational battlestation? 09:02 < darkmeson> Cleron: startx as your user, or systemctl start graphical.target, or service gdm start, etc 09:03 < Psi-Jack> mavorsa: Well, I have it connecting, via ssh, to my battlestation in my home office. :) 09:03 < Cleron> so startx will not work as root? 09:04 < Psi-Jack> Cleron: You shouldn't run X as root. 09:04 < luke-jr> startx as root will ideally exec halt 09:04 < Psi-Jack> You shouldn't run X on a server either. 09:04 < morf> nah just do whatever you want 09:04 < lnnb> but don't talk to strangers 09:05 < Cleron> can i shut down x any time so that it does not use up resources on a server, then just start it up when i need it? 09:05 < bartmon> mavorsa, i feel like i should know that reference... where's it from? 09:05 < luke-jr> Cleron: you may be interested in Xpra 09:05 < luke-jr> bartmon: Star Wars.. 09:05 < mavorsa> bartmon: star wars, emperor reference 09:05 < Psi-Jack> Anyway.. Sleepy time. 09:05 < bartmon> thank. 09:06 < luke-jr> anyone know if there's any modern standards for Ethernet that are bus instead of star? 09:06 < bartmon> s/\./s./ 09:06 < Psi-Jack> Cleron: ssh into the server when you need to do something. There's a whole list of reasons why running xorg on a server is bad. 09:07 < birdbolt1> yo isnt xorg like an input controller or some shit? 09:07 < nothos> Now'd be a bad time to mention xrdp then? 09:07 < ansraliant> luke-jr: bus ethernet. Now that's something I haven't heard in a long time 09:08 < luke-jr> ansraliant: thinking about putting a SoC at every outlet when I build a house, but it'd be a pain to have a switch port for each one.. 09:09 < luke-jr> not to mention all the Cat7 that'd entail 09:10 < ansraliant> so you want them to share the cable 09:12 < luke-jr> ansraliant: ideally, yes 09:12 < Cleron> When i do StartX i get -bash: startx: command not found. When i type "systemctl start graphical.target" it ask for password Authentication is required to start 'graphical.target'. than it sais ==== AUTHENTICATION COMPLETE === 09:12 < Cleron> . But does not start cinnamon. When i do "service gdm start" it sais -bash: service: command not found 09:14 < ansraliant> luke-jr: it should be possible. I don't remember much of the ethernet standards. Time to refresh my memory. I have somewhere my computer networks by Tannenbaum 09:14 < ansraliant> haha 09:19 < ansraliant> luke-jr: I think 802.3 supports bus topology out of the box. There is no need to do anything 09:20 < ansraliant> luke-jr: 802.3 defines CSMA/CD so.. in theory it should work 09:20 < luke-jr> ansraliant: but what on the physical layer? 09:21 < luke-jr> I mean, STP has two well-defined ends, right? 09:26 < ansraliant> yes, but wait. There was something you attached to the end of the backbone to solve that problem 09:27 < Triffid_Hunter> binTrl: fwiw I've mounted my phone using sshfs (a FUSE module) before.. way better than mtp which seems to crash all the damn time 09:27 < ansraliant> luke-jr: https://www.coursehero.com/file/p7fa9n/A-terminator-is-required-at-each-end-of-the-bus-cable-to-prevent-the-signal/ 09:28 < ansraliant> luke-jr: I'm not crazy... 09:28 < ansraliant> I found it by searching thing attached end cable 09:28 < ansraliant> haha 09:29 < ansraliant> luke-jr: so, you have your backbone, with the thing attached at the end, the beginning in your main pc or whatever. The SoC connected to the bus throught the sockets. Everything else should work out normally 09:29 < ansraliant> but you will have half-duplex 09:30 < ansraliant> But I don't think that will bother you 09:31 < luke-jr> ansraliant: sure this isn't talking about obsolete 10Base2 stuff? 09:32 < ansraliant> I think it was until 100BaseT... let me check 09:39 < peetaur2> is there a way to ask sshd which algo it is currently using for an existing connection? 09:39 < ansraliant> luke-jr: well.. it seems it was until 10Base5 09:39 < peetaur2> (kex, mac, cipher, host key) 09:39 < luke-jr> so ancient/obsolete :/ 09:39 < ansraliant> yeah, were back at square one 09:40 < ansraliant> guess you'll have to buy a big switch and a lot of UTP cable... 09:41 < momomo> I have a printer that I've successfully installed using WPS and without cables ... however, the Scanner functionality seem to work, at least with my program on Ubunut (Mint). Is there a program or an easy way to get the scanner functionality to work as well without cables ? 09:41 * luke-jr wonders how practical it is to do Ethernet over CAN 09:42 < ansraliant> luke-jr: or, you can jump into the wireless wagon 09:42 < luke-jr> pretty sure that many clients will overload a router 09:42 < ansraliant> depends on the router 09:42 < ansraliant> by the way, how many clients do you think you will have? 09:43 < peetaur2> if you can't scale horizontally, it doesn't matter which router...it still sucks. 09:43 < luke-jr> dunno, my current house has probably at least 30 outlets 09:44 < peetaur2> I think connections per second, packets per second, bandwidth, and whether you use traffic shaping, qos, etc. is more relevant than the number of ports 09:44 < ansraliant> 30 clients is nothing 09:57 < mavorsa> i wish there was a better xfce theme manager that actually changed all the desktop elements instead of just a few, so i can actually have a day theme and a night theme... #firstworldproblems 10:03 < searedvandal> what doesn't it change? 10:03 < searedvandal> the window manager stuff? 10:03 < well_laid_lawn> you could probably script it with calls to xfconfig mavorsa 10:03 < searedvandal> yeah 10:03 < lopid> hold down the brightness button on your monitor for the night theme 10:03 < searedvandal> haha 10:04 < mavorsa> well, for example, i have xfce-theme-manager, but the wallpaper section doesn't want to do anything when changing themes 10:04 < mavorsa> so ill change themes and the panels will change but the background wont 10:05 < searedvandal> xfce-theme-manager -t theme -b wallpaper.png 10:05 < mavorsa> somehow i also mucked up my gtkrc, as i have to manually define it in my home directory for the panels. 10:06 < mavorsa> searedvandal: hm, maybe i can tie that to an alias or something 10:06 < searedvandal> yeah 10:06 < searedvandal> take a look at the man page for all the cli options 10:07 < searedvandal> I believe you have to pass the -n option also in order to not start the gui 10:08 < searedvandal> if I remember correctly. currently on mate, so haven't used it in a while 10:11 < kevr> any clue why xrandr output positioning would not work in ~/.xinitrc? 10:13 < kurahaupo> kevr: first guess, .xinitrc isn't being read 10:14 < kurahaupo> Throw in a « date +'%F,%T xinitrc check' >> ~/.xinitrc 10:15 < kurahaupo> Gah 10:15 < kurahaupo> I meant throw in a « date +'%F,%T xinitrc check' >> ~/.xinitrc.log 10:15 < kurahaupo> » 10:15 < mavorsa> i'm not sure what im doing with the xfce-theme-manager cli. do is specify the theme by just the name or the path? and if the path, is it supposed to point to a file that xfce-theme-manager compiled from the theme components? and where is that? 10:21 < storge> if only there was some sort of manual 10:21 < mavorsa> man page doesn't mention it 10:22 < mavorsa> okay well meta theme im guessing means just the name, but that doesn't work 10:22 < mavorsa> (which is why i questioned it) 10:23 < searedvandal> another way is xfconf-query 10:24 < searedvandal> believe for themes it's xfconf-query -c xsettings -p /Net/ThemeName -s "Themename" 10:25 < searedvandal> can do a simliar line for wallpaper was well 10:26 < mavorsa> right, so this theme manager's cli doesn't use custom compiled references, which means ill have to define gtk, xfwm, icons, etc all manually in the same line 10:30 < mavorsa> right, so all of that worked, except for the background still (the xfce-manager cli). its like the xfce desktop settings seem to override it 10:31 < searedvandal> try to do wallpaper with xfconf-query 10:33 < searedvandal> xfconf-query -c xfce4-desktop -m to see what the location of backdrop entry is, and then xfconf-query -c xfce4-desktop -p /path/to/backdrop -s /path/to/wallpaper.png 10:34 < mavorsa> is there a way to specify both the gtk and the window manager theme separately (as i combine two different themes between these)? 10:38 < searedvandal> xfconf-query -c xettings -p /Net/ThemeName -s "Theme" and xfconf-query -c xfwm4 -p /general/theme -s "Theme" should set different themes 10:39 < searedvandal> the last one is for the window manager 10:39 < mavorsa> also the monitor just shows "/backdrop/screen0/monitor0/image-path" when i change themes 10:40 < mavorsa> i assume that's what is in a file somewhere but no idea what file if so 10:41 < searedvandal> for wallpaper? do xfconf-query -c xfce4-desktop -p /backdrop/screen0/monitor0/image-path -s /path/to/wallpaper.png 10:44 < mavorsa> the /Net/ThemeName, do i replace that with the theme name? and what if it has a space in it? 10:44 < mavorsa> or is that just the path to the theme? 10:44 < searedvandal> that's the property name of the theme 10:45 < searedvandal> so you should write it like that, not replace it with theme name 10:45 < mavorsa> okay, well its saying the property "/Net/ThemeName" does not exist on channel "xsettings", should i create it or something? 10:46 < searedvandal> do xfconf-query -c xsettings -m 10:47 < mavorsa> well, it is showing that its setting /Net/ThemeName 11:02 < mavorsa> i'm just wondering why the desktop settings are overriding the theme manager. 11:07 < JamesZhu> hi there 11:07 < JamesZhu> is there a channel for storage tech? 11:10 < searedvandal> think there is a ##storage channel, but don't know how active it is 11:10 < GNU\colossus> JamesZhu, try /msg alis list *storage* 11:10 < GNU\colossus> (and other terms that seem relevant to you) 11:11 < GNU\colossus> people concerned with ceph, glusterfs, lvm et al. might be interested in talking about that topic 11:11 < mavorsa> in my settings editor, it is showing that the wallpaper file is changing for it. 11:13 < mavorsa> brb, gonna restart session and see if wallpaper changes 11:13 < JamesZhu> I could not join "##storage" 11:13 < JamesZhu> nvm, i'm in that channel, but only 7 people 11:14 < JamesZhu> :) 11:20 < mzaza> I have disk space filled up, but I have no idea what is using it. https://paste.ubuntu.com/p/FPhdCXB2VX/ 11:21 < mavorsa> so, apparently, xfdesktop has full control over the wallpaper. im not sure how i can override it 11:21 < Night_Elf> mzaza: du -h --max-depth=1 ./ 11:21 < Night_Elf> That will start from the current directory and will tell you how much space each other directory there is occupying. 11:22 < tneva82> Hello. More configuration questions. /etc/httpd/conf.d/django.conf file I tried adding listen 9500 and lines and set up as before using previous instances(that hosting company had set up). Port number I just pulled out of my head setting for number between previous 2. It complained however that no listening sockets available. Guess I would need to open port for listening? 11:22 < tneva82> Where on server I could do that or is that in firewall which basically means I need to request hosting company to do it for us? 11:22 < Night_Elf> You can also place some other directory instead of ./ such as for example /var/log 11:25 < TyrfingMjolnir> Which files will "find . -atime +90" show me? 11:25 < geirha> files with access time older than 90 days 11:26 < djph> anything starting in the current directory that was last accessed ... 30 months ago 11:26 < djph> *3 months ago 11:28 < TyrfingMjolnir> This is on ext4 11:28 < TyrfingMjolnir> Which other options are there than ctime and atime? 11:28 < nekoexmachina> $man find ? 11:28 < mzaza> Night_Elf: Same info as the ones pasted above, Still I can't figuire what's taking all of that. 11:29 < nekoexmachina> man find|grep time works too 11:29 < TyrfingMjolnir> There appears to be no creation time 11:29 < Night_Elf> mzaza: that's odd 11:31 < mzaza> Night_Elf: Yes, and that's not the first time it happens. 11:31 < Night_Elf> mzaza: Maybe there's some mountpoint in your system that is full and when you mount, du shows the space of the mounted filesystem and not what the mountpoint itself contains ? 11:32 < nekoexmachina> TyrfingMjolnir: you can use -newer options to get that 11:32 < nekoexmachina> find -type f -newerct date_time ! -newerct date_time+1_day will find files which perms have changed from date_time to date_time+1 11:33 < nekoexmachina> if you didnt' change perms by hand it is creation date as well 11:33 < mzaza> Night_Elf: Can I check on the partition itself, it's /dev/xvda1? 11:40 < Night_Elf> mzaza: are there any mountpoints you use in the filesystem ? 11:43 < mzaza> Night_Elf: I am on AWS instance by the way, if that gives you any clue to anything :D 11:43 < mzaza> Night_Elf: no 11:46 < Night_Elf> Strange 11:48 < Night_Elf> mzaza: if it happened before, what was the cause of it ? 11:49 < jason85> How do I run a gtk application without a desktop environment? 11:50 < pingfloyd> like you do any other time 11:50 < likcoras> jason85: install X, run application 11:50 < pingfloyd> whether you have a DE installed or not doesn't matter for that 11:50 < revel> You don't need a desktop environment for GTK, do you? Just X. 11:50 < jason85> Thanks, didn't know it was that easy :) 11:51 < Night_Elf> Or connect remotely via ssh and run it redirected, maybe ? 11:51 < pingfloyd> the application you're trying to run just need its dependencies installed 11:52 < pingfloyd> Night_Elf: DE is basically a WM, bunch of extra daemons, and a bunch of apps bundled (many gtk+ based). 11:52 < pingfloyd> jason85: ^ 11:53 < pingfloyd> when you frankenstein a GUI together, you're basically constructing a custom DE 11:54 < pingfloyd> linux is for mad scientists! 11:54 < mzaza> Night_Elf: I don't know :D but a reboot solved the problem 11:54 < pingfloyd> mzaza: sometimes it's the shortest fix 11:55 < mzaza> pingfloyd: Yes :D, but it's a production server and I want to avoid this problem in the future. It brings the whole server down. 11:56 < pingfloyd> of course 11:56 < pingfloyd> mzaza: if a reboot fixed it though, it suggests the system was in an iffy state 11:57 < pingfloyd> one thing to do is note the date and time when the problem started. Then you can look in all the logs around that time for any clues. 12:00 < pingfloyd> that might help give an idea what proc(es) were having issues. If it happens again, you can take a closer look at those next time, as well as an idea of which one you may need to restart instead of resorting to a reboot. 12:00 < pingfloyd> s/one/one(s)/ 12:04 < rosco> Do you know a good way to keep a clean environment for building rpms? Maybe a container system? 12:05 < MrElendig> there are various tools to automagically build in a container/chroot 12:05 < MrElendig> that is how the distro packages are made 12:06 < MrElendig> see the devtools for your target distro 12:09 < pingfloyd> rosco: usually your dist will have an intended method 12:12 < Blinky_> morning all. I am rsyncing a loads of directories and files to a second server but need to preserve all the dates on the files and folders. I am using "-arhz" and that is preserving the files date but all the folders are being set to todays date. Is there any way of preserving the folder dates as 12:12 < Blinky_> well? 12:17 < ne2k> Blinky_, -a implies -r, so you don't need -r as well. -a implies -t and so should preserve the modification times. you can exclude dir from times with -O, you're not accidentally setting -O, are you? 12:18 < ananke> Blinky_: -z often results in a lousy performance. unless you really need to save on network traffic, I would skip that 12:19 < ne2k> Blinky_, are you seeing this behaviour during the run, or is still like that when it finishes? does the filesystem you're writing to support changing the mtime of a directory? (try doing it manually with touch -mt and see if it works) 12:20 < ne2k> Blinky_, try switching on verbose output and see what you get 12:23 < pingfloyd> -a also covers preserving mtimes 12:24 < ne2k> pingfloyd, I said that 12:24 < ne2k> [11:17] Blinky_, ... -a implies -t and so should preserve the modification times. 12:26 < mawk> hi 12:26 < mawk> you solved the pulseaudio issue ne2k ? 12:26 < ne2k> mawk, yep, just needed to disable PrivateNetwork for rtkit using a dropin 12:26 < ne2k> I think it's an ubuntu packaging bug, really 12:27 < ne2k> forums suggest lxc are denying responsibility 12:27 < mawk> rtkit inside the container ? 12:27 < ne2k> yes. as far as I can tell, the real issue is that rtkit tries to use a private network namespace for additional security, but within a container, it can't do that because you can't nest namespaces. or you can't the way I have lxc set up anyway 12:28 < mawk> yeah 12:28 < mawk> well you can nest namespaces if you have enough privileges in the enclosing namespace 12:28 < mawk> which you don't have with LXC 12:28 < mawk> but it's doable 12:29 < ne2k> this approach appears to be much more straightforward 12:29 < mawk> you'd need to disable apparmor and you already have enough namespace capabilities, and for a full blown container the enclosing container will need to be privileged 12:29 < mawk> yeah 12:29 < mawk> indeed 12:29 < mawk> and more secure 12:29 < Blinky_> Sorry guys, I have just tried using -t and it is still setting the directory datetime to today 12:30 < Blinky_> I have checked the source server and the dates on the folders are all different, just when it gets to the new server they are all set to today. I am not using -O either 12:30 < pingfloyd> what's FS is on the destination? 12:30 < veridiam> alias |grep rsync 12:31 < Kingsy> anyone had experience with getting pulse audio to work over a network? 12:31 < ne2k> Blinky_, can pastebin the output of t=$(mktemp -d); mkdir $t; stat $t; touch -mt 0101010101 $t; stat $t; rmdir $t; on the target machine? 12:31 < ne2k> by target I mean the machine where you are writing the files to 12:31 < ne2k> veridiam++ 12:32 < Kingsy> paplay -s 10.8.0.10:4713 a2002011001-e02.wav is how I am trying to connect to pulse audio and I am getting Connection failure: Access denied back, however I can telnet into that host on that port. 12:32 < Blinky_> ne2k, What does that do? 12:32 < Kingsy> any ideas welcome :) 12:32 < ne2k> Blinky_, does the test I asked you to do and you didn't do ;-) 12:32 < ne2k> Blinky_, caveat paster 12:33 < CyberCr33p> Does any SSH client interface exist like this: https://image.ibb.co/ceFHkJ/Screen_Shot_2018_06_19_at_13_31_19.png ? To support multi-tabs and also allow to type 1 command and send it to all tabs? 12:35 < djph> CyberCr33p: not that I'm aware of. 12:35 < Blinky_> ne2k, https://paste.linux.community/view/486b48e5 12:36 < ne2k> Blinky_, hm. I should have said to do it in the location you're trying to write to 12:36 < Blinky_> I did it in the destination directory I am writting too. 12:38 < ne2k> Blinky_, t=$(mktemp -d -p /path/we/are/writing/to); mkdir $t; stat $t; touch -mt 0101010101 $t; stat $t; rmdir $t 12:38 < ne2k> Blinky_, no, mktemp makes a file in /tmp. I made a booboo. try that instead 12:39 < ne2k> *some assembly required* 12:42 < Blinky_> ne2k, https://paste.linux.community/view/6de1e69c 12:53 < eset> I have two entries in /etc/sudoers.d/potsgres -> dev ALL=(postgres) NOPASSWD: /bin/bash and under neath dev ALL=NOPASSWD: /bin/systemctl restart postgresql.service and I get on dev user Sorry, user dev is not allowed to execute '/bin/systemctl restart postgresql' as root on localhost. 12:56 < Sitri> You're missing the .service 12:58 < eset> Sitri: yeah, facepalm :/ sorry 13:18 < TheDcoder> Is anyone familiar with perf report? 13:19 < TheDcoder> I am trying to debug my desktop environment and a functions seems to be on top of the CPU usage 13:19 < TheDcoder> https://paste.fedoraproject.org/paste/nGf2o5YkVYqHVH98A21qOA 13:20 < Lope> Gnome 3 sucks. What's a decent DE I can install on Ubuntu 18.04? 13:20 < TheDcoder> But I am not sure if I am correct because I never used perf before 13:20 < epicmetal> Lope: it depends on your needs and tastes 13:20 < TheDcoder> Lope: Try the other *buntu distros and choose your favourite 13:20 < epicmetal> Lope: the other major DE is Plasma 13:21 < Lope> Gnome 3 has almost no options nor functionality. I want a power user type of DE. I'd also like taskbars with window lists for each monitor 13:22 < Lope> do you mean kde-plasma-desktop? 13:22 < epicmetal> Lope: not sure what the Ubuntu package name is, but yeah probably 13:22 < epicmetal> Lope: if you've only just installed, you're probably better off reinstalling with the Kubuntu ISO for cleanliness 13:22 < Lope> epicmetal: surely I can just apt-get remove --purge gnome3 ? 13:22 < epicmetal> Although it depends on how well Ubuntu package/configure stuff 13:23 < epicmetal> Lope: that would only remove the top level metapackage 13:23 < Lope> I've had multiple DE's before, ubuntu handled it fine 13:23 < Lope> yes, then apt-get autoremove 13:23 < epicmetal> It'll handle it, but you might have stuff in /etc/xdg or whatever that bugs you 13:23 < Lope> it's probably only a few MB 13:24 < Lope> What about XFCE? 13:24 < Lope> Power user vibe, or not? 13:24 < epicmetal> Lope: sort of... 13:24 < revel> Do you want to be one or look like one? 13:24 < Lope> I was very happy with mate, which is a fork of gnome 2 13:24 < Lope> revel: work like one 13:25 < epicmetal> Lope: so why not continue with MATE then 13:26 < epicmetal> Lope: you basically just have to pick the set of bugs that you're happy with ;) 13:27 < Lope> mate allows taskbars for each monitor with a window list of only the windows on it's specific monitor. Mint mate had the mint menu, which was a search based app launcher. Some of the settings of mate (gnome2) I had to set using dconf or gconf or whatever. Then of course you can set keyboard shortcuts to launch programs or scripts (which probably any DE can do) 13:27 < Lope> But the version of mate that's available for ubuntu doesn't have the mint-menu 13:28 < epicmetal> Lope: apt-get install mate-menu ? 13:28 < Lope> The mate menu is not the mint menu. 13:28 < epicmetal> Oh 13:28 < Lope> mate menu just lists app in a static way that you have to configure with .application files. 13:29 < Lope> the mint menu is basically the mate menu plus a search based app launcher. 13:29 < epicmetal> Ok so "matemenu" got dropped from 18.xx? 13:29 < Lope> I used mate-desktop on Ubuntu 16.04. It was okay, but not so great without mintmenu of course. 13:29 < epicmetal> Hmm yeah 13:29 < Lope> mintmenu, not mate menu. 13:29 < Lope> Mintmenu is the one that came with linux mint with the search based app launcher. 13:29 < epicmetal> Sorry yeah 13:29 < epicmetal> I'm asleep 13:30 < epicmetal> Wait, mate-menu is a fork of MintMenu, at least in the Arch repos 13:30 < Lope> No worries. There are other search based app launcher things you can install but I didn't get around to finding and installing a decent one. There was one based on node.js that I thought looked cool, but then it had bugs and I wasn't bothered to persevere at the time. 13:31 < Lope> Weird. 13:31 < epicmetal> I've used mate-menu on Arch and it definitely has search 13:31 < Lope> But I don't sense much development for mate. 13:31 < Lope> Interesting, okay I'll give it a go on ubuntu 18.04 also. 13:31 < hendrix> yeah, IIRC mate-menu in opensuse is basically mintmenu 13:31 < Lope> But I'm keen to try out kde-plasma 13:32 < epicmetal> Lope: you'll love Plasma if you can handle the bugs 13:32 < hendrix> kde and xfce are fine too 13:32 < Lope> So does plasma mean it's all fancy pants like cinnamon? So it rapes your GPU on every click? 13:32 < epicmetal> (although the same can be said for any desktop really) 13:32 < GNU\colossus> has gvfs performance always been this _abysmal_? I'm on a ~100mbit link with <10ms RTT. listing a directory with ~1000 inodes in it takes around 0.02s on the server. listing the same directory over a GVFS mount takes _twenty seconds_. 13:32 < Lope> Is there a KDE without plasma? 13:32 < epicmetal> Lope: it's fancy but you can run it minimal-ish 13:32 < hendrix> Lope: you can disable about all effects and fanciness if you want 13:33 < Lope> hendrix: nice. 13:33 < epicmetal> Lope: no, unless you get Trinity DE which is KDE 3 13:33 < hendrix> kde on my machines runs much higher frame rate than gnome 3/cinnamon 13:33 < Sitri> GNU\colossus: what protocol? 13:33 < epicmetal> hendrix: I can vouch for that 13:33 < Lope> should I use gdm3 or sddm? 13:33 < epicmetal> gnome has known performance issues 13:33 < epicmetal> Lope: sddm if you're using Plasma 13:34 < GNU\colossus> Sitri, CIFS. gvfs mounts it with an "smb-share"-prefix. It's using IPv6. 13:34 < Lope> Gnome3 has the functionality of a cut-down android install. 13:34 < hendrix> I've i5 but gnome animations look like they are 30fps 13:34 < Lope> Gnome3 is like a donkey with makeup. Dumb and pretty. 13:35 < Lope> Thanks epicmetal 13:35 < revel> And great in bed. 13:35 < epicmetal> Lope: gnome does a few things fairly well, it is just limited in a few ways, and slow 13:35 < epicmetal> hahaah 13:35 < hendrix> doesn't ubuntu mate also have brisk menu or fork of it 13:35 < Lope> hendrix: never heard of that 13:35 < buoyantair> Guys, how do I install docker on manjaro? 13:35 < buoyantair> https://docs.docker.com/install/#supported-platforms 13:35 < Lope> Yes, there's mate-applet-brisk-menu 13:36 < buoyantair> I cant find a desktop edition for docker ce 13:36 < epicmetal> Lope: yw 13:36 < Lope> hendrix: will try it. 13:37 < hendrix> it may be under different name 13:37 < epicmetal> assuming kde doesn't butcher plasma with some new major release, it'll be the #1 desktop for linux Real Soon Now 13:37 < Sitri> GNU\colossus: haven't used it in a while, but IIRC samba isn't very fast. Try using something else? 13:37 < hendrix> https://ubuntu-mate.org/blog/ubuntu-mate-bionic-final-release/ "We have... updated Brisk Menu" 13:37 < GNU\colossus> Sitri, samba is not the bottleneck. 13:38 < epicmetal> they just need to not break stuff 13:38 < epicmetal> if they do, they're doomed 13:38 < hendrix> "traditional layout with the menu-bar (Applications, Places, System) replaced by Brisk Menu." 13:38 < epicmetal> also, there was some scary yet unconfirmed data loss bug with plasma 13:39 < epicmetal> https://old.reddit.com/r/kde/comments/1ak1lz/psa_dolphin_exhibits_data_loss_do_not_use_it_for/ 13:39 < epicmetal> so you basically Can't Have Nice Things, Ever 13:42 < MrElendig> buoyantair: pacman -Ss docker 13:42 < buoyantair> MrElendig: Thanks! I found a wiki article on it :D https://wiki.archlinux.org/index.php/Docker 13:43 < buoyantair> :D 13:45 < paul424> Hello, I am looking for some linux IRC client 13:45 < kingrodian> irssi is cool 13:45 < paul424> It should always remember the messages history 13:45 < paul424> and run in GUI 13:45 < kingrodian> oh 13:45 < epicmetal> paul424: hexchat 13:46 < paul424> hexchat : I tried that but it does not rembmeber the messages 13:46 < paul424> naah BLB 13:46 < epicmetal> paul424: did you enable logging? 13:46 < ne2k> paul424, I use xchat-gnome, it's pretty good 13:46 < epicmetal> paul424: it's in prefs 13:46 < paul424> epicmetal: no ... 13:46 < paul424> epicmetal: I would try , later 13:46 < epicmetal> iirc it's on by defaul 13:46 < paul424> ok thanks bye 13:46 < ne2k> kthxbai 13:47 < ne2k> ;-) 14:01 < no_gravity> In the "Displays" manager, I can only set a resolution up to 1680x1050. But I suspect the monitor can run a higher resolution. Is there a way I can try 1920x1080? 14:01 < GNU\colossus> http://paste.debian.net/plainh/c4eff606 - this is how broken gvfs acts in my setup. I have no idea why. :/ 14:03 < no_gravity> Should I try this? xrandr -s 1920x1080 14:05 < no_gravity> I will try it ... 14:05 < no_gravity> pfff "Size 1920x1080 not found in available modes" 14:07 < King_DuckZ> sorry if it's OT but I think this concerns us all - https://www.fsf.org/blogs/community/free-software-is-at-risk-in-the-eu-take-action-now the deadline is tomorrow, so please take a moment to read some info and sign the petition now if you didn't do it already 14:21 < hans_> what is awk doing here? sudo apt-get remove --purge `dpkg --get-selections | grep i386 | awk '{print $1}'` 14:22 < revel> Just printing the first word in each line. 14:22 < mavorsa> its printing money, duh 14:22 < revel> lol 14:22 < dnanib> It is the grep that is superfluous 14:22 < hans_> oh, so in other words, it removes anything except the first word of each line? 14:23 < dnanib> dpkg --get-selections | awk '/i386/ {print $1}' 14:23 < revel> So, it finds all the i386 stuff and prints the first word, which I assume will end up being all of the i386 packages. 14:23 < hans_> kk 14:24 < no_gravity> I'm trying to set the resolution of a Flatron E2210PM-BN monitor to 1920x1080. 14:24 < no_gravity> Any ideas how to go about it? 14:25 < mavorsa> use a hammer? :P 14:25 < revel> Make sure you have your drivers_ 14:25 < escalion> no_gravity: you looked at xrandr? 14:26 < revel> ? 14:26 < mavorsa> (dont mind me, in a silly mood from lack of sleep) 14:26 < no_gravity> escalion: Yeah, I tried some mode lines but they did not work. 14:29 < BluesKaj> Hi folks 14:30 < lilltiger_> King_DuckZ: I dont see a petition to sign on that link 14:30 < escalion> no_gravity: your displays max res is 1280x1024 14:32 < revel> This Romanian site seems to claim its max is 1920x1080 https://www.compari.ro/monitoare-c3126/lg/e2210pm-bn-p49805536/ 14:32 < no_gravity> escalion: Why do you think so? It's currently 1680x1050 and the specs say it supports 1920x1080. 14:33 < MrElendig> better to see the actual datasheet than some random comparason site 14:33 < lilltiger_> But Romanian sites are so trustworthy! 14:33 < MrElendig> how is the monitor connected? 14:34 < MrElendig> damn lg are running their servers on a piece of string 14:34 < revel> Well, the monitor names are quite cryptic, so he could've found the data for a different one. And I assume no_gravity has some better data. 14:34 < MrElendig> 38 seconds to download a 1.4mb pdf 14:34 < no_gravity> MrElendig: Via VGA-out to a vga->mini-display-port adapter. 14:34 < solidfox> MrElendig, lmao, not a potato but a piece of string 14:34 < MrElendig> no_gravity: do you only have vga out? 14:35 < MrElendig> also, data sheet says 1680x1050 14:35 < revel> solidfox: Yeah, their network cable is a wet piece of string :D 14:35 * azarus uses both VGA and DVI on his current monitors, but a new, better one will arrive tomorrow and I'll use DP 14:35 < no_gravity> the monitor also has dvi-d. 14:35 < dogbert_2> no_gravity, the max resolution on that monitor is 1280x1024 @ 75Hz...per the pdf 14:35 < MrElendig> but what *output* do you have 14:35 < solidfox> revel, lolol 14:35 < dogbert_2> you won't get anymore out of it 14:36 < MrElendig> dogbert_2: eh the lg pdf I found says 1680x1050 14:36 < no_gravity> Well, I'm giving up on it. I think it only supports 1680x1050. 14:36 < dogbert_2> and that's from LG itself 14:36 < MrElendig> same here 14:36 < King_DuckZ> lilltiger_: you have to go to the MEP list and send an email to the ones that represent your country 14:36 < revel> So, three conflicting numbers. 14:36 < dogbert_2> http://www.lg.com/uk/support/support-product/lg-E2210PM-BN 14:36 < dogbert_2> that's from LG itself 14:36 < MrElendig> links to file:///home/oh/Nedlastingar/E2210PM-Spec.pdf 14:37 < MrElendig> er bad url 14:37 < revel> lol 14:37 < MrElendig> http://www.lg.com/uk/products/documents/E2210PM-Spec.pdf 14:37 < MrElendig> stupid chrome messing up the clipboard 14:37 < escalion> the lg pdf I found stated 1280x1024 I clearly just found the wrong one :) 14:37 < lilltiger_> King_DuckZ: I was about to, then I read that he will oppose it. 14:37 < MrElendig> no_gravity: what *output* do you have available? 14:37 < escalion> but long&short if you made a modeline and it didn't work, it's not supported 14:38 < dogbert_2> the 2210 is VESA 1680 x 1050 @60 Hz 14:38 < no_gravity> MrElendig: Mini-Display-Port 14:38 < King_DuckZ> lilltiger_: who? 14:38 < MrElendig> use that instead of vga 14:38 < MrElendig> vga is just horrible 14:38 < King_DuckZ> lilltiger_: ah your representative? 14:38 < MrElendig> (and breaks edid) 14:39 < no_gravity> MrElendig: The monitor needs VGA *input*. 14:39 < dogbert_2> you wanna use DVI-D if at all possible or HDMI 14:39 < dogbert_2> then you 14:39 < no_gravity> MrElendig: So I have a converter in the middle. 14:39 < MrElendig> you said it had dvi 14:39 < Reventlov> do VGA supports hotplug? 14:39 < Reventlov> :> 14:39 < Lope> I'm busy trying out kde-plasma on ubuntu 18.04. It's very nice! 14:39 < azarus> Reventlov: sure 14:39 < MrElendig> Reventlov: yes 14:39 < dogbert_2> your stuck at 1280x1024 14:39 < lilltiger_> King_DuckZ: yes 14:39 < MrElendig> no_gravity: the manual also says it has dvi 14:39 < MrElendig> no_gravity: so go mini-dp -> dvi 14:40 < dogbert_2> when I bought my monitor, I made sure the video card would support it 14:40 < no_gravity> MrElendig: My theory now is that the monitor only supports 1680x1050. 14:40 < Reventlov> I think the standard says nothing about hotplug 14:40 < Reventlov> but it's supported usually. 14:40 < MrElendig> no_gravity: correct, but you will also avoid a lot of problems if you use the dvi in instead of going trough vga 14:41 < MrElendig> no_gravity: you will not have to make a modeline etc 14:41 < MrElendig> no_gravity: and you will get better picture quality 14:41 < no_gravity> Well, I'm only using this monitor today. 14:41 < reynierpm> hello there, I am having an issue while self-updating my composer installation and I don't like what I have found so far. First of all I am running composer as the vagrant user but such user can't update the composer executable at "/usr/local/bin/composer". I have tried with sudo composer self-update but for some reason I got this error "sudo: composer: command not found" 14:41 < MrElendig> can pick up a mini-dp -> dvi cable for 5-10€ 14:41 < reynierpm> I have read some recommendations and one of them says to add vagrant to the root group which I don't think is a good idea what would you suggest me on this case? 14:42 < MrElendig> converters costs about the same if you already have a dvi-d cable 14:42 * MrElendig generally prefers cables to converters 14:42 < MrElendig> specially for mini-dp, which is somewhat fragile 14:43 * hans_ bought 10x usb extension cables for a project, only to realize that it created too much signal noise, so had to buy 10x longer usb cables therafter 14:44 < hans_> ...from a chinese shop with no return policy~ 14:44 < MrElendig> you can get more reliable usb cables for cheap 14:44 < Dominian> /12 14:48 < MrElendig> hans_: how long did you need? 14:48 < notmike> MrElendig: what up fam 14:48 < MrElendig> hans_: anything over 5 meters requires an active cable 14:49 < MrElendig> (up to 20 or so meters) 14:49 < MrElendig> they basically have a hub chip at every 5 meter interval 14:51 < hans_> the original cables were about 50 cm, i needed about 100 cm in total, the extension cables were 150 cm, 50+150 didn't work out, bought 180cm cables now 14:51 < MrElendig> must have been really crap cables 14:51 < MrElendig> as expected from the random chinesium ones :) 14:52 < hans_> sigh, i guess 14:52 < MrElendig> but seriously, you can eg get cloth covered deltaco cables for <3.5€ each, so there isn't really any good reason to go with the chinese ones 14:53 < mcj> Hey, does anyone know why Samba has trouble sharing an external drive? I have set the mount point to /media/usr/Archive and set no authentication 14:54 < toffe> Does it work with an internal folder? 14:54 < mcj> I can see the folder, but I can't write 14:54 < hans_> think there's any hope for these cables? (1.8m usb3) https://www.aliexpress.com/item/QGeeM-USB-3-0-cable-Super-Speed-USB-3-0-A-Male-to-Male-USB-Extension/32787783151.html?spm=a2g0s.9042311.0.0.24a94c4d4ykSkY 14:54 < toffe> Might be shared as read only? 14:54 < mcj> I have 3 shares I would like to access. Two are drives, one just a folder in the user 14:54 < MrElendig> forget aliexpress 14:54 < MrElendig> *specially* for usb 3 14:55 < toffe> read only = no 14:55 < toffe> add to smb.conf as a config for each share 14:55 < mcj> That param has been set, along with force user = nobody, browsable = yes 14:55 < MrElendig> hans_: and you could buy from a reputable store for the same price 14:55 < mcj> I just want anyone who connects to our home wifi to have access to these without password. It seems Samba has trouble doing this 14:55 < toffe> Restarted smbd after that? 14:55 < mcj> Yes I did 14:56 < toffe> Hmm, usually works out of the box for me, any logs? 14:56 < mcj> I'm going to try again tonight toffe 14:56 < MrElendig> hans_: which country are you in? 14:57 < mcj> Was hoping maybe there was something that I was missing 14:57 < hans_> MrElendig, Norway 14:57 < toffe> mcj: public = yes, guest ok = yes 14:57 < azarus> mcj: checked logs, permissions? 14:57 < mcj> Oooh, that public setting may be the culprit 14:57 < toffe> guest account = nobody 14:57 < toffe> guest only = yes 14:57 < toffe> have these? 14:57 < MrElendig> hans_: just buy from multicom, they are the only ones with sane prices on cables :po 14:58 < MrElendig> hans_: 45nok for 1.8m m-m 14:58 < toffe> Not a bad price =) 14:58 < MrElendig> https://www.multicom.no/kabler-overganger/cat-l/c100116 14:59 < MrElendig> hans_: I bought a dp->hdmi cable from them, 106nok, second cheapest store was 200+, elkjøp/power/etc wanted 400! for the same cable 15:00 < toffe> Retailstores like Elkjøp and Power have most of their profit on cables. Multicom, komplett e.c. have more on the computerparts :) 15:00 < MrElendig> komplett wanted 200+ 15:02 < MrElendig> currently komplett wants 350 >_> 15:02 < toffe> ewww. bad komplett 15:03 < hans_> any idea what the difference between these are? they're all 1m http://www.deltaco.eu/en/Pages/SearchResults.aspx?k=USB3%2D210 15:04 < MrElendig> the colour and gold plated or not 15:04 < hans_> is there any actual benefit of gold plating, or does it just look cool? 15:04 < MrElendig> actually, probably just the colour and packaging 15:05 < MrElendig> hans_: more reliable connection, in most cases it doesn't matter 15:05 < MrElendig> just buy the cheapest one you find on multicom :p 15:06 < hans_> gotcha 15:07 < MrElendig> I like the deltaco prime just because of the stiffer cable, but that can ofcourse be a downside too 15:09 < hans_> .. but in theory, gold plating is somehow supposed to make a better connection than non-gold-plating? 15:11 < mquin> lower resistence, but it's pretty negligable 15:11 < MrElendig> actually higher resistance 15:12 < MrElendig> but, less corrosion, also mechanically better than some other plating, specially on card edge connectors 15:12 < mquin> it also doesn't oxidize, which can help with longivity 15:12 < hans_> kk 15:12 < MrElendig> oxide free cu/cu would give a lower resistance 15:13 < MrElendig> but cu oxides really fast so ... 15:13 < mquin> I've got some old hardware with tinned edge connectors thay have gotten pretty tempremental 15:13 < MrElendig> not all gold plating are made the same either, many are *really* thin and wears out fast 15:14 < MrElendig> typical on chinesium gold plated stuff 15:16 < MrElendig> had some banana plugs that didn't even last 5 insertions before it was worn off 15:16 < hans_> but only chinesium cables can pee! https://i.imgur.com/9T0LRcp.png 15:16 < Psi-Jack> Heh, chinesium. 15:16 < Psi-Jack> hans_: Heh, nice. 15:17 < MrElendig> that explains the smell of some of them :p 15:17 < widp> I have an application which needs to fetch a large amount of ".desktop" files(desktop entries) 15:17 < hans_> idd 15:17 < hans_> widp, neat 15:17 < widp> it runs slow because of whenever I launch it because of caching. 15:17 < widp> What can I do to make sure that this doesn't happen. 15:17 < MrElendig> widp: should only be slow on first run, unless you run out of ram 15:18 < MrElendig> the kernel will cache file access 15:18 < MrElendig> you could always keep your own cache too, in a more efficient format than the .desktop files 15:18 < MrElendig> so you don't have to parse them every time 15:19 < widp> MrElendig: The application that's causing problems is rofi, an application launcher. 15:19 < widp> How do I go about fixing this? 15:19 < MrElendig> rofi should cache itself 15:20 < widp> huh, what do you mean? 15:20 < MrElendig> it uses XDG_USER_CACHE 15:21 < widp> strange, that variable is not set to anything on my system 15:21 < azarus> widp: if it is not set it defaults to ~/.cache 15:21 < azarus> (i believ) 15:21 < azarus> believe* 15:21 < Psi-Jack> Man, I still cannot believe this. Work computer with chrome open with about 20 tabs, Rambox (electron app), hexchat, evolution, and other various apps... And it's only using 2GB RAM. And in this case, there's barely any cache too. 15:21 < MrElendig> er.. XDG_CACHE_HOME 15:22 < MrElendig> widp: that is correct 15:22 < MrElendig> widp: it should not be set unless you want to override the default, which is $HOME/.cache 15:22 < MrElendig> there was a bug where rofi would not create .cache if it wasn't present, but that was fixed ages ago 15:22 < widp> I see 15:23 < MrElendig> which version of rofi are you using? 15:23 < widp> 1.5.0 15:24 < MrElendig> https://github.com/DaveDavenport/rofi/issues/670 15:24 < widp> yeah, I looked at that issue. 15:24 < widp> and have trying to find a fix. 15:25 < hans_> oh god, AliExpress, a website where most sellers are chinese, does not support chinese characters in messages o.o 15:25 < MrElendig> widp: you could put the rofi cache on a tmpfs 15:26 < MrElendig> hans_: should do utf8 15:26 < hans_> and instead of actually adding support for utf-8, they just documented their lack of chinese support. https://service.alibaba.com/buyer/faq_detail/20111719.htm 15:26 < MrElendig> hmm 15:26 < widp> how would I go about doing that? 15:26 < MrElendig> I get messages from sellers in utf8 all the time 15:26 < hans_> weird 15:26 < hans_> > When you are sending messages using the “Contact Supplier” button, the system may inform you that Chinese characters are not supported. If this is the case, do not enter Chinese characters. 15:27 < MrElendig> mount -t tmpfs .... 15:27 * MrElendig runs off to torture his cat 15:27 < hans_> dafuq 15:27 < hans_> MrElendig, cat is not for torture! 15:27 < Psi-Jack> hans_: Heh 15:27 < escalion> ssh root@MrElendig 15:27 < MrElendig> he was in a fight last night and now smells really bad, so off to the shower 15:27 < escalion> rm -rf / 15:28 < hans_> ohhh, oki, carry on 15:28 < MrElendig> escalion: http://arch.har-ikkje.net/gfx/rm-rf.jpg 15:28 < Psi-Jack> escalion: Keep in mind, stating such a command here is not funny. It's usually insta-banned for it. 15:28 < hans_> MrElendig, rm -rf / --no-preserve-root (BUT DO NOT ACTUALLY WRITE IT) 15:28 < Psi-Jack> hans_: Same to you ^ 15:29 < azarus> or, let's be concise: "rm -rf /*" 15:29 * Armand bans hans_ 15:29 < Psi-Jack> azarus: And you too ^ 15:29 < escalion> I omitted --no-preserve-root intentionally in response to torturing a poor kitten xD 15:30 < azarus> Psi-Jack: because of stupid people that go "lol let's run this" 15:30 < azarus> ? 15:30 < Psi-Jack> azarus: Because this is a help channel, and that's not helpful. Ever. 15:30 < azarus> might be 15:30 < hans_> Psi-Jack, how else do you clear out a chroot? 15:30 < azarus> what if you need to trash your OS as quick as possible? 15:30 < azarus> legit use case 15:30 * epicmetal stifles a giggle 15:30 < Psi-Jack> hans_: You remove it, which wouldn't be / now would it? :p 15:31 < hans_> clear it, without removing it? 15:31 < azarus> (btw, before trashing a chroot, make sure nothing's mounted under it x_X ) 15:31 < hans_> hehe oki 15:32 < escalion> hey at least we didnt change the dram voltages to absurd values 15:32 < epicmetal> I'm coming to the inescapable conclusion that all DEs suck, and you just need to painstakingly build your own environment from the smallest, simplest components possible 15:32 < escalion> epicmetal: ratpoison 15:32 < hans_> 1V then? 15:32 < azarus> dwm! :D 15:32 < Psi-Jack> twm! 15:32 < azarus> cwm! 15:32 < escalion> or i3 15:32 < Psi-Jack> NO wm! 15:33 < hans_> epicmetal, did you try xfce before concluding? 15:33 < ZzZombo> Hello. I recently installed Manjaro on my laptop. However, during the installation I was unable to set my partitions up the way I wanted. The installer didn't let me. So I ended up with a pretty small (5GB) partition for the OS. 15:33 < ZzZombo> Now, using gparted I tried to remedy that, by adjusting my 2d partition with lots of free space, and expanding my 5th partition with Manjaro, but it doesn't seem possible to expand across other partitions. 15:33 < epicmetal> hans_: currently on it 15:33 < escalion> xfce looks like it's the result of a childs crayon drawing 15:33 < ZzZombo> What can I do? 15:33 < escalion> ZzZombo: did you set up LVM? 15:34 < hans_> epicmetal, tried the Adwaita-dark theme? :p 15:34 < epicmetal> hans_: the clock doesn't update after resume. Found some forum post with a solution that requires shell scripting. Which I can do, but sheesh. 15:34 < epicmetal> hans_: it's not a matter of aesthetics, although they suck too 15:35 < hans_> hrm, resume clock works fine for me, debian 10 pre-release 15:35 < escalion> return 418; 15:35 < hans_> (but it worked all the way back to debian 8) 15:35 < epicmetal> hans_: currently using materia dark compact 15:35 < Psi-Jack> ZzZombo: Manjaro is always a bad idea, just fair warning. 15:35 < azarus> dwm is so cool, works just about everywhere where Xorg can run 15:36 < azarus> (with very little porting required) 15:36 < hans_> >Materia is based on Adwaita by GNOME. 15:36 < hans_> huh 15:37 < epicmetal> hans_: it's GTK, it works for Xfce 15:37 < epicmetal> And there's an xfwm4 theme for it 15:37 < ZzZombo> escalion, I don't know. I didn't do it personally though. 15:37 < epicmetal> It is buggy however, requires CSS in Firefox to fix colours 15:37 < escalion> ZzZombo: You're better off doing a fresh install if no LVM, and also manjaro is an awfully bad idea imho 15:38 < epicmetal> Friends don't let friends Manjaro 15:38 < hans_> epicmetal, they inherited that from Adwaita 15:38 < escalion> You'll have more luch with Arch 15:38 < ZzZombo> Why, can I ask? 15:38 < escalion> s/luch/luck 15:38 < hans_> (and apparently didn't fix it) 15:38 < epicmetal> hans_: I hadn't noticed that in Adwaita under GNOME 15:38 < Lope> Is there a way to force xrandr to display a signal on HDMI-1 even though it thinks it's disconnected? 15:38 < hans_> well, i have 15:38 < epicmetal> hans_: but I've seen the bug manifest in MATE and Xfce 15:38 < hans_> hrm wait 15:38 < hans_> im not sure 15:38 < hans_> nvm 15:39 < escalion> Manjaro is rolling release but it comes with its own configuration 15:39 < escalion> hence why I said you'll be better off with arch 15:39 < Psi-Jack> escalion: Yeah, but Arch isn't for newbies. :p 15:39 < escalion> Neither is Manjaro, really 15:39 < escalion> Stick to ubuntu or some derivative thereof such as zorin or elemental 15:39 < azarus> arch and manjaro break in the funniest ways in my experience, and they're nowhere near as lightweight as they are praised to be 15:39 < epicmetal> Actually Manjaro came in handy when I needed a very quick laptop build at work 15:40 < epicmetal> Arch is "simple", not lightweight 15:40 < azarus> epicmetal: agreed 15:40 < epicmetal> Although less simple with systemd 15:40 < azarus> even more agreed 15:40 < escalion> azarus: they can be as light as you want them to be. It just depends how experienced you are 15:40 < epicmetal> Now this burden has fallen to Void 15:40 < azarus> escalion: not really 15:41 < azarus> arch base is already heavy 15:41 < ZzZombo> Does Ubuntu come with Nvidia drivers? 15:41 < escalion> azarus: arch base isn't overly heavy. If you want lightweight go for DSL 15:41 < epicmetal> ZzZombo: Ubuntu comes with everything except security and a good looking default theme 15:41 < azarus> escalion: nah, i have alpine for that 15:41 < escalion> Alpine too 15:41 < escalion> I use alpine for all my build VMs 15:42 < ZzZombo> "except security", what do you mean? 15:42 < azarus> i use alpine for pretty much everything except my gaming pc 15:42 < azarus> which runs on gentoo 15:42 < epicmetal> ZzZombo: Canonical don't bother patching the majority of the OS 15:43 < epicmetal> azarus: I keep looking at and then rejecting Alpine for some reason. Besides the fact that it didn't boot the last time I tried it 15:43 < ZzZombo> And why would I want it then? 15:43 < epicmetal> ZzZombo: because everyone makes things work for it 15:44 < epicmetal> ZzZombo: same reason Android is popular, really 15:44 < epicmetal> Although Android has a good looking default theme 15:44 < Psi-Jack> epicmetal: What? While canonical may be the douche of douches in the Linux community, they do security updates. 15:44 < Psi-Jack> And what they don't do, Debian does. :p 15:44 < epicmetal> azarus: how do you find Alpine? Any insights/comparisons you can impart from extended use? 15:45 < epicmetal> Psi-Jack: Universe/Multiverse are a "community effort" last time I checked 15:45 < azarus> epicmetal: works just as expected. doesn't get in my way. very easy to develop and good and small community 15:45 < epicmetal> i.e. at the disctretion of the individual maintainer to patch 15:46 < escalion> ZzZombo: what's your use case? 15:46 < epicmetal> Psi-Jack: Debian doesn't patch Ubuntu repos. Not directly or timely at least 15:46 < Psi-Jack> They don't. Ubuntu just pulls their patches in. 15:47 < epicmetal> Psi-Jack: not for Universe/Multiverse for a released distro. 15:47 < Psi-Jack> And yeah, universe and multiverse, those obviously are not supported by Canonical. 15:47 < epicmetal> That's the problem, it's not obvious to newbies, who are arguably the target market 15:47 < Psi-Jack> And are unsupported by Canonical as a result. 15:47 < ZzZombo> I'm finding ways to jump off Windows 10. 15:47 < Dominian> Psi-Jack: found something you might interested in looking at, rolling release distro, but still it's interesting: https://www.voidlinux.eu/ 15:47 < Dominian> Psi-Jack: can't really say 'found' someone I know told me about it 15:48 < Psi-Jack> Dominian: I've thought about it. But runit was a concern. I'm actually using Solus. 15:48 < epicmetal> Psi-Jack: what don't you like about runit? I'm considering moving to Void 15:48 < Psi-Jack> epicmetal: Because it's not systemd. 15:48 < azarus> runit is cool, but I'm more familiar to OpenRC, which is used in Alpine and Gentoo 15:49 < Dominian> Psi-Jack: Just thought it was interesting it was a 'LFS" type thing, build from the ground up 15:49 < Psi-Jack> Yep. So was Solus, actually. :) 15:49 < azarus> Dominian: so is Alpine 15:49 < azarus> so is Gentoo 15:49 < azarus> there are many ;) 15:49 < Dominian> Very true 15:49 < Dominian> I stand corrected. 15:49 < Dominian> :) 15:50 < Psi-Jack> Dominian: But.. 2GB RAM used, with Chrome (~20 tabs), Rambox (electron app), hexchat, gnome-session stuff, budgie DE, synergy, evolution, etc.. 15:50 < Psi-Jack> Dominian: Note /used/ 15:50 < kirk781> Electron apps are memory whores 15:50 < Dominian> Psi-Jack: this is with Solus? 15:50 < Psi-Jack> Dominian: Yes. 15:51 < Psi-Jack> Dominian: On Arch, this same utilization was easily 8~16GB RAM /used. 15:51 < Psi-Jack> Fedora and openSUSE too. 15:51 < Dominian> heh 15:51 < Dominian> I wish I used Linux full time in my job, but alas I do not 15:52 < UserUS> Same, that'd be fun 15:52 < Psi-Jack> I used to not. My last job my "computer" was a MacBook Pro. Not anymore. Now I have a nice convertable laptop and powerful desktop. :) 15:52 < Psi-Jack> Both running Linux 15:53 < UserUS> Man, I work at an MSP, and they can't even wrap there heads around Mac 15:53 < Psi-Jack> MSP? 15:53 < UserUS> Managed service provider. Kinda like IT for hire 15:53 < Psi-Jack> Ahhh 15:53 < Dominian> It's the lastes buzz word Psi-Jack 15:53 < Dominian> :P 15:54 < Armand> UserUS: You called ? 15:54 < Psi-Jack> Yeah. I worked for a company, left it before the titanic sunk. Few months after I did, they fired their in-house engineers and moved to MSP. 15:54 < Armand> lol 15:54 < Psi-Jack> Offsite MS Exchange services. Yeah, that'll buff out. 15:54 < escalion> ZzZombo: Try Zorin or elemental first, then go from there :) 15:54 < UserUS> Psi-Jack, You ever work in house? 15:55 < Psi-Jack> I /only/ work in house. :p 15:55 < Psi-Jack> Or do you mean at home? 15:56 < tsglove> Anybody know of a PDF "type writer" program? I know GIMP can write over PDFs, and LibreOffice Draw can do the same. Yet looking for something more "PDF type writer" and that's it. Any suggestions? 15:56 < UserUS> Psi-Jack, interesting, how do you like it? I was thinking about switching. Yeah, in house, not at home haha 15:56 < Psi-Jack> Well, I mean, I've done it like, most of the time. 15:56 < Psi-Jack> Should say something. 15:56 < twainwek> tsglove: i think okular has some of that functionality iirc 15:57 < ZzZombo> escalion, did you mean Elementary? 15:57 < tsglove> twainwek, on my way to check it out. 15:57 < UserUS> Psi-Jack, how long you been there? 15:57 < tsglove> twainwek, your name reminds me of the mid-90s TWAIN scanners! 15:57 < twainwek> haha good catch 15:58 < Psi-Jack> UserUS: Where I am now? Only a few months. My last company I worked for exactly 3 years. 15:59 < escalion> ZzZombo: That would be the one 16:01 < Psi-Jack> Solitary: The thing about Solus is... It's specifically designed for Desktop use, with support for developer-supportive workstations. For example, they will not package postfix, but will package apache (because apache supports development) 16:01 < Psi-Jack> Err.. 16:01 < Psi-Jack> Dominian: Was meant for you ^ 16:02 < Psi-Jack> Apparently I suddenly thought your nick was Solus, and tab completed to that other guy. :) 16:03 < Dominian> hehe 16:05 < Psi-Jack> That's pretty much something almost nobody does these days is desktop-optimized distros. 16:10 < azarus> -fomg-optimize 16:12 < post-factum> -fomg-optimize Reduces a function into a pure set of nop's. -O100500 enables this by default. 16:13 < azarus> http://funroll-loops.info/ 16:13 < azarus> http://funroll-loops.teurasporsaat.org/ * 16:17 < Lope> how can I use xmodmap to turn right shift into insert? 16:18 < c06> hi all 16:19 < benjwadams> What's the advantage of using `ed` over `sed`? 16:19 < c06> i have two machines both have dhcp server with different networks. i am creating VM and trying to get ip from dhcp server. but machines are conflicting when getting the ips 16:19 < Guest17347> hey, when I enter locale on my ubuntu 18.04 terminal, I was wondering if there should be quotation marks around the locales or not. Please see my output here: https://paste.linux.community/view/cb0986b5 I manually adjusted the LC_ALL and LANG locales 16:19 < Urchin> benjwadams: it's interactive 16:19 < benjwadams> I can't say I've gotten particularly deep in either, but definitely know more sed than ed 16:20 < benjwadams> Anything nontrivial i usually just end up writing something in python instead 16:20 < Urchin> benjwadams: to be fair, ed is these days mostly a historical curiosity 16:20 < Urchin> benjwadams: it would only be *needed* if you need to edit text on a computer with no screen involved 16:20 < benjwadams> was looking at this 16:21 < Urchin> benjwadams: ed is a text editor, not a programming language 16:21 < benjwadams> https://www.michaelwlucas.com/tools/ed 16:21 < Urchin> benjwadams: it was used on unix back when the main output was the teleprinter 16:21 < no_gravity> Sometimes my laptop takes ages to shut down. It seems to happen when I have logged into certain wifi spots while it was on. Any ideas to find out what is going on? 16:22 < benjwadams> Ok, what's the advantage of ed, over say, vim, which has a sizeable amount of ed for colon commands 16:22 < benjwadams> I feel bad. Apparently I'm not a "real sysadmin" for not knowing ed 16:22 < benjwadams> according to that author 16:22 < sharp15> is there a filesystem that allows straigt forward removable devices but isn't FAT? ext{2,3,4} didn't the last time i checked. 16:23 < azarus> sharp15: what? 16:23 < benjwadams> Probably still turing complete 16:23 < azarus> what are straight forward removable devices? 16:23 < benjwadams> there was that sed script that ran tetris 16:23 < Atque> no_gravity: When you shutdown your laptop, it sends kill signals to processes running on it. To fix your issue, I'd first check which processes are taking a long time to stop when the laptop shutdowns. 16:23 < sharp15> azarus: permissions confusion. 16:23 < post-factum> azarus: straight forward removable devices are the devices that cannot be removed straight backward 16:24 < sharp15> azarus: they use user numbers not user names. 16:24 < no_gravity> Atque: How do I find those processes? 16:25 < sharp15> sharp15: i'm sorry. i'll stop being confusing. ext-2,3,4 use user-numbers. it means that you can only acquire read permissions for files if your user number matches on the second machine. 16:25 < sharp15> azarus: ^^ 16:25 < Lope> `xmodmap -e "keysym Alt_R = Insert"` seems to work. But does anyone know why pressing Ctrl-Insert writes "5~" into my terminal. I don't want that. 16:25 < sharp15> i'll learn to type eventually. 16:27 < benjwadams> Wow, I actually know quite a bit of ed already from vim! 16:28 < sharp15> benjwadams: ed became vi which became vim. 16:29 < sharp15> benjwadams: visual mode was added to ed for vi and so on after. enjoy. 16:29 < Urchin> sharp15: actually, vim was directly inspired by vi-like emacs, vile for short 16:30 < sharp15> Urchin: ok. i'm certain about ed becoming vi at least. not sure if you are making a joke or not. there are still some heated debates related to vi/emacs. 16:30 < GunqqerFriithian> >not using nano 16:31 < Urchin> sharp15: there were debates between vi users and emacs users, but vim is closer in feature to emacs than vi back in the day 16:31 < Lope> I just switched to gnome-terminal. I'm feeling at home again. 16:31 < benjwadams> ed might yet be useful in some very minimal containers where vi/vim aren't bundled 16:31 < Urchin> sharp15: not to mention that those features did come from a version of emacs that was made to function like vi 16:31 < benjwadams> serious 16:32 < Urchin> even busybox manages to have vi 16:33 < killown> what is the best tool to run android apps on linux? 16:33 < benjwadams> I like to fantasize about situations where i need to use arcane, primitive technology 16:33 < sharp15> ok. i'll try a different question. does anyone know if btrfs is considered "stable" yet or should i stick with ext4? 16:33 < benjwadams> A man can dream 16:33 < revel> sharp15: By whom? It is according to some and definitely not according to others. 16:34 < Urchin> sharp15: XFS 16:34 < sharp15> benjwadams: you can always not bother with dreaming and do it for fun.. there are some pdp emulators and related chips around. 16:35 < sharp15> benjwadams: or come up with something more arcane than that if you wish. 16:41 < post-factum> sharp15: xfs 16:49 < Lope> I changed capslock into shift `xmodmap -e "keysym Caps_Lock = Shift_L"` that works nicely, now I want to make "Control_L Caps_Lock = Caps_Lock" ... Possible? 16:49 < napalmgrenade> Hey can anyone help me set my locales correctly, i'm trying to install oracle weblogic server 16:51 < napalmgrenade> I already tried manually editing the locale file, but when I type in the command locale, I'm seeing this: https://paste.linux.community/view/6432edd7 16:51 < napalmgrenade> concerned about the quotation marks surrounding some of the locales but not others... 17:00 < Psi-Jack> Heh, someone.. actually /wants/ to install weblogic? 17:02 < bb36e> perhaps they're being forced to 17:02 < bb36e> oracle has very sharp lawyers 17:04 < Psi-Jack> Well. Anyway. WebLogic is a propriatery product with a nice hefty licensing fee, but it comes with support. Support they can provide. 17:15 < funabashi> what is a small version of linux if i want to test some gui stuff? 17:15 < Lope> I've added another monitor to my computer but it didn't detect in the display manager (driver issue). I've forced the GPU to output on HDMI-1 and the monitor is functioning normally. Except xrandr is still falsely reporting HDMI-1 is disconnected, so my DE isn't letting me add panels to it etc. 17:15 < Psi-Jack> Linux is a kernel. 17:15 < Lope> Any ideas? 17:15 < twainwek> funabashi: define some gui stuff 17:16 < Psi-Jack> Lope: Have you tried turning it off and on again? 17:17 < funabashi> ok i want to test my touchpad. :p 17:17 < Psi-Jack> It works. 17:17 < funabashi> enough? 17:17 < funabashi> no 17:18 < dreadkopp> I've go a bashscript that starts X which is called at the very end of the bash_rc. the script itselfs checks every second if a X server is running and if not tries to start it ... running the bash script from a logged_in user works fine. however it's failing to start X on login . any ideas ? 17:19 < twainwek> funabashi: personally i'd suggest that if you want to test whether it will work or not not go with a light distro per se and instead go with some variant of *buntu since they include a lot of crap out of the box 17:20 < triceratux> lubuntu 18.10 lxqt ftw 17:20 < Psi-Jack> I kinda agree with twainwek for once on this, except not *buntu, but Solus, Fedora Live, or Debian Live. 17:21 < Lope> Psi-Jack: I've rebooted and tried 2 different HDMI cables. I'm using a HDMI > DVI adapter, that might have something to do with it. 17:22 < Psi-Jack> Nope. Not likely. 17:22 < Psi-Jack> HDMI and DVI basically use the same signals, just HDMI includes audio where DVI does not. 17:23 < Lope> Psi-Jack: The kernel has an option to force video outputs to work even if there's nothing appearing to be connected, will try that http://distro.ibiblio.org/fatdog/web/faqs/boot-options.html 17:25 < Lope> I'm busy trying out kde-plasma on ubuntu 18.04. It's very nice! But I can't get keyboard shortcuts working. Trying to launch gnome-calculator. I've tried just the program name and also the full path to it. I've tried it as a global shortcut and as an application shortcut. When I press the key-combo, nothing happens. I've setup the same keyboard shortcut on Mate desktop environment, and it works. I've set it to have 2 virtual desktops, now the hotkeys for 17:25 < Lope> switching virtual desktops is working... 17:25 < napalmgrenade> Hey I'm trying to install some software, but if I do it as root I get an error telling me that I am the root. If I try to do sudo followed by the installation, I apparently dont have permission to create /home/instance...what do I do? 17:27 < Psi-Jack> napalmgrenade: Welcome back. Still trying to install weblogic? 17:28 < napalmgrenade> ty, yes I am 17:28 < Psi-Jack> shtspk nt alwd hr. pls no. 17:28 < Psi-Jack> And.. Weblogic has commercial support for their commercial product. Call them. 17:30 < napalmgrenade> Wow, sorry to have offended you with my abbreviation of thank you. Thanks for letting me know about the support. 17:31 < Psi-Jack> Don't be so surprised. "ty" is rude. :p 17:31 < Armand> Innit 17:33 < napalmgrenade> I do have a general question, though. I installed Ubuntu 16.04 relatively recently, are there some procedures one should do with permissions and groups on a fresh install? 17:34 < napalmgrenade> I'm just wondering because this is the first time that sudo has not worked for me in terms of installing a program 17:34 < Lope> is there a way to make xrandr SAY that a monitor is connected (it actually is) when it thinks it's not? What does --set allow me to do in this regard? 17:36 < Psi-Jack> napalmgrenade: Weblogic is a rather "special" case entirely. 17:37 < V7> Hey all 17:38 < V7> When trying to boot from CD is shows SysLinux Copyright and then shows "boot:" prompt 17:38 < V7> This mobo is old enough 17:38 < V7> Also, when typing "ls" into this prompt it shows directories on CD 17:39 < Psi-Jack> napalmgrenade: And you shouldn't be installing software /to/ /home. Make a system account and group, make a home for it in /opt/weblogic, chown it to the user/group you created previously, install as system account. 17:39 < Psi-Jack> And contact WebLogic to complain about their stupid security issues. 17:46 < twainwek> Lope: what are graphics card and drivers? 17:48 < V7> Anyone ;) ? 17:49 < Psi-Jack> V7: Sounds like a bad ISO image, CD burn, etc. 17:49 < V7> Screen: https://i.imgur.com/1FpAEqW.jpg 17:49 < V7> Psi-Jack: On other machine boots properly 17:50 < Psi-Jack> Looks like EFI. 17:51 < V7> You mean, UEFI ? 17:51 < Psi-Jack> No, I mean EFI 17:52 < V7> Psi-Jack: This mobo is old and it's BIOS is old enough. I don't think that it has EFI 17:52 < Psi-Jack> That could be a problem if that ISO you're using doesn't support non-EFI. 17:53 < V7> It boots on other machine with Celereon D on board 17:53 < Lope> twainwek: Psi-Jack: It's a laptop with Nvidia Optimus GT540M (Optimus means it's a nvidia GPU and also Intel GPU and it switches between them to conserve power or provide performance) 17:53 < Psi-Jack> UEFI is the replacement BIOS, while EFI is the name/label of the partition where UEFI boots an EFI stub for. 17:54 < twainwek> Lope: nvidia prop driver? 17:54 < Psi-Jack> V7: Okay. That's nice. And useless information. 17:54 < Psi-Jack> Ewww, optimus. 17:54 < V7> Psi-Jack: I mean that this is not UEFI ( EFI ) related 17:54 < twainwek> last time i checked nvidia still didn't fully support linux... so it's always on if it's using the card 17:55 < V7> This mobo doesn't have anything EFI related 17:55 < Lope> twainwek: yes, using the latest ubuntu provided nvidia proprietary driver 17:55 < V7> Thank you for though, btw :) Psi-Jack 17:56 < Psi-Jack> Again, UEFI is in hardware. EFI is on disk. :) 17:56 < twainwek> Lope: i messed around with that few years ago and i had to explicitely change my xorg.conf and few other things to get it to "work" (with it staying on all the time) 17:56 < V7> Oh, you mean that this CD is for UEFI only ? 17:56 < Lope> The HDMI > DVI adapter I used had pins C1-C4 and the blade pin from DVI-I but my monitor only has a DVI-D connector https://en.wikipedia.org/wiki/Digital_Visual_Interface So I pulled out pins C1-C4 with longnose pliers, and that allowed the adapter to plugin to the monitor. The pins weren't even connected to anything internally. But still the laptop whether using the Intel or Nvidia GPU doesn't detect the monitor. 17:56 < Psi-Jack> V7: It's looking like a possability. 17:56 < V7> Again, Psi-Jack it boots on other machine which doesn't have UEFI on board 17:56 < Psi-Jack> V7: What is that, actually? 17:56 < twainwek> Lope: post xorg log 17:57 < V7> Arch32 ( i686 + 64 ) 17:57 < Lope> I can force it to turn on the display using xrandr, but then xrandr still tells the DE that the monitor is disconnected, and then I can't get a wallpaper and panel etc, but can drag windows there. 17:58 < Lope> I've borrowed another monitor, it's got an HDMI input, and when using the HDMI cable only (no adapter needed) the laptop detects the monitor and all works as expected. 17:58 < twainwek> post xrandr output to 17:58 < twainwek> too 17:58 < Lope> It's a different machine. The main thing is that xrandr reports HDMI-1-1 is disconnected, even when it's driving a signal through it. 17:59 < Lope> And I could make it all work if there was simply a way to tell xrandr to pretend the monitor is connected so it can tell that to the DE. 17:59 < twainwek> oh man you got physical adapters in there in addition to the optimus fiasco.. no clue 17:59 < Lope> Yeah, it's only with the HDMI > DVI adapter that it doesn't detect the monitor. With just a HDMI cable, it detects the monitor. 18:00 < Lope> So it's obviously a crappy adapter. 18:00 < twainwek> plus if that's all you're trying to do, you should be able to modify some conf file of your de to show the taskbar 18:04 < Lope> I just tried the adapter with my desktop computer that has only intel graphics, and it had the same issue. It didn't detect the monitor with the HDMI > DVI adapter. 18:05 < rqh4> i am new using curl 18:05 < rqh4> what does curl -X PUT does? 18:05 < rqh4> what request does it gets sent 18:06 < rqh4> all i know is PUT is for sending and GET is for recieving 18:06 < twainwek> look up the man pages. it tells it to send a PUT request 18:10 < Lope> Looks like it's a common problem with these type of adapters https://www.amazon.com/Importer520-Plated-Female-DVI-D-Adapter/product-reviews/B0035B4LJM/ref=cm_cr_othr_d_paging_btm_6?ie=UTF8&filterByStar=one_star&reviewerType=all_reviews 18:13 < Lope> The crappy adapter probably lacks the hot-plug-detect pin. 18:14 < MacWinne_> there is a /bin/bash process that is taking up 100% of CPU on one of my servers. It was started on June15th. I'm trying to figure out what it's doing and how it got started. Any tips? /proc/27699 shows the CWD it was started from 18:14 < Lope> The adapter is probably designed to output a signal from a DVI port of a GPU and convert that to an HDMI output, not convert an HDMI input to a DVI input. 18:15 < MacWinne_> ps -ef just shows "user 27699 1 13 Jun15 pts/1 14:33:03 /bin/bash" 18:15 < tim241> MacWinne_, are there any processes that are running under that /bin/bash process? 18:15 < tim241> check htop/top 18:16 < MacWinne_> tim241, htop doesn't show anything under it's tree 18:16 < tim241> hm MacWinne_ , try killing it then? 18:16 < MacWinne_> tim241, yeah, was going to. but trying to figure out what it was doing first so I can prevent it from happening again 18:17 < tim241> hm, no idea :3 18:17 < twainwek> MacWinne_: is there anything under /proc/27699/fd 18:18 < MacWinne_> twainwek, yep. /dev/pts1. a bunch of broken links like 9 -> pipe:[3344394] 18:18 < drumcode303> i don't have sound anymore. something seems to have killed my onboard sound, now i'm stuck with nvidia/hdmi. 18:18 < MrElendig> MacWinne_: tree view can be handy 18:19 < MacWinne_> MrElendig, yep. looked at tree view on htop.. didn't see anything under it 18:19 < MacWinne_> twainwek, interesting.. I see pts/1 which seems to be related to a multitail session 18:26 < drumcode303> how would I troubleshoot my system for sound when lspci and aplay list different devices? (lspci has two, aplay only one) 18:41 < bls> troubleshoot what? and aplay is going to use which every device alsa has set as the default 18:41 < RobertGridCoin> hii 18:42 < RobertGridCoin> Any humans this is my First time on irc 18:42 < Cmaj6> If linux was a women i'd f*cking marry her!!! 18:42 < RobertGridCoin> but its os 18:42 < RobertGridCoin> who u can make , its open source !!!!!! 18:42 < compdoc> we are all bots 18:43 < RobertGridCoin> haha humanBot 18:43 < revel> Did I pass the Turing test? 18:43 < RobertGridCoin> nope , even gogle dulex cannot do that 18:43 < compdoc> yes, but with a C- 18:44 < RobertGridCoin> google , can't do it 18:44 < RobertGridCoin> ok anyone can tell of you tell me ur best shell command 18:45 < revel> rm 18:45 < polprog> find . | xargs rm 18:45 < RobertGridCoin> haha 18:45 < cortexman> extract a website from a warc file.. 18:45 < polprog> dont do this at home kids 18:45 < cortexman> how 18:45 < revel> Yeah, do it at work. 18:46 < RobertGridCoin> haha yes at work , on linux server 18:46 < polprog> someones gotta somehow test if backups work, duh 18:47 < RobertGridCoin> so what os do u use , me Ubuntu 18:49 < triceratux> ubuntu lets me remain a n00b. thats never been so true as with lubuntu 18.10 lxqt. it will change everything 18:50 < RobertGridCoin> y bro 18:51 < sauvin> RobertGridCoin, "you", not "u", and "why", not "y". Use standard English, please. 18:51 < RobertGridCoin> okay , understood Commander 18:52 < triceratux> its the qt. it runs kde apps really easily without having to be kde itself. & i looks exactly like xfce so its usable from the first minute 18:53 < xormor> I use Debian "stretch" 9. I tried "Devuan" ASCII 2.0 but it remains a curiosity to me. I ran it from a USB thumb drive stick. I have Debian stable "stretch" 9 in my computers. 18:56 < RobertGridCoin> I am noob to Linux , windows does not even have comparison with Linux. 18:58 < hexnewbie> It's like comparing a car to car factory with the cars in it. 18:59 < RobertGridCoin> yes , very much kind of. 19:00 < Psi-Jack> polprog: snap piping find to xargs for that not do efficient. With -exec rm \{\} + or even just -delete you get the sane thing but better. 19:00 < Psi-Jack> Also... Not snap. 19:00 < polprog> Psi-Jack: yeah i know 19:01 < polprog> cant tell you that from the top of my head, dont use it that often 19:02 < Psi-Jack> xargs and find are both from the findutils package / sources. 19:02 < jim> RobertGridCoin, couple things... first, I hope you're enjoying your experience, I'm curious what you want to do with it mostly? 19:03 < RobertGridCoin> i do programming , it has java , python already installed , it respects and understands programmer 19:04 < bls> it does? good, because I'm tired of all the disrespect I get from other OSs 19:04 < Psi-Jack> It does not respect. That's a human thing. Software is not human. 😊 19:04 < Psi-Jack> Heh 19:05 < jim> RobertGridCoin, yep, and there are thousands of languages you can explore. linux and other unix-alikes were pretty much made for programming. ok, second thing: 19:05 < RobertGridCoin> its dev respects programmers 19:05 < morf> thousands? dude... what plat are you from? :) 19:05 < morf> planet* 19:05 < Psi-Jack> Made by programmers actually. Not exactly for. 19:05 < Psi-Jack> Hehe 19:05 < nixfreak> how do i make this work ? var=sed -n `echo ${linetwo}/p` $FILE1 19:06 < nixfreak> I know I am missing escapes somewhere 19:06 < bls> nixfreak: can't really tell from that snippet what you're trying to accomplish 19:06 < Psi-Jack> RobertGridCoin: lol. You haven't seen the lkml or the Linux source code have you? The comments are... Amazing. 19:06 < nixfreak> Ill post it to a paste 19:06 < jim> RobertGridCoin, I'd like to request that you spell out y as why and u as you... some newer english speakers have native languages where they don't hear y or u as why or you, and we would like to promote maximum understanding 19:07 < hexnewbie> Linus just tells it like it is, and you turn into a good kernel programmer because of it. Honesty is respect. ;) 19:07 < RobertGridCoin> nope , but if you recommend it ,then defiantly will do 19:07 < nixfreak> https://paste.linux.community/view/75c8b46e 19:07 < Psi-Jack> nixfreak: your missing $(...) 19:08 < nixfreak> the variable linetwoprint 19:08 < nchambers> nixfreak: :| 19:08 < bls> nixfreak: first off, try shellcheck.net, it's going to tell you to use $() instead of ``, which doesn't nest the way you need it to 19:08 < hexnewbie> The lack of comments in the kernel source code - or documentation, since comments are overrated - is truly daunting. I know I'm probably looking in the wrong place, but the VFS API changes with every version, I still adapt my filesystems by looking at the others, since I have no docs on the behaviour. 19:08 < nixfreak> it does but want to get this working first 19:09 < RobertGridCoin> yes , I will , thanks for mentioning it. 19:09 < Psi-Jack> nixfreak: shellcheck as sobbing at your code. Avoid using backticks. 19:09 < nchambers> nixfreak: get it working by listening to shellcheck 19:09 < nchambers> theres a reason it suggests the things it does 19:09 < dnanib> linetwo=`echo -e $linecount / 2 | bc` 19:09 < dnanib> - can be simplified as: linetwo=$(( $linecount / 2 )) 19:09 < bls> you can also .... ^ 19:10 < nchambers> avoid the use of echo with flags, and cat | anything is usually wrong 19:10 < bls> and cat $file | wc can be simplified as just wc $file 19:10 < nchambers> wc "$file" 19:10 < nchambers> don't forget quotes 19:10 < bls> well then don't forget {}s either: wc "${file}" 19:10 < Psi-Jack> Quote all the variables. 19:10 < nchambers> the {} doesn't make a difference there 19:10 < nchambers> the "" does 19:10 < nixfreak> what about the sed statement 19:11 < dnanib> And those sleeps... So totally Windows-y. "Please wait while I do nothing" 19:11 < bls> not if there's no space in the variable 19:11 < Psi-Jack> ^ 19:11 < nchambers> bls: could be tabs, could be new lines 19:11 < nchambers> could be special shell characters 19:11 < nixfreak> if your not using sleeps how do you slow down the script then 19:12 < Sitri> Why do you even want to slow it down? 19:12 < nchambers> might as well just quote it to be safe anyways. if it doesn't need the quotes, no harm. if it does, then you're already good 19:12 < dnanib> Why do you want to slow it down? 19:12 < bls> same can be said for {} 19:12 < Psi-Jack> bls there's a difference 19:12 < nchambers> bls: no 19:13 < nchambers> {} makes absolutely no difference thre 19:13 < nchambers> no matter what the variable contains 19:13 < Psi-Jack> Without quotes there could be unusual or even security issues. 19:13 < hexnewbie> Forgetting the {} in "${a}" is usually easily noticed, less destructive, and visible by your editor's syntax highlighting. 19:13 < bls> yes, but I'm saying if you're going to be defensive, then do it right 19:13 < nchambers> but thats not defensive... 19:13 < dnanib> One more thing. You are processing the entire file in lines 5, 7 & 8. That can be optimized a lot. 19:13 < Psi-Jack> Especially for the unknown. 19:14 < hexnewbie> "$a" is not wrong, $a - with the exception of a few corner cases - is usually both wrong and dangerous 19:14 < nixfreak> find 1st , middle and last lines of a file 19:14 < nchambers> hexnewbie: yes exactly 19:14 < nchambers> but ${a} by itself is pointless 19:14 < bls> `foo='a'; echo "$foo"` ...now I want to output a_b so I `echo "$foo_b"` 19:14 < bls> oops 19:14 < nchambers> yes, you would use it there 19:14 < nchambers> but that is not the case here 19:14 < nchambers> its by itself, so its pointless 19:15 < bls> so how does it hurt to add it to begin with? 19:15 < nchambers> i never claimed it hurt 19:15 < nchambers> i said it was pointless 19:15 < bls> it's pointless until it's not 19:15 < hexnewbie> bls: The 12 times I needed to do this, the syntax highlight of vim or kate told me. 19:15 < nchambers> and at that point you should use them 19:15 < bls> I could make the same claim for "s. it's pointless until you need it 19:16 < nchambers> except thats not true :) 19:16 < bls> so may as well use them 19:16 < nchambers> wat 19:16 < nchambers> you're not making any sense 19:16 < nchambers> quoting variables had much different behavior than wrappign it in {} 19:16 < Psi-Jack> Agreed 19:16 < hexnewbie> bls: It's not the same. The quotes may become needed at *runtime* as someone creates dangerous filenames. The curly brackets can only become needed at code editing time. 19:17 < nchambers> ^ 19:21 < Psi-Jack> Not just dangerous filenames, but dangerous outputs from running commands and trapping the output into variables. 19:22 < Psi-Jack> The unknown. 19:22 < vovioheler> is there anything stoping me from using 80 ports at the same time on a debian gateway? 19:22 < Psi-Jack> Nope. 19:23 < bls> right, and I'm saying both prevent a specific class of problem, and I consider them both worth preventing, not prevent one all the time and prevent the other some of the time, or hope someone editing your scripts knows better an remembers to add them in 19:23 < Psi-Jack> That's bash. :) 19:24 < Psi-Jack> You either know it well, or you know it enough to be damaging. 19:24 < V7> Just caught a line which could solve or say why it doesn't want to boot on thi machine 19:25 < nchambers> bls: if they don't know to put {} in the few times they use it, so that $foo is still separate in $foobar, then they shouldn't be editing scripts anyways 19:25 < bls> because not everyone knows better, so setting them up to not make breaking updates to a script is worth the time it takes to type "${foo}" instead of $foo 19:25 < V7> Just about for 0.1s and it completely freezes a computer after this line, caught via video: https://i.imgur.com/ZM8Me4R.jpg 19:25 < V7> What could this be 19:25 < bls> nchambers: if only we could fire people for being bad as shell scripting :P 19:25 < Psi-Jack> You can. :) 19:26 < nchambers> people are going to write shitty software. if they're not smart enough to look up what to do when you have a variable named $foo, but need to use it as $foobar, then you'll probably fuck up the script worse anyways 19:26 < Cache_Money> I’ve downloaded my public key to my server (long story but using the DigitalOcean console to access it) and now I’m trying to put the ocntents of that file into my /.ssh/authorized_keys file. I tried $ sudo cat gistfile.txt > ~/.ssh/authorized_keys but I get a “Permission denied” response. Any idea how to go about this? 19:26 < V7> nchambers: THis is option too, isn't it: $foo""bar 19:27 < Psi-Jack> or "$foo"bar 19:27 < nchambers> V7: there are several options 19:29 < Cache_Money> Do I have to manually type in my public key if I’m getting a Permission denied response when trying to cat it into my authorized_keys file? 19:30 < bls> Cache_Money: no 19:30 < Psi-Jack> Umm.. Definitely no/. 19:30 < bls> Cache_Money: the permission denied is due to permissions on one of the files, not the actual key data itself 19:30 < capn> tee? 19:30 < Cache_Money> bls: I tried using sudo but had no luck 19:30 < Psi-Jack> ... 19:30 < Psi-Jack> sudo? But... why? 19:31 < Sitri> sudo isn't a magic fix button 19:31 < searedvandal> set the correct permissions 19:31 < ntd> speaking of su/sudo magic 19:31 < bls> Cache_Money: then you've likely trashed the permissions on multiple files/directories in your home if you've been resorting to that 19:31 < Psi-Jack> Cache_Money: Explain what exactly you've done, and what exact problem(s) you are having. 19:31 < ntd> read the "latest"? (old timing attack when "su user -c foo") :P 19:32 < Cache_Money> I’m trying to regain access to my DigitalOcean droplet. I’m logged in via their web console and need to get my new public key added to my authorized_keys file so that I can ssh into it 19:32 < nchambers> Cache_Money: whats the output of: stat ~/.ssh/authorized_keys 19:33 < bls> also namei -l ~/.ssh/authorized_keys 19:33 < Cache_Money> Here are the permission on the authorzed_keys file: - r - - - - - - - 19:33 < Sitri> What about the .ssh directory? 19:33 < Psi-Jack> Not just permissions, but owners too. 19:33 < bls> ~/.ssh should be 700/rwx------ and ~/.ssh/authorized_keys should be 600/rw------- 19:34 < Psi-Jack> authorized_keys can be u=rw,go=r 19:34 < Psi-Jack> That's not as strict. 19:34 < thePacket> hi guys 19:34 < Sitri> .ssh/ has to be go-rwx 19:34 < thePacket> does anyone an implementation of l2tp that is not dead? 19:35 < Psi-Jack> .ssh has to be u=rwx,go= 19:35 < Sitri> 12tp is what? @thePacket 19:35 < Cache_Money> my user (deploy) is the owner of those files 19:35 < Cache_Money> So, I should chmod 600 ~/.ssh/authorized_keys ?? 19:35 < bls> Sitri: L2TP, (get a better font :P ) 19:36 < bls> Cache_Money: yes, it currently isn't writable 19:36 < Cache_Money> bls: okay. I’ll give that a try 19:37 < bls> Cache_Money: and as others mentioned, make sure you own that file and the directory it's in 19:37 < Psi-Jack> Oh perfect! 19:37 * Psi-Jack wraps bls up with Saran-Wrap 19:37 < Psi-Jack> :) 19:37 < Dan39> psi-wrap 19:37 < Saran-Wrap> :D 19:38 < Dan39> oh Saran-Wrap 19:38 < Dan39> creepy 19:38 * Saran-Wrap wraps Dan39 19:38 < Dan39> o_O 19:38 < Dan39> X_X 19:38 < Cache_Money> bls: that worked. THanks for the help~ 19:41 < LunaLovegood> How do I get data integrity with my dmraid ? If I understand it correctly, I can corrupt (with dd on /dev/sda or something) one of the two drives in my software RAID 1 thing, and Linux won't be able to tell which of the two disk has the real data. 19:42 < sh1ro> more drives 19:43 < LunaLovegood> and RAID 5? 19:43 < sh1ro> or 6 or 10 19:43 < sh1ro> or keep hashes on a flashdrive or something 19:44 < hexnewbie> LunaLovegood: That's the case for all RAIDs. Some filesystems get checksums, though. But the data can also get corrupted before reaching the filesystem (e.g. when in RAM) 19:44 < bls> you'd need something that can do bit-rot protection 19:45 < LunaLovegood> Is there something with LVM or whatever that would give me per-sector checksums? 19:45 < azarus> i use raid6 and scrub periodically 19:46 < LunaLovegood> Would a 3-drives RAID6 make sense? 19:46 < azarus> LunaLovegood: not possiblöe 19:46 < azarus> possible* 19:46 < LunaLovegood> why not? one data and two parity drives? 19:46 < azarus> RAID6 needs at least 4 19:46 < LunaLovegood> it has triple parity? 19:46 < azarus> no 19:47 < hexnewbie> MD RAID can do some non-orthodox things 19:47 < azarus> LunaLovegood: https://en.wikipedia.org/wiki/RAID 19:47 < tds> I think ZFS raidz2 can work with 3 drives 19:47 < azarus> "RAID 6 requires a minimum of four disks." 19:47 < hexnewbie> For example, it can do 3-disk RAID-10. However, the mathematical reduction of RAID 6 to 3 disks would be a 3-way mirror, so RAID 1. It's not a genuine RAID level. 19:48 < tds> You might as well run a triple mirror at that point though 19:49 < LunaLovegood> Yeah I suppose but I'd hoped it could work with two drives mirroring, with checksum metadata stored somewhere else 19:49 < azarus> Remember: RAID is not a backup 19:50 < hexnewbie> Well, hopefully MD is intelligent enough to see what the disks are voting, so if one drive differs, it takes the data from the other two. But that would exist in the code only if 3-way mirrors were popular enough for the programmer to consider that case, there's no guarantee it will happen 19:50 < Sven_vB> hi! I'd like to automate the task of reading and modifying numbers at specific positions in ELF binaries. currently I have one long ugly command "dd … | od …" to read and another "printf … | dd …" for writing. that's for unsigned integers; it gets waaaay worse for floats. does Ubuntu have better tools for that? 19:50 < hexnewbie> azarus: And backups are also not a way to certify the integrity of the current live data 19:51 < LunaLovegood> I can't just use ZFS or something, because I want to run DRBD somewhere between the disks and the filesystems 19:54 < hexnewbie> Hm, XFS and ext4 have both added checksums, but only for the metadata, so far. 19:55 < hexnewbie> NILFS has checksums, but it's something I've never heard anyone use (I've been tempted by the feature set) 19:56 < LunaLovegood> What I'd initially thought would have been : ( SATA x 2 -> DMRAID1 ) x 2 -> DRDB -> EXT4 or other 19:57 < hexnewbie> LunaLovegood: If you format this combination with NILFS on top, you'd probably make up something you're the only person on the planet running 19:57 < LunaLovegood> It would have been great for bad sectors or power outages, but it does nothing for data integrity. EXT4 or whatever I use at the end would have no way to know it has 4 locations to read from, even if it had data checksums. 19:58 < hexnewbie> Although I guess RAID underneath DRBD is common in itself 19:59 < LunaLovegood> isn't there some overlay block device thingy I can stick between SATA and software RAID1 that would consume some space in order to store block checksums every 128 blocks or so? 19:59 < sh1ro> ZFS is amazing if you can deal with the annoyance of dkms or custom kernels 19:59 < lupine> I wasn't impressed last time I used it 19:59 < lupine> very poor scalability 20:00 < sh1ro> yeah it's really nicer on bsd :\ 20:00 < spreeuw> and high ramuse and low performance from a user pov 20:00 < spreeuw> nice for data safety though 20:01 < Combined2857> I used the command "badblocks -v" on a hard disk with data in, will that erase the data ? 20:01 < sh1ro> no 20:01 < spreeuw> hows the linux version of zfs coming along? 20:01 < Combined2857> thanks sh1ro 20:01 < lupine> (gpt (luks (pvs (vg (lv (ext4 data)))))) 20:01 < sh1ro> -w nukes drives 20:02 < spreeuw> plus you kinda need ecc ram 20:02 < sh1ro> not need but it's preferable 20:03 < spreeuw> for my simple fileserver dm raid mirror with ext4 suffices 20:03 < spreeuw> easy to recover too 20:06 < sh1ro> reminds me. it's worth backing up your luks header 20:06 < sh1ro> if that sector fails you're boned 20:06 < spreeuw> I think it might be using a copy somewhere 20:07 < spreeuw> but that would also at the start somewhere 20:11 < kubast2> Can someone provide me with /me proof LXC tutorial ,trying to get it running on virt-manager ,but it says there's no connection driver avalible for lxc:// 20:11 < GlenK> hi there. is there a canonical way to run virtual machines these days? or does it all sorta vary from distro to distro? 20:12 < tonyt> glenk use virtual box 20:12 < bls> it's all the same tooling under the hood. the differences are mainly going to be in the frontends 20:14 < bls> so libvirt and kvm underneath, then virt-manager or the proxmox UI or the RHEL interface 20:15 < kubast2> still no idea where to find "connection driver" for lxc 20:19 < Psi-Jack> tonyt: Ugh.... 20:19 < Psi-Jack> Here's some slime. Use this! It'll work for you! ;) 20:20 < sauvin> I use Virtualbox. :P 20:20 < Psi-Jack> I'm sorry to hear that. 20:20 < morf> use libvirt gez 20:20 < sauvin> I also use qemu, but I don't like it for running XP. 20:21 < LunaLovegood> Okay so I guess I'll try a BTRFS RAID1 filesystem since it has data checksuming, and then I'll make a big empty file and use DRBD on it with a loopback device. 20:22 < spreeuw> XP? 20:22 < spreeuw> eol amirite 20:22 < sauvin> Yes, but it's still what I need when running legacy engineering software. Those virtual machines have zero net access. 20:22 < spreeuw> https://twitter.com/marcan42/status/1008981518159511553 HP servers 20:23 < DLange> yup, best exploit 2018 award, strong contender! 20:23 < bls> LunaLovegood: I'd read up on the bugs in btrfs' RAID implementation very carefully before adopting it for anything other than an experiment 20:23 < spreeuw> almost better than the sun solaris where you could just root a box via telnet 20:25 < bls> although BTRFS + RAID + DRBD + a loopmount sounds pretty experimental 20:29 < Siecje> What is the proper way to re-parent a long running process to init.d? 20:30 < LunaLovegood> bls: it's actually BTRFS/RAID1 + loopmount + DRBD + BTRFS 20:31 < LunaLovegood> Oh, and then it gets exported over NFS 20:32 < LunaLovegood> man, what a mess 20:33 < LunaLovegood> Individual SATA drives can read about 200 MB/s, I wonder what I'll benchmark this at. 20:33 < LunaLovegood> *my SATA drives, I mean 20:40 < lnnb> Siecje: init.d, is that a pid 1 program? 20:41 < LtL> lnnb: init is, yes 20:41 < lnnb> LtL: init is pid 1, is init.d the pid 1 portion of an init program? 20:42 < lnnb> if so you can do the daemon thing i forget the steps is it fork() setsid() fork() (then exit in the parent fork?) 20:42 < LtL> lnnb: init.d/is a directory with init scripts 20:42 < LtL> init is #1 PID 20:43 < lnnb> if init is a directory, how can you reparent a program to a directory? 20:43 < lnnb> chroot!? 20:43 < lnnb> init.d 20:44 < LtL> lnnb: i dunno, systemd magic? 20:44 < phogg> lnnb: "init" is what we call the first program to run (PID=1), /etc/init.d/ is a directory. 20:44 < lnnb> is Siecje your alt account, LtL ? 20:45 < LtL> lnnb: no it is not. 20:45 < phogg> when a process loses its parent it is re-parented under PID 1 (init) 20:45 < lnnb> i don't need this lesson but thanks for the refresher for the people in here without a clue 20:46 < LtL> lnnb: i have no fefault accounts, had this one over 12 years. 20:46 < LtL> *default, rather 'other' accts. 20:47 < phogg> ah, having scrolled up I see it now 20:48 < phogg> Siecje: "init.d" is a directory, not a daemon. The init daemon is just called "init", or these days may be called "systemd". To re-parent a long running process you merely need for its parent to exit without signaling for it to be terminated, or for the parent to otherwise disown it. 20:51 < LtL> lnostdal: i beg your pardon, did i speak out of turn? e.g. Siecje ? 20:51 < LtL> lnnb: i beg your pardon, did i speak out of turn? e.g. Siecje ? 20:51 < LtL> lnostdal: sorry 20:55 < phogg> Siecje: more generally the process itself should fork twice and exit the original process and the first forked process. The remaining process will now be a child of pid1 and you have more or less created a daemon. 21:03 < Siecje> phogg: How do you exit without singalling a sub process? 21:07 < bls> don't believe it's possible once something is already started, but using the daemonize command to start it /should work 21:12 < jprjr> Whenever I write code nowadays I don't bother putting a daemonize option in it. Every modern init system doesn't need the doublefork, if somebody insists on using sysvinit they can use start-stop-daemon to daemonize it themselves. 21:20 < qrvpzvb> su requires a valid username right? 21:21 < qrvpzvb> I can't just use an arbitrary numeric ID 21:24 < evanesoteric> What exactly is happening when I do rsync -e "ssh .." ... as opposed to without -e? Is the information encrypted, does it now use the ssh logs? 21:28 < koala_man> evanesoteric: you're just customizing the network transport 21:28 < koala_man> rsync uses ssh by default, but you can choose any other transport tool, or just customize how ssh is invoked to do it 21:28 < evanesoteric> So using port 22 instead of 879 or whatever? 21:33 < ananke> evanesoteric: '-e' and 'without -e' can mean the same thing, pending how the url is constructed 21:35 < bls> and whether or not there's an rsync daemon running/listening on the other end 21:49 < binTrl> Why did MTP had to be this bad :( . I cut some files from my device to my PC and now they're nowhere 21:50 < Amm0n> binTrl, did you "sync" the device? 21:51 < binTrl> I did nothing, just mtp cut from Thunar. 21:52 < binTrl> you'ren't supposed to sync from gui apps like thunar. I hate thunar for this as well 21:52 < binTrl> I am screwed :( 21:53 < lukey_> rule of thumb for me is to always copy files and then delete them from the source when dealing with different disks/devices; moving on the same filesystem is fine 21:54 < binTrl> lukey_: New rule of thumb, fsck mtp. *Always* use ADB 21:58 < jim> newer rule of thumb: if you don't have two, you have one or less 22:09 < birdbolt1> hallo 22:10 < birdbolt1> im curious why do some channels have two pound symbols 22:10 < kerframil> birdbolt1: I think it's to denote that they are not officially associated with the project or theme in question 22:11 < kerframil> birdbolt1: for instance, ##linux is not administered by Torvalds or the Linux Foundation 22:11 < birdbolt1> hmm interesting. I tried to join #linux and it was invite only so i was curious 22:12 < ayecee> probably because you were already in ##linux 22:12 < birdbolt1> Do you have to be a recognized contributor or something before you can find your way in there 22:12 < ayecee> otherwise it would just redirect you to ##linux 22:12 < birdbolt1> wait... arent they different? 22:12 < kerframil> birdbolt1: here you go: https://freenode.net/kb/answer/namespaces 22:12 < jim> (Torvalds being the author and trademark (on "linux") holder 22:12 < jim> ) 22:12 < ayecee> they're different in that there's one more #, yes 22:12 < ayecee> #linux is a placeholder channel only 22:12 < birdbolt1> ohh interesting bit about irc i didnt know. Thanks kerframil 22:13 < kerframil> birdbolt1: well, freenode - not irc at large 22:13 < birdbolt1> ohh okay i see 22:14 < jim> #channels are on-topic project channels with the owner of the project involved 22:14 < jim> ##channels are "about" channels, like, this channel is "about" linux 22:14 < louisdk> Doing some mysql packages updates with apt I get this "Failed to connect to database: Access denied for user 'root'@'louis-server.lan' (using password: YES) at -e line 5, <> line 1." I just can't figure out where it gets louis-server.lan from and how I can change it to localhost. Looking in /etc/mysql/ didn't gave any clues. 22:15 < ayecee> louisdk: /etc/mysql/debian.cnf, probably 22:15 < ayecee> i think that's what package updates use for credentials when working with the database. 22:16 < ayecee> though usually the user is debian-sys-maint, so maybe it's not that. 22:20 < kerframil> the louis-server.lan part is probably the result of a rerverse DNS lookup, as applies to those who haven't cottoned on to the "skip_name_resolve" setting 22:22 < graff> how do i make fortune run evertytime anyone logs in to an ssh server? 22:23 < bls> graff: you can't 22:23 < graff> or is this potentially a bad idea to have so much crap 22:23 < bls> graff: you can try adding it to /etc/profile.d/ but people are free to override the settings there 22:23 < graff> i just want to sort of convey the unix experience, sort of like i am used to it from the olden days 22:23 < graff> but I don't want to mess it up 22:23 < graff> bls: ah so just the shell rc file? that makes sense 22:24 < graff> hmm 22:24 < graff> btw, while I am here. what is the proper practice for upgrading debian 9? 22:24 < graff> they said -u upgrade on #debian. but that seemed sort of weird 22:25 < graff> i don' recall anything about -u from back in the day. 22:25 < qrvpzvb> can runuser take UIDs? it seems to me that it requires valid usernames only 22:25 < ayecee> the infamous, cryptic "they" 22:27 < pol0> hi , when im trying to open any video there is sound 22:29 < boblamont> do I have to install cups to print/merge to file labels in openoffice? 22:32 < bls> what do you mean by "print/merge to file labels"? 22:34 < boblamont> the options are to print or to print to file, both of which do mail merge (pull data from the database into the label template) 22:34 < bls> it technically should be able to print to file without the cups daemon installed 22:36 < bls> was going to suggest asking its channel, but pretty sure that project is dead 22:36 < pol0> i can see the sound bar 22:37 < boblamont> yeah, I tried there, seems dead 22:37 < pol0> but its not working 22:37 < bub_> 4 22:37 < bub_> whops.. 22:37 < bls> pol0: tried asking the channel for your video player? 22:38 < pol0> bls im using linux 22:39 < pol0> and its audio not video 22:39 < pol0> it was working before 22:45 < pol0> no support? 22:45 < V7> Whata ... https://www.wish.com/feed/tag_53dc186421a86318bdc87f20/product/5ae098f16bcfa31c870bb50c 22:46 < jprjr> pol0: honestly that's the kind of thing that's hard to troubleshoot over IRC because there's a lot of details involved 22:46 < jprjr> pol0: what kind of sound hardware you've got, what distro your own, what version of the kernel being used, are you using pulseaudio (probably) or just ALSA (unlikely), what player is it, etc etc 22:48 < pol0> jprjr, im using kali linux , when i open setting i found speaker - built-in audio 22:49 < pol0> jprjr, look when i open youtube or something there is voice 22:49 < pol0> jprjr, sorry when i open youtube there is no voice 22:49 < pol0> no sound* 22:49 < jprjr> pol0: right off the bat, you're using Kali Linux, which nobody should use for a day-to-day distro. I'm not going to help just because of that, being brutally honest. 22:50 < kminor> heya guys, I can't remove cpp-4.1-doc package. It's not pinned or on hold... it just wont remove 22:50 < jprjr> If you get the same problem on like, any other distro, then we can talk. But I don't help with Kali 22:51 < kminor> started when dist-upgrading from debian wheezy to jessie ... but those guys are kinda terds to ask so figured I'd try you guys first 22:52 < birkoff> in old times did people only have screens that occupied TTYs of one linux server and actual machine? 22:54 < jprjr> birkoff: depends on how far back you go I guess 22:55 < jprjr> It would also depend on who's running the server right, like even today you'll have people setup a "server" who do everything via physical console just because they don't know/get ssh very well 22:55 < orbisvicis> I want to make sure I've linked the right version of a library, can gdb help with this ? 22:56 < jprjr> orbisvicis: I think you can do it with readelf or objdump 22:57 < jprjr> orbisvicis: ldd is also a quick-and-easy way to show what libraries you'll be using, but that's assuming you're not dealing with cross-compilation 22:58 < birkoff> as far as the linux TTY go jprjr 22:58 < jprjr> birkoff: I'm not 100% sure what you're really asking here 22:59 < birkoff> this is why there's alt+F1 to alt+f7/8 22:59 < birkoff> i read somewhere once that in the past people would use those ttys on different computers 22:59 < birkoff> i think 23:01 < jprjr> Yeah kinda. The default virtual terminal on Linux only works with one seat. 23:01 < memphisto> terminals 23:01 < memphisto> dumb computers 23:01 < jprjr> But I'm pretty sure you can setup multiple physical terminals on serial ports and whatnot 23:02 < jprjr> I've never had a need to do that so I don't know how to go about it 23:03 < bls> yes, you can. used to do it in smaller labs where we didn't have a serial terminal server...just connect two machines with crossover cables, fire up a getty on one serial port on each 23:03 < birkoff> serial ports connected to where ? 23:03 < birkoff> and communicating how ? 23:04 < bls> you can connect physical terminals or you can connect other computers, and they usually communicate using RS-232 23:04 < hexnewbie> TTY literally means a serial port connected to a TeleTYpe. That's why on Linux (where tty[0-9]* stands for virtual consoles), the serial ports are ttyS[0-3] 23:04 < kerframil> to each other i.e. with a null modem cable 23:05 < bls> in the past, it meant adding a getty line for ttyS# to /etc/inittab, then running minicom or cu on the other end 23:06 < birkoff> so the other teletypes had to have some sort of computer with its own OS ? 23:06 < bls> no 23:06 < jim> and, serial ports have been known to connect to other things, like different kinds of phyaical terminals, and modems and such 23:06 < bls> just enough hardware to interpret the serial comms and the terminal codes 23:06 < bls> like: https://en.wikipedia.org/wiki/VT100 23:07 < birkoff> so multiple teleypes without cpu/ram and harddisks would connect to one computer 23:07 < bls> birkoff: correct, they were essentially just a screen and a keyboard 23:07 < birkoff> which had those 23:07 < birkoff> that's cool 23:07 < bls> you'd have a mainfram/big iron system with lots of serial ports, then you had a bunch of dumb terminals connected to it from which people did their work 23:08 < memphisto> history class 23:08 < Haled> Is it wrong to expand a primary Linux partition to the left? 23:08 < hexnewbie> birkoff: An original teletype, before glass terminals: https://www.youtube.com/watch?v=MikoF6KZjm0 23:08 < balletjebal> at amazing speeds :P 23:08 < birkoff> it seems like an easy way to divide computer use to people, instead of like today having to buy several licenses for OS and multiple complete computer machines 23:09 < birkoff> Haled depends if the mBR sits on that HD 23:09 < Haled> birkoff, it does 23:09 < birkoff> can one do something simliar today with like a keyboard a screen and i dont know, some way link them to one machine running linux ? 23:09 < bls> although the modern equivalents usually have more substantial local memory and use netbooting 23:10 < birkoff> pxe something loading an OS from the net? 23:10 < memphisto> you could use thinclients today 23:10 < bls> you could, and although it was neat for the era, it's extremely ugly and limiting given the technology available these days 23:11 < jprjr> I think it's possible to run Xserver in a multiseat setup 23:11 < bls> birkoff: yes, pxe boot a kernel, then load the rest of the OS over NFS 23:11 < jstvz> hexnewbie: thanks for that, never seen one before 23:11 < Haled> birkoff, I can't delete the leftmost partition because the hdd has the mbr there?! why? 23:11 < Haled> I don't mind re-installing grub if necessary or whatnot 23:11 < bls> Haled: what does "left" even mean in this context? 23:11 < memphisto> but if you look at it like that you have full blown PC that boots over network 23:11 < birkoff> bls how is it limited with todays technology and not improved? 23:11 < Haled> or is it worse than that... 23:11 < memphisto> thin client is like small pc that remotes to Server 23:12 < birkoff> Haled but grub needs to sit on the first sector of the drive 23:12 < birkoff> first 512 bytes IIRC 23:12 < hexnewbie> Haled: Non-orthodox resizing is more difficult and risky, which is why people choose LVM where it's not an issue 23:12 < jprjr> Haled: I assume birkoff was trying to avoid having to re-install grub? So long as you leave some empty space for it you should be fine 23:12 < Haled> bls, means closer to the beginning of the drive 23:12 < birkoff> but for a pxe i still need a motherboard and ram 23:13 < jprjr> Haled: my first question is why, like what are you trying to actually do? 23:13 < bls> birkoff: we now have accelerated, bitmapped graphics, a full TCP stack, interfaces that aren't arbirarily designed for terminal vendor lock-in 23:13 < birkoff> Haled show us the whole partition table 23:13 < Haled> jprjr, I have a machine with NTFS (windows), swap, Ext4 (Linux). Decided I don't want the windows and I want its space 23:13 < birkoff> bls rs232 was designed for vendor lockin ? 23:14 < bls> birkoff: no, but the terminal control protocol on top of it was 23:14 < birkoff> Haled then yo uwould lose swap too i think 23:14 < birkoff> bls what was its name? 23:14 < Haled> birkoff, you make it sound so naughty 23:14 < Haled> there you go: https://pastebin.com/W8rDCw03 23:14 < jprjr> Haled: yeah you should be able to delete the windows partition no problem. I don't know if you listed your partitions in order or whatever. Also depends on if anything is a physical vs logical partition 23:14 < bls> birkoff: read the history of the termcap file and terminfo database for some of the details, or check out the terminal chaper in the Unix Haters Guide 23:14 < Haled> all physical 23:15 < Haled> I can unlink the swap, obviously 23:15 < Haled> sudo swapoff ... 23:15 < jprjr> Haled: See you didn't specify all this earlier, and I assume people thought you were "expanding left" into free sectors that were being used by grub 23:15 < birkoff> then like I said you will lose swap and possibly the boot loader 23:15 < birkoff> those are not fatal 23:16 < jprjr> Haled: But yeah you should be able to trash /dev/sda1 no problem, maybe move /dev/sda2 to the same start sector that /dev/sda1 was at, then expand /dev/sda3 to fill the space, badda-bing badda-boom 23:16 < Haled> jprjr, thanks 23:16 < bls> birkoff: Chapter 6 of http://web.mit.edu/~simsong/www/ugh.pdf 23:16 < Haled> birkoff, thanks 23:17 < kurahaupo> A swap partition may still be useful if you want to "suspend to disk" 23:17 < jprjr> Alternatively, just resize /dev/sda1 to like 500MB or something, reformat it, make it your new /boot partition 23:17 < Haled> jprjr, for speed or something? 23:17 < jprjr> Then move /dev/sda2, expand /dev/sda3, then you're not stuck with weirdly-numbered partitions that start at 2 23:17 < V7> Hey all 23:18 < Haled> because it's the beginning of the disk? 23:18 < V7> Could anyone suggest a linux without cx8 and cmox CPU instructions ? 23:18 < jprjr> Nah I just like having a separate /boot partition, always have 23:18 < V7> cmov * 23:18 < birkoff> 500mb is overkill for /boot not? 23:18 < jprjr> It's probably not really necessary tbh 23:18 < ayecee> that would be a distribution that supports i586 processors 23:19 < jprjr> birkoff: Well that's why I said "or something". I've filled up 100MB boot partitions before so I tend to stick with like, 500MB 23:19 < jprjr> It's virtually nothing in terms of total disk space so who care 23:19 < kurahaupo> I like to have at least two root partitions, with /home & /var separate, so that I can upgrade & roll back. 23:20 < jprjr> I do think having a separate /boot is needed when you get into setting up full-disk encryption, root on LVM, etc 23:20 < jprjr> Key phrase "think", I'd have to research specifics 23:21 < Haled> jprjr, stack exchange agrees with you https://unix.stackexchange.com/questions/256/is-it-good-to-make-a-separate-partition-for-boot 23:22 < kurahaupo> jprjr: /boot must not be encrypted; whether it can be managed through LVM I haven't checked lately, but the answer used to be "no" 23:23 < Namarrgon> when you use a recent version of grub then /boot can be encrypted, but then you might not want to split out /boot in the first place 23:23 < bls> I've been seeing reports of encrypted /boot also taking minutes to decrypt 23:25 < Namarrgon> if you run it on a potato, sure 23:25 < Namarrgon> the key derivation count makes it "slow", if you don't have a fast cpu then pick a lower count 23:26 < searedvandal> I did the whole encrypt everything for a while, and I can't say that it took any significant amount of time to decrypt boot or root 23:27 < bls> not saying it's an issue, just some people obsess over the 30 seconds - 2 minutes of booting they do compared to the hours they spend using the computer once booted 23:28 < searedvandal> sure, if you boot multiple times a day it can be annoying, I guess. 23:31 < V7> ayecee: Thank you, but ubuntu-14.04-server-i386 tells that it needs same instructions 23:31 < ayecee> V7: despite the name, that's for i686 and above 23:32 < V7> oh dear 23:32 < V7> Whata a confuse 23:32 < V7> Thank you very much ayecee 23:34 < V7> By the way, is there any was to know which instrunctions needs each kernel ? 23:35 < bls> i386 is the traditional name used for 32-bit x86. some distros started using i486 or i586 as old i386/486 cpu features got dropped from the kernel, but not all 23:37 < V7> birkoff: So, is there ay possibility to get a set of instruction of specific kernel ? 23:37 < V7> For example, for Kernel 3.2 ? 23:37 < V7> Does it need pae, cx8, cmov ? 23:38 < birkoff> what? 23:39 < V7> I mean, what's the way to know if this kernel/OS will work on specific CPU 23:39 < RayTracer> you try to run it 23:39 < birkoff> you must have meant addressing bls , V7 23:39 < V7> For example, ayecee told that i586 doesn't need cx8 and cmov instructions 23:39 < V7> birkoff:, oh, sorry 23:40 < bls> the kernel will include support for all the instructions it can leverage, and will use other instructions/workarounds when they're absent 23:41 < bls> it's when those workarounds start getting dropped that old CPUs lose support 23:42 < V7> bls: Yes, but why then I can't install OS which needs cx8 and cmov instruction ? 23:42 < V7> instructions * 23:42 < V7> There should be a way to check which kernel needs correct set of instructions 23:45 < kexec> hello please for christ sake, how to get cron running on windows bash? i keep the bash instance opened, cron service is running, user is in crontab group 23:45 < twainwek> kexec: sorry but i gotta say this, the easiest way is to stop using windows 23:46 < bls> kexec: going to need to ask a windows or wsl related channel 23:47 < kexec> yeah i am aware of that :D but still id like to fix it 23:48 < bls> the kernel calls out which architectures it supports. it doesn't call out the individual instructions introduced/dropped from those architectures 23:50 < bls> so the lack of CMOV means it at least requires i586 support, which some distros have dropped 23:50 < V7> bls: So there's no way to get if it needs cx8 instruction or not ? 23:50 < V7> Oh, the instruction is not for kernel, but for distro ? 23:50 < birkoff> kexec whats the problem ? 23:51 < V7> bls: How did you know that CMOV for i586 ? 23:51 < kexec> birkoff: it a wget cron but it does nothing 23:51 < birkoff> kexec in the last windows update you could run background tasks without a terminal open 23:51 < bls> different distros can apply/remove patches and config items as they see fit, so some drop old processor support to slim down their kernel 23:51 < birkoff> kexec show us your crontab 23:51 < bls> V7: https://askubuntu.com/questions/5807/what-is-the-significance-of-the-cmov-cpu-instruction 23:52 < V7> bls: Oh, so they can modify kernel and remove staf from them and it will be called as custom kernel ? 23:52 < bls> V7: correct 23:52 < V7> So, kernel supports all instructions by default ? 23:53 < kexec> birkoff: 23:53 < kexec> * * * * * /usr/bin/wget http://.pdf --output-document=/home/user/document_`date +%Y-%m-%d_%H:%M:%S`.pdf 23:53 < bls> no, the vanilla kernel can also pick and choose, they've recent dropped i386 support 23:53 < bls> hehe, recently as in 6 years ago 23:54 < V7> Time goes fast 23:54 < birkoff> i thought linux was all about supporting weird or out of date hardware 23:55 < V7> bls: IS there a way to check which hardware support which kernel ? 23:55 < triceratux> linux still supports plenty of weird or outofdate hardware 23:55 < V7> I mean, each kernel * 23:55 < bls> it is, until you reach the point where it's impossible to run the current system on such hardware, then you drop it from the mainline to save time and effort. people will fork from that point if they still want to keep it working 23:55 < birkoff> kexec you would have to wait for someone who is more proficient in cron than me, meanwhile try to see if there actually is a cron instance running in the background and monitor your logs to see if it is executed and what is reported 23:56 < benjwadams> Is there any way to view what sort of things are stored in swap? 23:56 < bls> this likely isn't a cron issue and is more of a WSL issue 23:56 < birkoff> you cant know for sure bls 23:56 < benjwadams> I'm trying to figure out why swap usage is so high 23:56 < birkoff> although I had this idea about a Freenode WSL channel for quite some time 23:56 < V7> benjwadams: Doesn't it mean a RAM ? 23:56 < bls> no we can't, because we have no clue how to debug/triage things like this on windows 23:57 < benjwadams> I guess i could use dd :) 23:57 < V7> Check in "htop" and sort by "RAM" 23:57 < simbalion> I seem to have done something that has made all my F# keys no longer usable for their functions, however the keys themselves are still working? 23:58 < bls> simbalion: does your keyboard/bios have the Fn modifier key or Fn-lock? 23:58 < simbalion> bls: it's an apple keyboard, I'm still learning how it works but it does have an Fn key 23:58 < jada> hey, I'm looking for a good cheap linux laptop only for skype/ssh/web. I'm finding some x230 thinkpads on craigslist in $150-$200 price range, do you people know of some alternatives? Maybe some chromebooks? 23:59 < jada> I wouldn't mind eMMC and/or AMD/ARM cpu 23:59 < kexec> birkoff: i can see cron in "ps -A". and i couldnt find any logs, "/var/log/syslog" doesnt exist 23:59 < benjwadams> V7: that's not what i want 23:59 < bls> simbalion: ah, in macOS you there's a setting for whether the top keys are the control keys or function keys by default, then you need to hold Fn down to get the other behavior --- Log closed Wed Jun 20 00:00:00 2018