--- Log opened Tue Apr 03 09:00:42 2018 09:02 < notmike> You're right. That was uncalled for. I'll ban myself now. 09:02 < jim> if there's even one woman here, you just made her feel like crawling under a rock for about 10 years 09:02 < jim> so... 09:02 < jim> be -careful- 09:04 < xdije> hi 09:04 < aaa_> hi 09:05 < xdije> after i finished rsync a huge data to new location, if i run the rsync again with -o in addition will this cause the data to sync again or only the owner attribute will be sync 09:07 < jim> I'm pretty sure some kinds of rsync will compare the file hashes 09:09 < madghost> https://ibb.co/bPgh6c 09:09 < madghost> that is my example :( 09:09 < madghost> why doesn't work ? 09:10 < madghost> windows has 192.168.1.3 ip address 09:17 < jim> I guess 0.0.0.0 is probably special 09:20 < jim> we should probably try again... I'll be right back 09:20 < madghost> route add default gw 192.168.2.1 get me SIOCADDRT: Network is unreachable as well :(((((( 09:29 < jim> can you run ... route -n > ~/route-listing 09:36 < jim> to add this default route: route add default gw 192.168.2.1, you need already a net route like this: route add -net 192.168.2.0/24 09:36 < jim> in other words a net that the gateway is on 09:37 < jim> it's actually -two- routes, the net route and the default route 09:37 < madghost> I did route add 192.168.2.0/24 and ping 192.168.2.1 works 09:38 < jim> ok, now try to add the default route (now that you can reach the gw) 09:38 < madghost> but when I try to do: route add default gw 192.168.2.1 it gives me SIOCADDRT: Network is unreachable 09:39 < jim> ok, do this... route del 192.168.2.0/24 09:39 < jim> then route add -net 192.168.2.0/24 09:40 < jim> maybe you have to add the netmask separately like in the old days 09:41 < jim> but let's try this for now 09:41 < jim> wait, you're on the net now... on a different machine? 09:41 < madghost> I can't ((((( 09:42 < madghost> cause this computer is other city, and then I do it, I don't want to go there :(((( 09:42 < jim> when will be tne next time you go there? 09:43 < madghost> I hope not soon 09:43 < madghost> ))) 09:43 < jim> or is it just too far? 09:43 < madghost> yep ( 09:43 < jim> ok 09:44 < jim> what had you want to do all this? 09:45 < madghost> I don't know (( I will read about routing in Linux, and maybe something find. Thanks for time on me 09:46 < jim> you'll want to learn ip route command 09:46 < madghost> I emulate this problem on other computer, and will find a solution 09:46 < madghost> thanks 09:47 < jbit> madghost: the network route needs to be link local 09:47 < flipper887> It took a month but using Lubuntu 16.04.4 on the 4.4.0-x series of kernels i was able to get thoes amdgpu-pro drivers up and running. 09:47 < jbit> sounds like it's not, and that's why it thinks hte address is unreachable 09:48 < jbit> madghost: you need to do like "route add 192.168.2.0/24 dev eth0" (or whatever your ethernet adapter is) 09:49 < naim42> madghost: what's with the ((((( ? 09:49 < madghost> I did it, and my network thru dev enp10s1 09:50 < madghost> 192.168.2.0 192.168.1.254 255.255.255.0 UG 0 0 0 enp10s1 09:51 < madghost> 64 bytes from 192.168.2.1: icmp_seq=1 ttl=59 time=0.912 ms 09:51 < jbit> madghost: the "G" means it's a global route, and not link local 09:52 < jbit> madghost: sorry, that's slightly misleading... the "G" means it uses the gateway 09:56 < jim> so it's working now? 09:56 < jim> maybe there's a bad route or something unfortunate in iptables 10:02 < okamis_> I dont understand readlink command. E.g. the -f flag, refers to component, and all but the last component must exist. What is a component? 10:03 < the_drow> How do I connect two computers with an Ethernet cable (without a router) and ensure that a route between the two computers is established automatically? 10:04 < MrElendig> okamis_: a part of a path 10:04 < MrElendig> the_drow: set up a profile in your network manager 10:04 < autopsy> the_drow, you need a poatch cable to do that it is specific. 10:04 < MrElendig> autopsy: no 10:04 < autopsy> MrElendig, no patch cable? 10:05 < okamis_> MrElendig: Last component, would that be the filename? 10:05 < MrElendig> the_drow: or run dhcp server on one of the machones 10:05 < MrElendig> autopsy: you need a ethernet cable of some kind, and I suspect that you are confusing patch with crossover 10:05 < autopsy> MrElendig, if you don;t use a patch cable the connectors are reversed so it won't work ethernet to ethernet. 10:05 < MrElendig> autopsy: there is nothing special about a patch cable, except that they are usually stranded 10:05 < the_drow> MrElendig, How do I do that? 10:06 < autopsy> MrElendig, yeah a crossover cable sorry. 10:06 < the_drow> autopsy, I know 10:06 < MrElendig> autopsy: that is a crossover cable, not a patch cable, and any hardware made in the last decade as auto-mdi/mdix 10:06 < MrElendig> autopsy: all gbit cards has it because it is required in the standard 10:06 < the_drow> MrElendig, also known 10:06 < MrElendig> so not corssover cable needed in the wast majority of cases 10:06 < autopsy> MrElendig, ok you're right I'm referring to old cables. 10:06 < aloy> okamis_: the last component in foo/bar/baz will be 'baz', with -f, the path foo/bar must exist, baz doesnt have to exist 10:07 < MrElendig> autopsy: not old cables no, it was not due to the cables, it was due to the electronic at each end 10:07 < MrElendig> autopsy: some devices had a physical switch to switch between mdi and mdi-x btw 10:07 < the_drow> How do I configure the NetworkManager to create the route automatically? 10:07 < rendar> rasberry freezed, i have booted it again and now my external ext4 driver can't be mounted: journal is corrupted...is ext4 so fragile?! it's a power black out enough to destroy a ext4 fs?! 10:08 < the_drow> I can find info about using ip ad add but not with network manager 10:08 < autopsy> the_drow, can't you set it in /etc/sysconfig/network-scripts/ifcfg-eth0 as GATEWAY=192.168.43.1 ? 10:08 < the_drow> I'm not sure 10:08 < the_drow> The thing is, sometimes the ethernet card is connected to a router and something it isn't 10:08 < MrElendig> rendar: no filesystem likes unclean shutdowns 10:09 < rendar> MrElendig i know, but in +30years of NTFS, FAT32 and FAT16 it never happened! 10:09 < MrElendig> rendar: and you just don't notice all the times it have happened where there wasn't any journal corruption afterwards 10:09 < MrElendig> rendar: lol 10:09 < the_drow> The computer in question has two network cards, one that goes into the router and the other to another computer. We have no way to know which is connected to which in a customer's deployment 10:09 < autopsy> the_drow, well if you're not using DHCP you can set the default gateway with route command. 10:09 < MrElendig> rendar: fat doesn't have a journal, there you just get data corruption with no aparent errors instead 10:09 < MrElendig> :p 10:09 < aaa_> dns over dhcp 10:10 < MrElendig> and ntfs likes to blow up too 10:10 < peetaur2> fat is so modern...glad to see it used in EFI 10:10 < MrElendig> atleast it is not riddled with patents 10:10 < the_drow> Please explain this to me like I'm 5 10:11 < the_drow> Because I've done some stuff but not this 10:11 < autopsy> the_drow, that just sounds silly why are they unplugging the cables and mismatching them afterwards? 10:11 < peetaur2> rendar: oh...we're talking about raspi? is it an SD card? rapsi is notoroious for corrupting things due to hardware, not software's fault.... use a card on this list https://elinux.org/RPi_SD_cards 10:12 < autopsy> the_drow, use route add default gw 192.168.0.1 10:12 < MrElendig> you can use arp do somewhat reliable detect what is at the other end 10:12 < peetaur2> s/notoroious/notorious/ 10:12 < MrElendig> but this is a "why are you doing this in the first place?" kind of thing 10:12 < autopsy> MrElendig, why is the customer mis matching the ethernet cables though? 10:13 < autopsy> the_drow, they should know to keep them plugged into the right RJ-45 port. 10:13 < the_drow> Ok so one computer is connected to the ethernet, the other is isolated from the intranet. The other is also connected to instruments measuring some stuff and it transmits them over LAN 10:14 < the_drow> autopsy, We can't tell how they will deploy the product 10:14 < autopsy> the_drow, I see. 10:14 < autopsy> the_drow, what instruments? 10:14 < NGC3982> i am about to build a live feed for my web site. i have an old smart phone i will use for this, and some kind of scheduling. my initial solution is to build a tasker script that takes a picture every minute, which in turn uploads it and replaces a picture shown in an automatically updated html document. im doing this on my standard apache server. any suggestions on how to do this "more properly", or is 10:14 < NGC3982> this actually a good solution? 10:14 < the_drow> autopsy, Naval sensors 10:15 < the_drow> It is also likely a customer will connect a sensor directly to a computer with an Ethernet cable 10:16 < autopsy> the_drow, your sensors have RJ-45 jacks? 10:17 < autopsy> NGC3982, doesn't sound like a feed of video. 10:17 < the_drow> autopsy, yup 10:17 < autopsy> NGC3982, Every minute is kinda long for it to be a feed. 10:17 < the_drow> Sometimes only them 10:18 < autopsy> the_drow, so what are you asking here is the real question the ethernet cables might be plugged into the wrong port for your deployment? 10:19 < autopsy> the_drow, routers can do DHCP or static addressing. 10:19 < the_drow> They might be connected to any port that is available or even disconnected from the Intranet in order to connect it to a sensor 10:19 < the_drow> autopsy, But they are unwilling to connect these sensors to routers 10:19 < the_drow> Oh don't ask why 10:20 < MrElendig> wouldn't need a router, just some controller which runs dhcp 10:20 < MrElendig> and it would make your life a lot simpler 10:20 < MrElendig> sidenote: this is less of a issue if everything supports ipv6 slaac 10:22 < autopsy> the_drow, you need to use a dhclient script maybe. 10:22 < the_drow> MrElendig, So if I have dnsmasq can I set it up? 10:22 < MrElendig> sure 10:22 < MrElendig> and have all the sensros send a id/hostname in the dhcp request, so that you can automagically do dns 10:22 < MrElendig> or use mdns 10:23 < MrElendig> because having a hardcoded list of ip addresses is silly 10:24 < autopsy> MrElendig, maybe he doesn't want them to change ever. 10:24 < the_drow> MrElendig, not all sensors have a hostname. Some are just FPGAs 10:25 < MrElendig> autopsy: hardcoded lists works until the day someone makes a single mistake and then the oven catches fire 10:25 < MrElendig> been there, seen that (and spent two weeks redoing a lot of cabeling due to it) 10:25 < the_drow> Anything can connect to our product 10:26 < the_drow> But we need to set up a computer to another computer LAN first. I don't know the hostnames in advance 10:26 < autopsy> MrElendig, yeah but it's only two RJ-45 jacks he's worried about. 10:27 < autopsy> the_drow, you can use iptables and forwarding to reach the internet from the LAN computer. 10:27 < NGC3982> autopsy: true. still, i dont think i need more than every minute. its for monitoring plants. 10:27 < s10gopal> how to disable "hwclock --systohc" during shutdown ? 10:27 < autopsy> the_drow, are you trying to reach the internet from both computers? 10:27 < NGC3982> autopsy: still, sounds like a reasonable idea, or a botch job? :/ 10:27 < autopsy> s10gopal, is your distribution using systemd? 10:27 < NGC3982> i do not really know what services exist for this already. 10:27 < s10gopal> autopsy, no upstart 10:29 < autopsy> s10gopal, it's in /sbin/init I think when the shutdown proceedure is going. Check if /sbin/init is a text file script or a binary file. 10:29 < s10gopal> autopsy, thx for help , bugzbunny is helping me on #ubuntu 10:29 < autopsy> s10gopal, did you get it fixed yet? 10:29 < s10gopal> autopsy, no 10:30 < autopsy> s10gopal, on your machine you should have /sbin/init 10:31 < autopsy> s10gopal, upstart SYSV init and systemd all differ across the many available distributions. 10:31 < autopsy> s10gopal, or /etc/rc.d you should have also. 10:32 < BCMM> autopsy: curious - on which distros is init a text file? 10:32 < the_drow> autopsy, no, I'm trying to reach the intranet only from one 10:32 < BCMM> (also, missing option - on Debian Sid, by default, it's a symbolic link to /lib/systemd/systemd) 10:32 < autopsy> BCMM, on Fedora's old LiveCD initramfs it is a nash shelll script. 10:32 < s10gopal> autopsy, content of /sbin http://paste.ubuntu.com/p/3H7WyfYVB4/ 10:33 < BCMM> autopsy: presumably one that does a few weird liveCD things, and then execs real init? 10:33 < autopsy> BCMM, yes it does pivot_root $newroot /sbin/init 10:34 < s10gopal> autopsy, and init is binary file 10:34 < BCMM> autopsy: oh, i missed that you said initramfs. that's not an unusual configuration for an initramfs 10:34 < autopsy> s10gopal, ok so forget that approach. 10:34 < s10gopal> autopsy, #ubuntu also dont know about upstart 10:34 < s10gopal> s10gopal: i have no idea; maybe /etc/default/hwclock (it overrides some settings found in hwclock.sh 10:36 < autopsy> s10gopal, nah you need to find the call to the binary in your initialization system The one that controls what happens on shutdown or Runlevel 0 and runlevel 6 10:36 < s10gopal> autopsy, how ? 10:38 < BCMM> i'm sorry, but the idea that it's normal for that to be called directly by /sbin/init is just wrong 10:38 < BCMM> it will be an init script (service) that does it 10:39 < BCMM> and /etc/default/hwclock *should* modify the behaviour of that script (but it's also *possible* that that file is just a holdover from debian and the ubuntu script doesn't actually check it) 10:40 < Ben64> oh man this is annoying 10:40 < Ben64> sometime between when i learned how to use inkscape and now, they decided to remove Object -> Objects 10:40 < Ben64> now i don't know where to see a list of objects 10:41 < BCMM> Ben64: what version? 10:41 < Ben64> 0.91 10:42 < s10gopal> autopsy, BCMM, can i install systemd on ubuntu 14.04? 10:42 < the_drow_> MrElendig, Is there a tutorial or something I can read on how to do it? 10:42 < the_drow_> autopsy, ^ 10:44 < BCMM> s10gopal: you *can*, you probably *shouldn't*. stuff will break. 10:45 < BCMM> s10gopal: any particular reason for not running a newer ubuntu? 10:45 < BCMM> s10gopal: you've still got support for another year on 14.04, but there's been a new LTS out for a while 10:46 < s10gopal> BCMM, https://bugzilla.kernel.org/show_bug.cgi?id=198665 , if u use old kernel on ubuntu 16.04 laptop becomes super laggy 10:46 < BCMM> s10gopal: well, there's a *new* new LTS coming this month 10:47 < BCMM> so i guess that might solve your problem 10:47 < s10gopal> i dont happen upto kernel 4.12 10:47 < s10gopal> 4.13 onwards it happen 10:48 < autopsy> s10gopal, I am a Fedora user I don't know about upstart either. 10:50 < s10gopal> autopsy, plz for giving your precious time 10:50 < autopsy> s10gopal, I have to check google for upstart on shutdown proceedures I don't have Ubuntu. 10:50 < s10gopal> http://paste.ubuntu.com/p/GMP8PMKxFB/ 10:50 < autopsy> s10gopal, why do you want to disable hwclock --systohc anyways? 10:51 < BCMM> that's a good question 10:51 < s10gopal> https://bugzilla.kernel.org/show_bug.cgi?id=198665 10:51 < s10gopal> a developer told me 10:52 < autopsy> s10gopal, ahh you found it that looks like it right there what file is that actually? 10:52 < autopsy> s10gopal, what is the full path name? 10:53 < s10gopal> i did cat /etc/init/hwclock-save.conf | pastebinit 10:56 < autopsy> s10gopal, so everything is in /etc/init now that you found it edit it in a text editor like gedit and comment out that line #exec hwclock --rtc=/dev/rtc0 --systohc $tz --noadjfile $badyear After you put in a pound sign before the call to exec hwclock save the file and reboot to test it. 10:57 < s10gopal> autopsy, thx 10:58 < autopsy> s10gopal, you can use nano or vim too to edit files of type text. 11:05 < s10gopal> autopsy, done , how i can verify that it is disabled on shutdown? 11:13 < autopsy> s10gopal, switch to runlevel 0 or runlevel 6 using /sbin/init 6 11:14 < autopsy> Runlevel 0 is a poweroff and runlevel 6 is a reboot. 11:16 < s10gopal> autopsy, can you please guide me how i can learn linux kernel and driver development ? from where i should start ? i only know very basic linux commands and little bit c programming 11:19 < autopsy> s10gopal, I've been reading the manual pages on system calls in C 11:19 < autopsy> s10gopal, you want to develop drivers in C? 11:19 < s10gopal> autopsy, yes 11:20 < kurahaupo> s10gopal: what sort of things do you want to write drivers for? 11:20 < s10gopal> kurahaupo, for embeded devices 11:21 < milp_2> hi, could anyone give me advice how to tell openvpn not to route intranet-subnets through vpn? As soon as i connect my office machine to vpn i cant access any of the intranet-subnets anymore 11:21 < kurahaupo> Lots of devices already have drivers; what do you expect doesn't? 11:22 < kurahaupo> milp_2: use different subnets at each end. If that's not possible, use 2-sided NAT 11:23 < s10gopal> kurahaupo, still new devices are coming 11:24 < milp_2> kurahaupo: this is a normal anonymity-vpn provider, nothing on my own. cant i just tell it to ignore all internal requests and only work as gateway for ther est? 11:24 < haventermianl> hi, someony can help with alipne and web.de? How to show all folders? 11:24 < kurahaupo> milp_2: what does "ip route show table all" look like before and after connecting the vpn? 11:25 < kurahaupo> milp_2: this should be the default behaviour so first we need to figure why it's going wrong 11:26 < peetaur2> s10gopal: http://kernelnewbies.org/ 11:26 < s10gopal> peetaur2, thx 11:26 < kurahaupo> milp_2: and "ip -o addr show" for good measure 11:54 < kuri0> Do applications ever store sensitive data in /usr ? 11:55 < autopsy> kuri0, sensitive in what manner? 11:55 < kuri0> autopsy, ssh keys passwords and stuff like that 11:55 < sauvin> And what applications? 11:56 < sauvin> Mostly, applications that *users* run can't do that. 11:57 < autopsy> kuri0, oh hell no. 11:57 < kuri0> autopsy, so its safe to run an untrusted application in sandbox giving it read only access to usr ? 11:57 < kuri0> *assuming the sandbox is secure* 11:58 < autopsy> kuri0, are you afraid of someone hacking the machine? 11:58 < kuri0> autopsy, i use a sandbox for untrusted stuff from github 11:59 < autopsy> kuri0, what is an untrusted application? 11:59 < kuri0> autopsy, application which could be malicous 11:59 < junka> what is a sandbox 11:59 < autopsy> junka, like a jail. 12:00 < junka> does it have a name? 12:00 < autopsy> kuri0, I'd check the source code of any untrusted application first. 12:00 < kuri0> autopsy, what about if it has no source ? 12:00 < kuri0> junka, i use bubblewrap 12:00 < junka> oh 12:00 < autopsy> kuri0, then you aren't providing it under GPL like the kernel. 12:01 < sauvin> Untrusted stuff like that, run it in a VM that has no net access. 12:04 < junka> kuri0; why not firejail 12:04 < kuri0> idk 12:04 < kuri0> i just used bubblewrap 12:04 < kuri0> is firejail better ? 12:05 < sine0> I tried to migrate a client to another host for ecommerce, alot of hte filenames are in chinese for the pictures. when I scp/ftp the files over they become url/unicoded and the file urls are invalid 12:05 < junka> kuri0; seems easier 12:05 < sine0> I hate tried testing this with local box and the same happens, so what gives and how can I fix this? would it be down to the host locales and therfor I would be limited in what the would enable ? 12:06 < kuri0> junka, bubblewrap seems to allow more configuration too 12:08 < kuri0> junka, also firejail is a suid binary which requires root ? 12:09 < kuri0> i installed bubblewrap from source without needing root 12:10 < kurahaupo> sine0: scp just copies the bytes of the filenames; check that you have the same locale settings at both ends 12:11 < sine0> kurahaupo: sure, so I guess if its a paid host, I would not have control over that, right ? 12:11 < junka> kurahaupo; i think so 12:12 < bipul> What could anyone assume? if we get this output while installing any package via apt "E: Sub-process /usr/bin/dpkg returned an error code (1)" 12:13 < lupine> bipul: I'd expect to see some error output from dpkg in that case 12:13 < bipul> What's the meaning of "dpkg returned an error code (1)" ? 12:13 < autopsy> bipul, maybe a package scriptlet failed to run properly to completion. 12:13 < kurahaupo> bipul: what error is reported if you run dpkg directly? 12:14 < lupine> the meaning is exactly what it says. `dpkg ... ; echo $?` would output 1 12:14 < bipul> kurahaupo, where will i get the error report? 12:14 < kurahaupo> Stderr 12:15 < autopsy> bipul, doesnt apt have an error log mechanism? 12:15 < autopsy> I call it a mechanism. 12:16 < bipul> autopsy, Yes it has. 12:16 < autopsy> bipul, I know dnf has a history list it can undo actions it already completed using that history list. 12:16 < autopsy> bipul, what is the norm for error logs in dpkg? 12:17 < bipul> autopsy, https://paste.ubuntu.com/p/QG56TJdpJv/ 12:20 < bipul> autopsy, https://paste.ubuntu.com/p/PPtTGhKryh/ /var/log/dpkg.log 12:26 < azarus> I'm trying to decide between 3 wireless cards. One is the Intel Wireless AC 7260, one is the Compex WLE600VX and the other Compex WLE600VX. 12:26 < azarus> The Intel one is supported by iwlwifi, the others by ath10k. 12:27 < Polar_Bear> I am having some trouble with grub2. I skipped installing it when installing the rest of debian stretch earlier, which I have dual booting in UEFI mode with windblows10, so its normal install scripts didn't get to run. I've since installed the package since, but haven't got grub-install to play nicely. I think it's trying to install it in MBR mode and is failing with the message "grub-install: error: will not proceed with blocklists." 12:27 < azarus> Which ones will prove less problematic, and which ones will have better speed? 12:29 < dgurney> all of them will probably work just fine 12:30 < dgurney> I've never had problems with Intel or Atheros cards 12:30 < azarus> I'll probably go for the Compex WLE600VX, as it uses less power and support 802.11AC 12:30 < dgurney> okay then 12:30 < MrElendig> Polar_Bear: --target 12:37 < bipul> could i expect the solution of my problem which i have shared? 12:42 < Polar_Bear> MrElendig, thanks. That doesn't get me any further. If I specify --target=x86_64-efi I just get "grub-install: error: /usr/lib/grub/x86_64-efi/modinfo.sh doesn't exist. Please specify --target or --directory" 12:43 < Polar_Bear> should I be mounting the efi partition somewhere special first? 12:43 < MrElendig> find out which package owns that file and install it? 12:44 < MrElendig> azarus: ath10k is a steaming pile of crap 12:45 < MrElendig> azarus: https://media.ccc.de/v/32c3-7375-wireless_drivers_freedom_considered_harmful 12:45 < MrElendig> not that iwlwifi is much better 12:47 < autopsy> bipul, check dpkg logs 12:47 < autopsy> 02:00.0 Network controller: Ralink corp. RT5390R 802.11bgn PCIe Wireless Network Adapter 12:47 < autopsy> MrElendig, is RaLink no good? 12:48 < bipul> autopsy, Are you referring to this /var/log/dpkg.log ? 12:49 < MrElendig> autopsy: ymmv 12:49 < MrElendig> and do watch that video 12:49 < autopsy> bipul yes use tail on that file. 12:49 < MrElendig> also wireless.kernel.org 12:50 < EdLin> I can't seem to get alsa sound working in kernel 4.16 with my Ryzen APU.... 12:50 < autopsy> bipul, paste bin your log file for me to look at about the error. 12:50 < EdLin> tried unmuting it, no joy 12:51 < MrElendig> lspci | grep Audio 12:52 < EdLin> https://paste.pound-python.org/show/rZSg4kGVBo1fdKHyRtsy/ 12:52 < EdLin> only my video card and apu video audio is showing. 12:53 < oiaohm> MrElendig: its not just wifi drivers is all transmit software define radios. Regulators around the world are a pain in tail with them. 12:53 < EdLin> I can't use the apu video, as I'm using DVI-D with it, and can't use the video card audio, as that's in a VFIO-bound video card for virtualization 12:53 < EdLin> MrElendig: ^ 12:54 < MrElendig> oiaohm: it is perfectly possible to comply to the regulations without being assholes, sadly the wifi chip makers likes being evil 12:55 < oiaohm> MrElendig: not really. https://www.law.cornell.edu/cfr/text/47/2.944 This is the 2005 regulation for you. 12:56 < oiaohm> MrElendig: you can fairly much bet on getting a blob of random unknown quality. 12:56 < MrElendig> oiaohm: the video actually covers this in part 12:57 < MrElendig> oiaohm: there is a difference between hardlocking certain settings and making everything a non-free blob and actively sabotaging people who write free drivers 12:57 < MrElendig> eg atheros are doing both of the later 13:01 < azarus> MrElendig: can't listen to video now, is there a tldr 13:01 < MrElendig> "we are screwed" 13:02 < autopsy> Is RaLink sabatoging people who writre free drivers? 13:03 < azarus> so -- buying a AC 7260 is preferable? 13:03 < oiaohm> MrElendig: Manufacturers must take steps to ensure that only software that has been approved with a software defined radio can be loaded into the radio. << Its the first bit of the regulation. Where does this say that a software defined radio with open source is allowed as this is not taking all steps to prevent other stuff loaded into the raido. 13:03 < MrElendig> they are certainly not helping in any way, don't know if they are actively sabotaging the eforts 13:03 < oiaohm> Really if the regulation stated signed but source displayable would help open source drivers a lot. 13:03 < MrElendig> oiaohm: where does it say that that requires the software to be closed source? 13:03 < MrElendig> (it doesn't) 13:03 < autopsy> MrElendig, my wireless cards work fine here. 13:04 < azarus> I currently have a (ath9k supported) AR9285, but I'm longing for 802.11ac 13:05 < MrElendig> secureboot is an example where it is done without requiering that the software is closed source 13:05 < triceratux> MrElendig: i downloaded yesterdays swagarch 18.04 liveiso. the opera 51.0 download was going to take 2days. the firefox 59 download was going to take 2hrs. the wget brought it down in 9min. is there something weird about github.io that makes it particularly suited to wget ? https://swagarch.github.io/ 13:05 < MrElendig> azarus: all the current AC chips are quite troublesome, including the intel ones 13:06 < junka> swag 13:06 < azarus> MrElendig: which ones are the least troublesome, in your opinion? 13:06 < junka> none? 13:07 < autopsy> What is AC? 13:07 < MrElendig> rtl and broadcom (cypress now really) requires out of tree drivers, the iwlwifi is quite buggy and the radio on the cards are not terrific (and ofcourse, pcie only) 13:07 < azarus> 802.11ac is a wireless standard, autopsy 13:07 < junka> NO ITS POWER SOURCE 13:07 < azarus> ALTERNATING CURRENT 13:08 < junka> fixed 13:08 < autopsy> Oh haha I got ya. 13:09 < autopsy> azarus, is AC better than BGN? 13:09 < MrElendig> unless you are american, then it is air condition 13:09 < MrElendig> autopsy: yes 13:09 < autopsy> MrElendig, how is it better? 13:09 < triceratux> anonymous coward ftw 13:09 < junka> am confused 13:09 < revel> I'm a nonymous coward. 13:09 < MrElendig> autopsy: better performance, deals with congestion better, supports beam shaping and more advanced power saving etc et 13:10 < autopsy> MrElendig, I got ya. 13:10 < oiaohm> MrElendig: with 802.11ac cards going to windows does not help you a lot of the time either.as they still load up the the same insanely stuffed up firmware. 13:14 < BluesKaj> Hi folks 13:15 < triceratux> sup 13:16 < autopsy> BluesKaj, hi. 13:17 < JimBuntu> heya BluesKaj 13:18 < BluesKaj> hey triceratux, autopsy, just morning coffee atm...trying to wake up :-) 13:18 < triceratux> BluesKaj: yep coffee is a priority. i moved out of that lousy neighborhood & theyre not vandalising & stealing my car so much 13:18 < autopsy> BluesKaj, its 4:18 AM here. 13:19 < JimBuntu> Pacific Time 13:19 < BluesKaj> autopsy, yeah EDT here 13:19 < autopsy> JimBuntu, are you asking if I'm using Pacific Time Zone? 13:20 < JimBuntu> autopsy, Not really, I was stating that the time you mentioned,... would also be true in the Pacific Time zone 13:20 < autopsy> BluesKaj, isn't it 7 AM where you're at then? 13:21 < JimBuntu> Eastern shows ~7:20am 13:21 < autopsy> JimBuntu, nah MST here Mountain Standard Time. 13:21 < autopsy> I'm in Arizona. 13:22 < JimBuntu> autopsy, Ah... in one of those areas that doesn't follow daylight savings 13:22 < autopsy> JimBuntu, yeah we don't have to switch our clocks here. 13:22 < BluesKaj> triceratux, so you've had changes in your life too. I'm still tring to deal with the death of my wife on Feb 6 ..totally unexpected, still in shock and disbief 13:23 < BluesKaj> disbelief actually 13:23 < JimBuntu> I'm saddened to hear that this is still leading to shock and disbelief BluesKaj... then again, I can only imagine what such a thing would be like, I have only lost parents 13:23 < autopsy> BluesKaj, you run Linux? How long? 13:23 < autopsy> Yeah that must suck. 13:24 < BluesKaj> JimBuntu, it's tough ...sometimes, but it's helped my family come together 13:25 < triceratux> BluesKaj: zomg thats actually serious. my best wishes at this difficult time 13:25 < BluesKaj> autopsy, been running linux for about 12-13 yrs , but just a home user 13:25 < BluesKaj> triceratux, thanks 13:28 < BluesKaj> gotta keep going tho ..linux helps me keep a routine every morning, it's my hobby 13:28 < autopsy> BluesKaj, how long have you been using Linux? 13:28 < BluesKaj> testing kubuntu bionic beta atm 13:29 < BluesKaj> autopsy, read above 13:29 < autopsy> BluesKaj, oh 13 years. 13:31 < BluesKaj> most people here are veteran linux users abd devs I would think ..right triceratux , you're an old timer so to speak ? 13:34 < triceratux> BluesKaj: my youth was misspent on cobol, cics, & windows. so im an oldtimer with computers but i didnt get around to linux until 2009. im a bonafide n00b compared to most of you folks. so i multiboot to try to recoup the experience i dont have 13:35 * triceratux is testing lubuntu-next 18.04 lxqt with the rebranded Falkon 3.0. qupzilla is no more 13:36 < autopsy> triceratux, I've learned a lot from Fedora's LiveDVD capabilities about Fedora Linux and it's development model. 13:38 < triceratux> autopsy: i keep the current fedora xfce spin bootable as a reference install. i dont run it for long periods because its a little more crashprone than some other distros. vital to know what those guys are up to tho 13:39 < autopsy> triceratux, so xubuntu has systemd? 13:41 < triceratux> autopsy: yep. all the buntus are systemd. there is no escape. except mx-17 & a few others 13:41 < BluesKaj> autopsy, since 14.04 yes 13:42 < MrElendig> but still with the terrible sysv compat enabled :/ 13:42 < BluesKaj> well 14.04 was still non systemd 13:43 < triceratux> actually since 15.04. 16.04 was the first lts with systemd 13:43 < autopsy> Have they been slow to adapt systemd? 13:44 < triceratux> autopsy: nope they jumped right in once they made the decision. even beat debian by a few hours 13:44 < autopsy> triceratux, I see. 13:45 < autopsy> triceratux, isn't Ubuntu based on Debian? 13:47 < triceratux> autopsy: yep it is which is why they both switched at the same time. ironically its the ubuntu users who are slow to deal with systemd, because many of them only take lts seriously, & even then only after numerous point releases. so theres still a yuugh segment running 14.04 & the 16.04 conversions are ongoing even now 13:50 < JimBuntu> Why should the masses take none LTS seriously? If we are not testers, then why concern ourselved with non-LTS? 13:52 < triceratux> JimBuntu: thats exactly my attitude. canonical let you view the development cycle by way of the interim releases, but youre not forced to endure it if youre interested solely in a functional wellsupported system. works for me 13:53 < JimBuntu> triceratux, About the same for me. I listen/watch what is being said about non-LTS (regular) releases, but I don't use them... I can't imagine what it would be like to have to upgrade every ~9 months, lol, sounds a wee bit too MS like for me. 13:54 < kuri0> How can I extract the contents of a folder in a tar.gz ? 13:54 < MrElendig> man tar 13:54 < JimBuntu> yup kuri0 'man tar' is your friend. 13:55 < JimBuntu> kuri0, basically though, tar -zvxf 13:56 < hehehe> hey folks 13:57 < BluesKaj> personally I found LTS releases more unstable in their early days than the non-LTS OSs, but as a tester I've never found LTS releases particularly advantageous over the non-LTS OSs as a general rule for a home user like me. Production machines are a differnt story i'm sure. 13:57 < revel> I don't think you have to specify -z with a tar.gz 13:57 < jim> hi 13:57 < ananke> nor '-' 13:57 < JimBuntu> Hiya jim 13:57 < lupine> well, you wouldn't use ubuntu on production machines anyway 13:57 < hehehe> I have mounted a vps in a rescue mode, then I have mounted the partition and used passwd - idea is to reset root pass 13:57 < JimBuntu> revel, I think you are correct, if will probably recognize it's a gzip and work by default. 13:57 < revel> Right, tar flags don't need - either. 13:57 < hehehe> is that all I have to do? i 13:57 < ananke> hehehe: yes 13:57 < jim> JimBuntu, heya 13:58 < revel> -v is technically not required either, but hey, it's nice anyway. 13:58 < JimBuntu> lupine, You can clearly use Ubuntu on production machines, lol 13:58 < JimBuntu> revel, very true, I like being verbose though ;-) 13:58 < lupine> yeah, but could isn't the same as would 13:58 < lupine> you can clearly also use windows on production machines 13:59 < JimBuntu> lupine, Ok then... I DO use Ubuntu on a few production machines, as do many others 13:59 < revel> Depends on the machines. 13:59 < lupine> condolences 13:59 < MrElendig> people use win95 on production machines sp.. 13:59 < MrElendig> so* 13:59 < JimBuntu> lupine, in 10 years, no issues... please quantify your concern 14:00 < JimBuntu> MrElendig, Yes, many ATM machines (and banking computers in general, as well as medical facilities) do still use Win95, although, I don't really see the comparison 14:00 < lupine> dodgy upgrades, heavy churn, etc 14:00 < lupine> then there's the governance problems 14:00 < revel> Churn? 14:01 < lupine> e.g. most server distros have gone through one init system change in the last 10 years. ubuntu has gone through 2 14:01 < lupine> this is churn 14:02 < jim> MrElendig, is there something about earlier versions of windows that make it more suited as a remote server? 14:02 < jim> (briefly, I don't mean to derail present discussion) 14:02 < JimBuntu> "dodgy upgrades" -> I have not yet had an issue... churn? idk... governance, no issue at all as I can install various programs to handle better security, asset management, logging/etc... 14:02 < JimBuntu> Perhaps I am simply lucky. 14:03 < ananke> JimBuntu: or have a small enough number of systems where these issues are mitigated with hard work 14:03 < JimBuntu> jim, many companies had custom apps written around the time of Win95... apps that don't work 100% as expected on newer versions. That's the excuse I have been given time after time. 14:03 < cloudbud> how to remove new lines in a file 14:04 < jim> oh ok, that makes sense 14:04 < JimBuntu> ananke, possibly... 40+ machines serving the general public 24/7/365.25 - No "hard" work, barely any work in regards to updates/etc (unattended mode active) 14:04 < ananke> cloudbud: first hit on google. seriously. 14:04 < rumpel> cloudbud, I'd look for "sedoneliners" 14:04 < jim> "new lines for old... new lines for old..." 14:04 < lupine> "I don't see it" isn't the same as "it's not there" 14:05 < lupine> but to each their own. you have my sympathies, and an install key of your choice if you ever do encounter any of the problems 14:05 < jim> cloudbud, do you mean remove the line endings so that what you're left with is one (unterminated) line? 14:06 < ananke> JimBuntu: which releases? I'd be curious as to your experience with dealing with such annoyances as automatically removing unused kernels, or the obnoxious automated install (although the latter is also shared with debian) 14:07 < hehehe> ananke are you sure? 14:07 < cloudbud> jim : I have a file that have a record then so many blank lines then another record then lines how can i remove the blank lines 14:07 < JimBuntu> ananke, LTS only, Everything is either 14.04.x ( a few dev boxes) or 16.04.x. Increase the /boot partition and never run out of space, or purge old kernels/etc once every few blue moons. 14:07 < hehehe> there suppose to be a change from # to some other sign 14:07 < ananke> hehehe: yes. _if_ you were using passwd from chroot, not just running 'passwd' period 14:07 < hehehe> I recall 14:07 < hehehe> ananke: how would I use passwd from chroot 14:07 < hehehe> i just run passwd 14:08 < ananke> JimBuntu: which exactly proves my point, it's a hackish solution that doesn't scale well 14:08 < kuri0> JimBuntu, doing that creates extracts the files in that folder to the folder with the same name 14:08 < kuri0> what i want is it to extract the files from that folder 14:08 < ananke> hehehe: you need to chroot to your root filesystem first 14:08 < JimBuntu> kuri0, then once you are done, mv the files from the folder into ./ or . if you prefer 14:08 < hehehe> ananke how? 14:08 < jim> cloudbud, oh, so you mean to remove the extra newlines (the ones that match spaces \n? 14:08 < ananke> hehehe: chroot /your/mounted/root/filesystem 14:09 < cloudbud> jim : yes 14:09 < hehehe> cool 14:09 < hehehe> ty 14:09 < hehehe> that is the missing bit 14:09 < ananke> cloudbud: 'blank lines' are not the same as 'new lines' 14:09 < JimBuntu> ananke, if I were inclined, I guess I could cron the apt purge, simple. Which distro do you suggest that doesn't keep previous kernels? 14:10 < JimBuntu> remove all lines that are only /r/n or /r or /n 14:10 < ananke> JimBuntu: distros better suited for production, where automation is necessary. rhel/centos/sles. you can easily define how many previous kernels you want to keep 14:10 < hehehe> ananke: ty 14:11 < JimBuntu> ananke, not a fan of rhel or centos, I'll check out sles, thanks. 14:11 < ananke> JimBuntu: ubuntu even in LTS mode is ill suited for what typically constitutes 'production' 14:11 < jim> cloudbud, I'd first look at those newlines and see if there are spaces before the newline 14:12 < cloudbud> jim : sample file https://pastebin.com/WDtEfMQD 14:13 < fSharp> hello, my encrypted luks partition rejects my correct password. after having tried all sorts of things to find the reason and open it, I can only reduce it down to two causes; 1. clipboard pasted the password incomplete (I saw that it happened with pasted passwords on gedit) 2. gnome disks on ubuntu 17.10. errorenously created the encrypted partition. have you had such experiences with these? 14:13 < lupine> centos shares many of hte same governance issues as ubuntu, unfortunately 14:14 < JimBuntu> Clouder, 'egrep -v " +" >output.txt' 14:15 < cloudbud> JimBuntu : not working 14:15 < JimBuntu> Clouder, please pastebin or termbin the terminal 14:16 < JimBuntu> Clouder, I saved your example file, and ran it with that command, only output was the non-space/CR/newline lines. perhaps your real file is different from example 14:17 < fSharp> are there any clipboard pasting errors that are seen with certain i.e. non-ASCII characters? 14:17 < NGC3982> hm, my system keeps telling me my grub is broken when i try to update. 14:17 * NGC3982 doesnt dare a reboot. 14:18 < cloudbud> JimBuntu : https://pastebin.com/WDtEfMQD 14:18 < the_drow_> What could cause a USB wifi stick not to connect to Wifi. demsg claims the link is not ready 14:18 < the_drow_> The firmware is there 14:18 < JimBuntu> cloudbud, same link? 14:18 < the_drow_> and loaded 14:19 < cloudbud> JimBuntu : https://pastebin.com/WDtEfMQD 14:20 < rockdarko> When doing the updates remotely on my dad's laptop I get something that makes me scared to reboot when pacman (Manjaro) runs grub config: Incompatible libdevmapper 1.02.146 (2017-12-18) and kernel driver (unknown version). 14:20 < rockdarko> Full output: https://darko.vision/pub/blackmage01 14:20 < JimBuntu> cloudbud, I have saved that file from raw, ran the egrep command, and it worked for me. What distro are you using and what is the output from 'egrep --version' ? 14:22 < cloudbud> JimBuntu : done 14:23 < Kingsy> is it possible to change any tray icon? I assume its a file somewhere you can replace if you find it/ 14:42 < kuri0> JimBuntu, any way to do without that ? 14:46 < rocketmagnet> hi all 14:47 < rocketmagnet> i've a question: i'm using kali linux and i want to change the kernel boot parameters so they stay permantently, normaly i can do this with /etc/default/grub but since it's a live distro i've no idea where to put the kernel boot parameters so they stay when robooting 14:49 < spreeuw> if its a true live boot distro you cant 14:49 < spreeuw> unless you installed it rw somewhere 14:50 < spreeuw> or when you adjust the ro iso when creating the distro 14:50 < spreeuw> or enter it manually in grub each time you boot 14:54 < azarus> MrElendig: After watching that talk you linked (thanks, btw) I decided something like this is nice: https://v.gd/M4rO7O Supported by the mt76 driver 14:58 < autopsy> rocketmagnet, is it on a DVD or CD that you boot from? 14:59 < fSharp> are there any clipboard pasting errors that are seen with certain i.e. non-ASCII characters? 15:00 < fSharp> in other words, can copy-paste make errors with non-ASCII characters? 15:02 < JimBuntu> fSharp, it *could*, but I wouldn't expect it to, not specifically for non-ASCII... 15:03 < fSharp> JimBuntu, I created a luks partition by copy-pasting the password, and it doesnt open now. it is either the clipboard or gnome disk utility 15:04 < JimBuntu> fSharp, This should be somewhat easy to test... copy the password and paste it into a few different places, such as a text editor, possibly even a text editor set to something less than UTF-8 and see what happens 15:04 < JimBuntu> I'm not up on my luks/gnome disk utility... is there any chance your password was too long?... and was truncated? 15:05 < fSharp> JimBuntu, I tried it..the password in question is shown properly, but just yesterday I witnessed that non-ASCII was not copied from browser to gedit 15:06 < fSharp> so I wonder, even if it is shown properly now, maybe at the moment of partition creation something was missing. 15:06 < fSharp> whether or not that's possible 15:06 < fSharp> I dont think it is too long. longer passwords had no problem 15:07 < JimBuntu> fSharp, I'm unsure. I can say that since you asked, I went to a browser page with what I consider non-ascii chars... copied and pasted one (Chrome) and it worked fine. I may not be the best test subject though as I am using Synergy, which ties into my clipboard 15:08 < fSharp> yes, I guess it is not easy to detect if anything goes wrong 15:09 < fSharp> the other thing that comes to mind is having added c,a or v to the password while copy-pasting..but that brings a whole lot of combinations to test 15:10 < JimBuntu> fSharp, I guess it stand to reason that I should ask... is there anything in the encrypted partition that is worth the time/effort or do you have backups? 15:10 < fSharp> JimBuntu, I 15:10 < fSharp> I've been asking this, too..I think there is 15:11 < fSharp> backup..was the partition itself. 15:11 < fSharp> and the system has been reinstalled 15:11 < JimBuntu> ouchies. 15:12 < Hrtlnd> I have a VM a former assistant locked down so I can't get remote access to it through putty. I can log on locally, and I've tried to find where he did the lock down, but having no luck. 15:12 < Hrtlnd> I've checked IPtables port 22 is allowed for the entire subnet I'm on 15:13 < Hrtlnd> Is there another place I should look? 15:13 < JimBuntu> Hrtlnd, did you verify the SSH daemon is running? 15:14 < Hrtlnd> Yes, it's started 15:14 < fSharp> JimBuntu, thank you for your time 15:14 < JimBuntu> Hrtlnd, i.e. 'sudo service ssh status' and if running, perhaps a 'sudo cat /etc/ssh/sshd_config' to verify it's set to use port 22... or optionally, 'sudo netstat -tunelp | grep -i ssh' 15:14 < fSharp> I will see how/what I can do 15:15 < JimBuntu> fSharp, You are welcome. 15:15 < fSharp> bye! 15:15 < Hrtlnd> Running since 2017-11-26 15:16 < Hrtlnd> listening on 22 15:17 < JimBuntu> Hrtlnd, ok, does it only accept keys or also passwords? If it accepts keys and you can locally log in, verify the key you are using is properly set within the ~/.ssh/authorized_keys file 15:18 < JimBuntu> Hrtlnd, it would actually be good to see what error (if any) you are getting when you try to SSH in 15:19 < Hrtlnd> So one of the tasks I gave him over a year ago was to implement a process where an email was fired on login so we could see when a person had gained access as part of an audit trail. 15:19 < Hrtlnd> So when I SSH from my PC it takes me to logon, but password that works on console fails. 15:20 < xdije> when i run pvs i get WARNING: Device for PV 9pFXew-uqNo-T6yR-emB5-1sRe-Bf34-0Bwvbj not found or rejected by a filter. 15:20 < Hrtlnd> It was supposed to be password only 15:20 < xdije> and is only for sbd 15:20 < azarus> Is it enough to have CONFIG_WLAN_VENDOR_MEDIATEK=y and CONFIG_MT7601U=m to have support for the mt76_pci driver? 15:20 < Hrtlnd> It's possible he created a key 15:20 < Hrtlnd> but he didn't supply me with one if he did 15:21 < Sitri> You should've supplied your public key... 15:21 < xdije> lvmconfig devices/filter filter=["a|^/dev/mapper/mpath|","a|^/dev/sdb|","r|.*/|"] 15:21 < JimBuntu> Hrtlnd, I would check the /etc/ssh/sshd_config and the /home//.ssh/authorized_keys to be sure. I would suggest against ever letting passwords be used, but that's obviously up to you 15:22 < Hrtlnd> It's an internal server not internet accessible so I'm not majorly concerned 15:22 < autopsy> JimBuntu, I use secure shell to my online hosting account servers using a password all the time. 15:22 < JimBuntu> autopsy, Everyone is welcome to do as they wish :-) 15:23 < Hrtlnd> So /home is an empty directory 15:24 < JimBuntu> Hrtlnd, when the password fails, are you presented with a line that reads something like 'Permission denied (publickey)' ? 15:24 < JimBuntu> Hrtlnd, what distro? 15:24 < Hrtlnd> No, just Access denied 15:24 < Hrtlnd> I believe CentOS 15:24 < JimBuntu> Hrtlnd, are you logging in as root? 15:24 < Hrtlnd> Yes, root 15:24 < JimBuntu> ok, then check /root 15:25 < autopsy> Oh root over secure shell on a LAN. 15:25 < Hrtlnd> directory has an .rpm file and that's it. 15:27 < JimBuntu> Hrtlnd, Ok, sounds like it's set up to accept a password, and there are no keys configured in the default way that I am aware of. 15:27 < JimBuntu> The port is open, since you get the Access Denied message 15:27 < Hrtlnd> So prior to his leaving I was able to log on to this server. 15:28 < Hrtlnd> I unfortunately was away on his leaving 3 days, and things were relatively amicable from what we could see. 15:28 < Hrtlnd> Since he left, we found some questionable stuff that makes us believe this was deliberate. 15:29 < JimBuntu> Hrtlnd, if there is any doubt, I would move the important data from the VM into a new one... never know what odd kind of tunnels could be lurking around 15:30 < Hrtlnd> Looks like that's the thing I need to do. 15:31 < Hrtlnd> Thanks for your help. 15:36 < autopsy> Hrtlnd, someone hacked your secure shell server or what? 15:38 < lowin> How do I make new policykit rules take effect? don't wanna restart the whole system 15:39 < Hrtlnd> No, former assistant left last year, I can only log on locally now, not ssh. The server is physically on-site and doesn't touch the net. 15:53 < xdije> my centos7 system hangs on Reached target Shutdown, untill i hard rest it, any explaination for this 15:55 < autopsy> xdije, why not just use Fedora 27? 15:56 < xdije> autopsy: it is a productive server 15:56 < Sitri> xdije: modern mobo? Because really old systems didn't have a software means to tell the mobo to completely the shutdown. 15:57 < autopsy> xdije, yeah how old is the motherboard? 16:00 < xdije> autopsy: HP G9 16:01 < xdije> now while we are talking it has rebooted but took long time for that 16:07 < puff> I'm running google chrome on ubuntu 16.04 LTS. Lately chrome has been spiking to 100+ % of CPU. When I open the chrome task manager it's not some particular window or extension, it's "Browser". This seems to be recent and it's incredibly annoying. Anyone know what's going on? 16:08 < dysfigured> is there a way to review some log of processes run by a user? 16:09 < puff> dysfigured: Not particularly processes by user, though /var/log/* probably has some relevant info. If you want to view logs *currently* run by some user, "ps -fu username" 16:10 < puff> dysfigured: Er, s/view logs/view processes/ 16:10 < dysfigured> i have a vps i let a couple friends have accounts on, and i'm wondering about some sort of monioring system 16:10 < dysfigured> checking shell history isn't reliable.. 16:11 < puff> dysfigured: That's correct. 16:11 < dysfigured> checking the current process list is only an instantaneous view, i'm wondering about tracking processes over time by specific users 16:11 < puff> dysfigured: Given the history of unix as a multi-user system, I'm sure there's something out there, but how much hassle it'll be to use it is another question. 16:11 < dysfigured> hell, myself and root too, good data as well 16:12 < puff> dysfigured: You could run a program that constantly logs the ps output. 16:13 < dysfigured> i was semi thinking there would be systemd logs somewhere 16:15 < puff> Good question. I'm new at systemd so I don't know. 16:17 < monkeyisl> ln is like "shortcut" on windows? 16:18 < amosbird> hi, how can I clear the effect of sched_setaffinity() ? 16:24 < ayecee> set the affinity to all cpus 16:32 < Richard_Cavell> Hi everyone. I want to ask about the use of the ldd command. Is this a good place to ask? 16:32 < ayecee> Richard_Cavell: one way to find out is to ask 16:32 < ayecee> about the use of the ldd command 16:32 < ayecee> not about whether it's a good place to ask about the use of the ldd command 16:33 < Richard_Cavell> I don't really understand how libraries work. I have some coding knowledge but I want to understand what the output of ldd means here: 16:33 < Richard_Cavell> if you can't cure the cause, mitigate the symptoms! 16:33 < Richard_Cavell> oh dear hang on 16:33 < ayecee> heh 16:33 < Richard_Cavell> https://ideone.com/iPFgmY 16:33 < ayecee> site doesn't load 16:34 < Richard_Cavell> loads for me. 16:35 < ayecee> there it goes 16:36 < ayecee> now, what's the question 16:37 < Richard_Cavell> lines 9 through 11... I completely don't understand them 16:37 < kabads> It's DNS and I'm beat! 16:37 < djph> linux-vdso.so.1 is at memory location (whatever) 16:38 < djph> libc.so.6 is /lib/x86_64[...] and at memory location (whatever) 16:38 < djph> same for the last one 16:38 < Richard_Cavell> okay so my code uses 3 libraries, is that correct? 16:38 < ayecee> Richard_Cavell: yes 16:38 < ayecee> three that are dynamically linked anyways 16:39 < Richard_Cavell> You can see how simple it is. Should I be bothered that it uses 3 libraries, or is that just normal? 16:39 < ayecee> that's normal 16:39 < Richard_Cavell> okay 16:39 < Richard_Cavell> And the fact that they have memory locations, does that mean that they were already loaded when I ran ldd, or that they have been loaded for the purpose of ldd, or what? 16:40 < djph> that's normal. first is linux headers, second is the standard C library, third is I forget what. These three get pulled in *every* time you compile a C program 16:40 < ayecee> they were loaded when ldd ran the executable 16:40 < ayecee> third is the dynamic library loader 16:40 < ayecee> first isn't "headers" 16:40 < ayecee> though i don't know offhand what that is. 16:41 < djph> they're pointing to memory locations, because they're "shared objects" -- that is, they're dynamically called, rather than being compiled directly into the program you wrote. 16:41 < djph> ayecee: dammit ... 16:41 < Psi-Jack> ayecee: What about onhand? 16:41 < ayecee> i have the manpages onhand 16:41 < ayecee> i guess i could hunt through those. 16:41 < Psi-Jack> hehe 16:41 < djph> vdso - overview of the virtual ELF dynamic shared object 16:42 < ayecee> something to do with making system calls faster, looks like 16:42 < djph> yup 16:43 < Richard_Cavell> So when I run ldd, it actually runs the program, and the program then loads up those 3 libraries? 16:43 < djph> no 16:43 < ayecee> yes 16:44 < djph> wait what? 16:44 < ayecee> now you know 16:44 < prussian> ldd will indeed execute code 16:44 < djph> hm, it appears I've misunderstood the manpage then 16:45 < prussian> that's why I prefer objdump if possible 16:45 < ayecee> i know, it's kind of surprising :) 16:45 < djph> not really no, manpages suck sometimes. :) 16:45 < puff> monkeyisl: Re: ln, sort of but not at the technical level. 16:45 < Richard_Cavell> So the second one is "the standard library". Does that mean that the entire C standard library is one binary chunk called libc.so.6? 16:46 < puff> monkeyisl: In windows, if I recall correctly, a shortcut is a file that contains several settings, one of which is the location of the executable file that you actually want to run. 16:46 < djph> I had interpreted it as "ldd /some/executable" just inspected the executable, and didn't actually *run* it. 16:46 < ayecee> yeah, that's the surprising part 16:47 < djph> ... oh ... it's hidden in another manpage ... ld.so(8) 16:47 < djph> errr LDCONFIG(8) 16:48 < djph> whatever ... I'm too tired after dealing with bang..where-ever to think straight 16:48 < puff> monkeyisl: In linux, the ln command is used to create a "link" to a file. In a limited sense, a link is similar to a windows shortcut in that it contains the location of another file. But how links are implemented at the very low level is a bit different. 16:49 < puff> monkeyisl: https://www.linux.com/learn/intro-to-linux/2017/6/understanding-linux-links 16:50 < gump> anyone else notice that NPM in CentOS 6 stopped working on friday? 16:51 < prussian> i iamgine all of node ecosystem in centos6 should stop working 16:51 < oiaohm> puff: monkeyisl https://developer.gnome.org/desktop-entry-spec/ shortcut file on windows the closest on Linux is a .desktop file. 16:51 < prussian> get the latest nodejs stable release from some scl or external sources 16:52 < gump> prussian: yeah the official node repos still work fine 16:58 < Polyonymous> I'm looking to throw a linux (CentOS) install on my local server. I'd like to use a pair of old laptop HDDs in such a way that, when one of them dies, I can simply slot in yet another old laptop HDD and have everything chug along happily. With minimal effort on my part. Although this seems like an obvious case for a software RAID1, I've not tinkered with it in years and don't know if there's a good way to configure booting in this 16:58 < Polyonymous> Is there a clever way to boot from mirror raids nicely? (Preferably something I wouldn't have to sit down and re-jigger every time something broke.) Am I better off digging up an old RAID controller and booting through that? Would using a RAID controller make it unfeasable to boot from either drive (degraded, obviously) should the controller fail? 17:01 < ananke> Polyonymous: what happens is that you have raid1 for your boot filesystem, and the computer boots from one of the sides. after linux is booted, it assembles a raid1 array 17:01 < ananke> Polyonymous: there are plenty of guides out there showing how to do it 17:02 < Polyonymous> Aye, I've found guides on how. I'm trying to learn about the failure modes before I put the whole mess together. 17:03 < Psi-Jack> Heh, the deeper I dig in, the more curious I get. The original author of Grav CMS is actually the co-founder of Joomla. A former Drupal core dev left Drupal and is now working with Grav. Iiiiinteresting. :) 17:04 < Polyonymous> More clearly stated: my question is if there is a 'best practise' implementation. For example, would an older, wiser user just back up key files and lose less time to tinkering? 17:05 < Psi-Jack> Always backup. 17:05 < Polyonymous> Always do 17:06 < Polyonymous> But off-site backups are for when the entire box melts into slag, I'm curious about the 'get it back online with little effort because one of the HDDs decided to release the magic smoke' case 17:09 < ayecee> Polyonymous: grub can boot from raid, though you'd want to install the bootloader on each of the disks in the boot raid set. 17:10 < ayecee> raid1 at least, not 100% sure about raid5/6 17:10 < Polyonymous> Just raid1 17:10 < Polyonymous> And great. 17:10 < JimBuntu> Polyonymous, Then RAID is your friend. RAID1/5 17:10 < ayecee> raid can be a fickle friend 17:10 < Polyonymous> Aye. I tend to be a bit paranoid about backups, so it's allowed to be fickle 17:11 < Polyonymous> And I've found raid1 to be much better behaved, historically >.> 17:11 < fendur> ayecee: fickle how? 17:11 < fendur> you can just point to an article/blog 17:11 < azarus> as I understand, mt76 supports this wifi card: https://v.gd/M4rO7O 17:11 < Polyonymous> Anyway. Thanks for sanity checking me 17:12 < ayecee> fendur: fickle in that if you don't set it up right, it don't work right. 17:12 < ayecee> like, if you install the bootloader on one disk but not the other. 17:12 < azarus> but in my distro I don't see a mt76.ko? CONFIG_WLAN_VENDOR_MEDIATEK=y is there... :s 17:14 < the_drow_> I'm trying to mount an ntfs volume with udev. I'm getting Transport endpoint is not connected 17:14 < the_drow_> From my research on the internet it seems like it's not possible to mount a ntfs usb volume with udev 17:14 < the_drow_> I'd rather not walk away from my implementation and move to udisks2 if possible 17:15 < the_drow_> Is there a way around it? 17:15 < ayecee> the_drow_: that doesn't sound like a message related to ntfs. how are you mounting it? 17:15 < ayecee> also, does your research say why that's the case? 17:15 < the_drow_> ayecee, '/bin/mount -t ntfs-3g -o gid=100,dmask=000,fmask=111,utf8,flush,rw,noatime,users /dev/sdc1 '/media/flash_drive_sdc1'' 17:16 < the_drow_> no, the debian mailing lists says just because 17:16 < the_drow_> ayecee, https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=774149;msg=59 17:16 < ayecee> weird. that sounds more like a samba or nfs error. 17:16 < ayecee> oh, usb transport. 17:17 < ayecee> huh. i don't have any ideas, but i appreciate the problem. 17:18 < the_drow_> mounting manually works just fine 17:25 < Psi-Jack> oh good, the cafe has arrived with coffee? ;) 17:26 < Pentode> coffee? where? 17:26 < absurdistani> COFFEE! 17:26 * Pentode looks around frantically 17:32 < Psi-Jack> Well, now I have my name in Grav's code history. Muahaha. 17:32 < edge_hay> Hi 17:33 < edge_hay> So I'm coming from ##windows… 17:33 < edge_hay> They didn't like me mentioning that I was running a stable os 17:34 < edge_hay> Hope you guys don't mind 17:34 < fendur> I also think stable OSs are stupid. 17:34 < edge_hay> You mean any os apart from windows? 17:35 < fendur> just stable OSs 17:35 < edge_hay> You're in a linux chan :/ 17:35 < fendur> and? 17:35 < edge_hay> What os are you running? 17:35 < fendur> fendos. It's terrible. 17:35 < edge_hay> System Information: Model: MacBook Pro (13-inch, Retina, Late 2013) • OS: macOS High Sierra (Version 10.13.4, Build 17E199) 17:35 < edge_hay> It's terrible too 17:36 < edge_hay> Kernel panic per year: 0.02 17:37 < edge_hay> s/0.02/0.2/r That's one every 5 years for me. 17:37 < zaruba> I just downloaded the latest firefox, now I'd like to create a launcher for it but I'm not sure how to do that. Anyone? 17:37 < edge_hay> A launcher? 17:38 < edge_hay> What do you mean by that? D: 17:38 < revel> https://frugalfun4boys.com/2013/06/06/how-to-build-a-catapult-out-of-dowel-rods-and-rubber-bands/ 17:39 < edge_hay> Still more stable than windows 17:39 < zaruba> a shortcut link to open that application, edge_hay 17:39 < revel> zaruba: So, a .desktop file? 17:39 < edge_hay> what distro? 17:39 < zaruba> yes 17:39 < zaruba> debian 17:39 < edge_hay> Oh. Never ran debian with a gui :/ 17:39 < revel> .desktop files are sort of universal. 17:40 < the_drow_> ayecee, Well, udisksctl mount works. I just had to sed into fstab for it :( 17:40 < edge_hay> revel is it like a .bash_profile file? 17:40 < mnemon> zaruba: https://wiki.archlinux.org/index.php/desktop_entries has some example files 17:40 < revel> ^ 17:41 < revel> edge_hay: No, it's like... Well, you can have a specific image, process, description etc. tied together so you can use them as desktop shortcuts or in menus or whatever. 17:42 < revel> /usr/share/applications/ contains a ton of 'em. 17:42 < revel> (don't stick your own there) 17:43 < edge_hay> I don't have /applications folder on macos 17:44 < delt> Hello 17:44 < Richard_Cavell> So may I ask about libraries again? This file libc.so.6 - does the so stand for anything? 17:44 < delt> what's the name of the utility to draw directly on the root window? 17:44 < revel> Richard_Cavell: Shared object. 17:44 < revel> edge_hay: I doubt .desktop files are a thing on MacOS. 17:44 < Richard_Cavell> revel, ok thanks. And what does the number 6 mean in libc.so.6 ? 17:44 < delt> draw text i mean 17:45 < JimBuntu> revel, macOS has an equiv 17:45 < mnemon> Richard_Cavell: shared object or something like that iirc? denotes shared libs 17:45 < revel> Some sort of version, probably. 17:45 < revel> JimBuntu: So does Windows, but it's not the same. 17:45 < edge_hay> revel that's my /usr/share/ https://pastebin.com/PxhGbkc3 17:45 < JimBuntu> revel, in macOS, they are .command files. I can't remember if you can assign icon/etc. 17:46 < revel> edge_hay: If you really are on MacOS, then having .desktop files would be pointless. 17:46 < Richard_Cavell> So on my machine my libc.so.6 file is a link to another file that is 1.8 megabytes. Am I to take it that this 1.8 megabytes implements the whole of the standard C library? 17:47 < edge_hay> oh 17:47 < revel> 1.9 megabytes for me :D 17:48 < azarus> When was the mt76 driver introduced? 17:48 < revel> Richard_Cavell: Yes. 17:48 < ayecee> Richard_Cavell: sounds reasonable 17:48 < edge_hay> :D 17:48 < edge_hay> C OP 17:48 < Richard_Cavell> Now, I'm assuming that many programs on my machine use the C standard library. Is this library always loaded into RAM? 17:49 < djph> Richard_Cavell: for some values of "always" 17:49 < edge_hay> Your kernel most likely is made in C Richard_Cavell 17:49 < jelly> Richard_Cavell: most of it, some stuff is separate (floating point is in libm.so*, some localization stuff is in libintl.so*) 17:49 < revel> edge_hay: It's a "freestanding" binary that doesn't have a C library. 17:49 < ayecee> Richard_Cavell: pretty much, yeah 17:50 < mices> no matter how many drives in a raid array i only need one for parity to be able to protect against data loss if a drive in the array fails? 17:50 < djph> mices: no 17:50 < ayecee> mices: yes 17:50 < revel> Richard_Cavell: I don't think every memory gets its own libc loaded into memory, but it's there. 17:50 < revel> s/memory/process/ 17:50 < edge_hay> revel doesn't mean that the binary doesn't contain library functions :/ I think they do 17:51 < edge_hay> s/they do/it does/r 17:51 < ayecee> mices: however, that would only address a single disk failure scenario. 17:51 < mices> ayecee: i'm thinking about building a raid array from usb sticks 17:51 < ayecee> neat 17:51 < edge_hay> It's so nice to see people use regex in irc :D 17:51 < mices> does anything else make sense comparitively 17:51 < edge_hay> ttyX D: 17:51 < edge_hay> nuuuh 17:51 < edge_hay> who 17:52 < ayecee> mices: sounds like a fun experiment, but doesn't sounds practically useful. 17:52 < ayecee> i couldn't immediately say why though. 17:52 < revel> edge_hay: The kernel? Well, uhh, I'm not a professional on the topic. 17:52 < mices> ayecee: and i backup to what then? 17:53 < ayecee> mices: what do you back up to now? 17:53 < revel> edge_hay: But no, kernels can't depend on libraries. 17:53 < mices> another usb stick array? 17:53 < edge_hay> Yeah neither am I but functions like malloc are implementation defined so they must be defined 17:53 < mices> if my data was on a usb stick array 17:53 < ayecee> it's usb stick arrays all the way down. 17:53 < mices> i'd backup to another usb stick array 17:53 < edge_hay> yeah I guess it's not using libc 17:53 < djph> revel: I believe that the kernel, if compiled from source, uses internal static links to the C libraries that it needs, as opposed to dynamic links via libc.so.X 17:53 < fendur> mices: I have done this, and I use it regularly to back up work files on my work machine. I assume it's safer than using one thumb drive. 17:54 < revel> I meant dynamic libraries, yeah. 17:54 < djph> s/if compiled/when compiled/ 17:54 < edge_hay> eitherway 1.8Mb in memory isn't much 17:54 < mices> what's a better way to format usb sticks with gpt or msdos partition tables 17:54 < djph> not at all 17:54 < absurdistani> revel: the kernel has its own C lib. from there, if you are using dynamic linking, then you likely have C lib loaded in RAM. if you are using some distro like Stali, then no. 17:54 < mices> to do this use gpt? 17:54 < revel> Since it's just standing on top of the hardware on its own with nothing there to manage memory or handle filesystems for it. 17:55 < absurdistani> mices: I prefer to use MBR on USB sticks. Makes them more portable. 17:55 < BCMM> mices: as long as the USB stick doesn't need to be bootable, it's quite possible to put the filesystem on the whole block device 17:55 < BCMM> i.e. with no partition table at all 17:55 < BCMM> (and yes, windows supports that) 17:56 < absurdistani> BCMM: some version of macOS freak out if you do that 17:56 < BCMM> huh, didn't know that 17:56 < fendur> mices: mine is built on a raspberry pi zero (W), so it effectively acts as a portable, potentially batter powered NAS. 17:56 < BCMM> it used to be relatively common to get USB sticks like that out of the box 17:56 < absurdistani> BCMM: i discovered this in a very inconvenient way :( 17:56 < Trel> When SSH'd into a machine using a key, is there a way to see WHICH key in the authorized_keys file I'm using? 17:57 < BCMM> Trel: ssh -v 17:57 < edge_hay> If it's compatible with windows it works on macos 17:57 < BCMM> Trel: it will tell you all about which keys its offering and which one is accepted 17:57 < mices> fender: you're using a powered hub? 17:57 < mices> fendur ^ 17:57 < Trel> BCMM: that just prints the hel 17:57 < fendur> mices: no, it's passive 17:57 < Trel> *help 17:58 < BCMM> Trel: i mean, add the -v option to your existing ssh command line 17:58 < fendur> er, whatever they call not-powered usb hubs. 17:58 < mices> fendur: i'm looking at hubs like this on ebay https://www.ebay.com/itm/7-Port-USB-2-0-Multi-Charger-Hub-High-Speed-Adapter-ON-OFF-Switch-Laptop-PC-US/382164931804?hash=item58face14dc:g:760AAOSwhilZaT67 17:58 < BCMM> Trel: ssh -v user@host 17:58 < edge_hay> so just make it compatible with Windows and Apple will do the rest. Unless it's ntfs. In which case god bless your soul 17:58 < Trel> BCMM: I'm connecting with Putty, I need to see which in the remote machine 17:58 < BCMM> oh, no idea 17:58 < djph> Trel: you don't see "in the remote machine" -- the remote only holds PUBLIC keys. 17:58 < repys> I had a RAID1 in my computer and suddenly I got that the system didn't start anymore. at the boot it didn't find the efi boot file from both disks. few hours before I changed power supply of my computer, I am not sure if that would be related to. but my question is, I can't see any bad sector from smartctl from both disks. 1. what could be it happe 17:58 < repys> ns to these disks? 2. can smartctl see if there problems in the first sectors of the disk (where efi files are)? 17:59 < BCMM> Trel: might be logged by sshd, depending on the settings. are you root? 17:59 < djph> you're presenting the remote with a private key, and it's comparing a sig against the public keys it knows are authorized. 17:59 < Trel> djph: I'm trying to see on the remote machine, which pubkey in authorized_keys is being matched for the connected session 17:59 < BCMM> djph: he asked "which key in the authorized_keys file" 17:59 < Trel> BCMM: I have sudo access, but I'm not SSHing to root 17:59 < BCMM> djph: so i don't see a reason to think he doesn't know that 18:00 < djph> BCMM: all kinds come to all places :| 18:00 < edge_hay> mices: https://www.amazon.com/gp/product/B01N9M32TA/ref=od_aui_detailpages00?ie=UTF8&psc=1 I bought that, works great… No idea about having 7 usb ports tho 18:00 < BCMM> Trel: hang on, putty doesn't let you offer multiple keys anyway, does it? 18:00 < repys> any idea guys? :) 18:00 < BCMM> Trel: you're using whichever key putty is configured to use, surely? 18:00 < ayecee> several ideas! 18:00 < ayecee> just not about your thing 18:01 < edge_hay> Probably the best usb hub I've ever bought. All the ports support charging. The cable is thick (that's what she said) and the speed is the same 18:01 < fendur> mices: mine is only 2 drives. But if you're thinking of many drives, yeah powered might help. 18:01 < djph> Trel: perhaps state the problem / query in a different manner? 18:01 < absurdistani> repys: boot to a livecd and then do: mdadm --assemble --scan 18:01 < absurdistani> repys: note the disk names 18:02 < absurdistani> repys: then you can do like: mdadm --detail --scan >> /mnt/targetAssembledArray/etc/mdadm.conf 18:02 < repys> absurdistani: yes I can mount the disk but of course I can't boot anymore the computer from these disks because I get that efi boot files are not found 18:02 < absurdistani> repys: ah okay 18:02 < edge_hay> Trel `whoami` ? 18:02 < repys> absurdistani: smartctl doesn't show any bad sector but something is happened :( 18:02 < repys> absurdistani: so I wanted to know how can I check the real state of these disks? 18:03 < Trel> BCMM and djph, I know my private key though it's in putty's ppk format. I want ot know WHICH line in authorized_keys I would need to delete, to make it so the key I'm currently connected with will no longer work. I have five separate lines. I do not know which pubkey corresponds to the private key I'm currently using. 18:03 < noway96> I have three USBs flashed with same software. Two of them booted successfully. One of them 'mounting /dev/loop0 failed invalid argument'. I then copied the rootfs over to the failing drive and it booted. Does that mean that the USB wasn't ejected properly before which was causing the issue? 18:03 < BCMM> Trel: to confirm, you don't have the public key half in putty? 18:03 < absurdistani> repys: so in that case, try something like... chroot to that mounted location, then try: efibootmgr -c -d /dev/sda -p 1 -l '\efi\kernelNameHere' -L "your linux system" -u "root=/dev/sda2 nomodeset intel_iommu=on" or whatever 18:04 < BCMM> Trel: doesn't a .ppk file contain both halves of the key together? 18:04 < absurdistani> sounds like the efi entry got dead, so you just need to recreate it 18:04 < mnemon> Trel: you can get the public key from the private key via putty-gen or ssh-keygen 18:04 < mnemon> then just remove the matching line 18:05 < repys> I tried to boot another computer with these disks and I get the same error 18:05 < djph> or remove 'em all ... granted that may have the side effect of "nothing works" :) 18:06 < rocketmagnet> hi all, does someone know if there is an opensource calculator for linux where i can write my formulars with my hand ? 18:06 < rocketmagnet> i found many great calculators but i'm always missing this feature 18:06 < ayecee> like with a pen? 18:06 < rocketmagnet> yes 18:06 < rocketmagnet> i've a touchscreen 18:06 < Trel> Guys, what I have access to right now is authorized_hosts on the remote machine, and the currently connected putty session. The key is loaded into the agent, so I don't currently have access to the file itself or I could just convert it to a SSH key and then use ssh-keygen -y to get the pubkey. 18:06 < ayecee> i'm virtually certain that there isn't. 18:07 < BCMM> Trel: sorry, what precisely do you mean by "the key is loaded into the agent"? 18:07 < rocketmagnet> but i'm sure there are such calculators around 18:07 < BCMM> what agent, on which machine? 18:07 < ayecee> me, i'm sure there isn't. 18:07 < rocketmagnet> :( 18:08 < kurahaupo> Trel: ~ ctrl-Z cat keyfile fg 18:08 < noway96> How do you debug initscripts? 18:08 < noway96> also asking, how do you write initscripts 18:08 < noway96> are they just a bunch of bash? 18:08 < ayecee> start at the first line, and go from there 18:08 < revel> I use nano :D 18:08 < ayecee> they are just a bunch of bash, usually 18:09 < noway96> But I suppose with limited commands in bin? 18:09 < ayecee> they don't have to be, but they usually are 18:09 < ayecee> noway96: nope, though you may have to set up your own path 18:09 < revel> Or use absolute paths. 18:09 < BCMM> Trel: one option is to increase the logging level on sshd 18:10 < noway96> K. 18:10 < BCMM> at LogLevel VERBOSE it will give fingerprints of keys used 18:10 < Trel> BCMM: on the machine I'm using pagent is running with the private key loaded in (to which I no longer have access). I'm connected to the remote server, to which I have access to authorized_hosts. I want to remove the line that allows the key I'm currently connected with to ensure it cannot be used again. I don't want to remove any other key. 18:10 < BCMM> Trel: right, but you can reconnect putty? 18:10 < BCMM> Trel: increase loglevel on sshd, restart sshd, reconnect 18:11 < BCMM> Trel: the log will show the fingerprint of the key used 18:11 < Trel> Does it output to a specific log file or the general syslog? Right now I don't seem to have any ssh related log files in /var/log 18:11 < mnemon> Trel: Is it not visible in pageant? also if the ssh-agent forwarding is on you can just ssh-add -L in the session to show the loaded keys. 18:12 < Trel> mnemon: pagent shows what appears to be the fingerprint, not the the actual pubkey, I don't see anything corresponding in authorized_keys. Let me check the output of that command though 18:12 < Trel> Nope, not on 18:13 < rocketmagnet> ayecee: i found something: https://stackoverflow.com/questions/3570220/ocr-lib-for-math-formulas 18:13 < BCMM> Trel: general log, unless you tell sshd otherwise. i think. 18:13 < kcrowboat> systemd-journald[504]: Deleted empty journal 18:14 < kcrowboat> grrr 18:14 < kcrowboat> deleting all useless journald-junk 18:15 * ayecee pictures a neanderthal whacking a server with a wooden cudgel 18:16 * kcrowboat whacks the server with a larger club. . . . . 18:16 < ayecee> percussive maintenance :) 18:16 < kcrowboat> I got the beat !!!! 18:16 < mnemon> Trel: ssh-keygen -lv -E md5 -f .ssh/authorized_keys will get the fingerprints 18:17 < mnemon> or -E sha256 ... but I think pageant still shows md5? 18:18 < Trel> mnemon: that got it, I see which line matches from the fingerprint (it's md5) 18:18 < Trel> thanks! 18:18 < mnemon> np 18:19 < Trel> in authorized_keys, the last 'field' is a comment? 18:20 < ayecee> kind of, yes 18:20 < Trel> I'm labeling the remaining ones to ensure I don't run into this again. 18:20 < ayecee> it's part of the key, but it's a comment within the key 18:21 < Trel> I can modify the key as well right? 18:21 < Trel> For now I'm storing the label in a separate file to reference 18:21 < ayecee> using ssh-keygen, yes 18:22 < Trel> Cool, thanks. 18:23 < Trel> Thanks everyone for the help 18:23 < prussian> you can add environment=gibberish=key_ident as well or use an AuthorizedKeysCommand 18:31 < TyrfingMjolnir> prussian: Where are you from? 18:35 < Richard_Cavell> another newb question: In /proc/cpuinfo I have a whole lot of information. Is this information just a text file, or is it somehow generated when I type cat /proc/cpuinfo ? 18:35 < BCMM> Richard_Cavell: it's not a "real" file, in that it's not actually on the disk somewhere 18:36 < BCMM> Richard_Cavell: what's happening is that there is a special psuedo filesystem mounted at /proc/ 18:36 < Richard_Cavell> ok 18:36 < BCMM> the contents of which is generated on the fly by the kernel 18:36 < Richard_Cavell> Is there anything else in /proc that I might find interesting? 18:36 < lupine> everything, more or less 18:36 < BCMM> Richard_Cavell: yes! those numbered directories inside proc correspond to currently running processes 18:37 < BCMM> Richard_Cavell: each directory name is the PID of a process 18:37 < lupine> consider /proc/meminfo 18:37 < revel> Plus mount info. 18:37 < revel> And info on memory usage. 18:37 < revel> And some kernel stuff. 18:38 < BCMM> Richard_Cavell: for example, you can look at /proc/[some number]/cmdline to get the command that was used to start a given process 18:38 < revel> Or /proc/cmdline for the kernle cmdline :D 18:38 < revel> s/nle/nel/ 18:39 < revel> And /proc/version is a bit more informative than `uname -a` 18:39 < BCMM> s -l /proc/${PID}/fd - list of files a program has open 18:39 < flipper887> That amdgpu-pro drivers botched last night installation of Lubuntu and went to Ubuntu Server 16.04.4 and installing xubuntu-desktop and lubuntu-desktop packages 18:39 < BCMM> ^ls -l that is 18:44 < x0d> /win 2 18:50 < RonWhoCares> I have an rsync question https://askubuntu.com/q/1021667/453336 18:51 < MrBech^imac> I think this is the right channel to ask, but don't crucify me if it's not. I'm having a issue where I can't connect to mysql with "localhost", but "127.0.0.1" works fine. I now know it's the permissions on the /var/lib/mysql/mysql.sock file that is wrong. The permission is: "srwxrwxr-x" and should be "srwxrwxrwx". What causes this problem I do not know. Would someone care to investigate with me and figure this one out? Thanks. 18:51 < ayecee> RonWhoCares: hopefully the forum members can help you with that. 18:51 < ananke> RonWhoCares: you should look into using ssh keys instead of supplying password on command line like that 18:52 < djph> MrBech^imac: it's probably a user permission (in mysql) rather than file permission issue... 18:52 < RonWhoCares> ananke: The reason I pasted the link is because it is too much to paste here 18:52 < MrBech^imac> @djph Let's check. Could you guide me through it? 18:53 < ananke> MrBech^imac: define 'can't' 18:53 < RonWhoCares> ananke: I started with ssh keys. The cron job doesn't have acess to the key 18:54 < MrBech^imac> @ananke Making a PDO connection in PHP, using "host=localhost" I get permission denied. When using "host=127.0.0.1" there is no problem. 18:54 < mawk> that's a classic MrBech^imac 3~ 18:54 < mawk> just use 127.0.0.1 18:54 < ananke> MrBech^imac: check how your access control is set up in mysql 18:54 < mawk> localhost will get through the unix socket 18:54 < djph> MrBech^imac: look at whether or not user@localhost is allowed. 18:55 < MrBech^imac> @mawk What's that mawk? 18:55 < mawk> if some special plugin is enabled 18:55 < mawk> and it is by default with mariadb 18:55 < MrBech^imac> I'm running mariadb 18:56 < mawk> it's some means of communication that uses endpoints living on the filesystem 18:56 < MrBech^imac> So is this something I should fix? or just use 127.0.0.1? 18:56 < mawk> it's not a bug 18:56 < MrBech^imac> I read somewhere that using "localhost" would be "faster" 18:56 < mawk> it's for instance to allow root to use mysql without logging in 18:56 < mawk> mariadb will detect root is on the other side using properties of unix sockets 18:56 < mawk> it uses the unix socket instead of a TCP socket, it's maybe faster but you don't need that 18:57 < SCHAPiE> 127.0.0.1 is the real IPv4 localhost; the name 'localhost' is just a local name bound to that IP. 18:57 < MrBech^imac> Why won't I need "faster"? 18:57 < Celmor> I want to use gzip to compress a log file while it is being written (long-running process in background) but it seems it won't 'flush' content to file till process exits so I guess gzip keeps allocating more memory the more I write and if PC hangs it won't write at all? 18:57 < SCHAPiE> So 127.0.0.1 is technically faster, because it saves a lookup, even locally 18:57 < solidfox> what database do you guys prefer the most? 18:57 < ananke> solidfox: depends on the purpose 18:57 < SCHAPiE> solidfox: postgresql 18:57 < MrBech^imac> @SCHAPiE Hm. I read the opposite. that localhost was faster. But ok 18:57 < MrBech^imac> Thanks for clarifying @mawk. I will forget about it, and just use 127.0.0.1 18:58 < uplime> MrBech^imac: who told you that? 18:58 < ananke> MrBech^imac: and you never even bothered to check how your user permissions are set in your database 18:58 < SCHAPiE> the difference should be negligeable, or maybe not even noticeable 18:58 < MrBech^imac> @uplime I read it somewhere.... That browser tab has since been closed :) 18:58 < SCHAPiE> MrBech^imac: Firefox has a "Undo Closed Tab" function, Ctrl Shift t 18:58 < uplime> i wouldn't believe that site then :) 18:58 < mawk> SCHAPiE: here localhost makes it travel through the unix socket, if the relvant plugin is enabled 18:58 < mawk> that's the faster we're talking about 18:58 < MrBech^imac> @ananke I have checked them before. They are as they should be. 18:59 < uplime> also its not necessary to prefix user names with @ 18:59 < SCHAPiE> mawk: ah, okay, a socket is something completely different 18:59 < ananke> MrBech^imac: which is? you never provided that info 18:59 < MrBech^imac> @SCHAPiE Sorry, it's 100 closed tabs since 18:59 < mawk> MrBech^imac: you don't need faster because it's already fast enough 18:59 < SCHAPiE> MrBech^imac: it keeps a chronological history of closed tabs, not sure how many though :p 18:59 < za1b1tsu> I see online that people save their configuration files in github, nicely organized, but configs are usually split all other the place. How do they do it? 19:00 < uplime> za1b1tsu: symbolic links 19:00 < SCHAPiE> mawk: you could argue that using a socket implies filesystem access as well, which might make it slower 19:00 < uplime> ie, ~/.bashrc is a symbolic link to ~/dotfiles/.bashrc 19:00 < SCHAPiE> depending on the storage 19:00 < mawk> that's just constant time complexity SCHAPiE 19:00 < uplime> at least, thats how I do it 19:00 < mawk> setup times and stuff 19:01 < mawk> you need to measure the variable part 19:01 < SCHAPiE> hmkay, makes sense 19:01 < mawk> and for the variable part I guess the unix socket is faster, but it's just a guess 19:01 < za1b1tsu> uplime, that's interesting and they make a script to link everything toghether? 19:01 < mawk> the canonical idea is to take a very large data string, and measure transmission time 19:01 < uplime> za1b1tsu: what do you mean link everything together? 19:02 < mawk> the larger it is, the less important that constant part of complexity is 19:02 < AnAverageHuman> za1b1tsu: stow is a symlink "manager" if you're looking for programs. 19:02 < za1b1tsu> like I have to link tmux, vim, etc. I create a script to link everything when I'm on a new pc 19:03 < za1b1tsu> AnAverageHuman, is it needed though? Maybe if I want to remove them 19:03 < BCMM> well, most things will be in ~/ or ~/.config/ 19:03 < BCMM> so you could probably just do a couple of ln -s * ../ sorta things 19:04 < autopsy> BCMM, what about system wide configs don't they get copied to home directory too? 19:04 < autopsy> BCMM, that would imply read access on configs for the root filesystem. 19:04 < autopsy> By all or other. 19:07 < uplime> za1b1tsu: i guess. I just do it by hand 19:07 < uplime> it doesn't take very long 19:07 < uplime> or write a short loop in bash to do it for me 19:08 < edge_hay> uplime "I just do it by hand" 19:08 < edge_hay> So do I. Smh 19:09 < uplime> something like `for file in ~/.{bash,vim,nano}rc; do ln -fs "dotfiles/$file" "$file"; done` 19:09 < ancient_dog> should I use systemd or openrc when installing a new distro? 19:09 < edge_hay> imma pretend you pretended to laugh so I can feel better about myself 19:09 < edge_hay> :) 19:10 < edge_hay> eheh 19:10 < uplime> i didn't 19:10 < edge_hay> Lame 19:10 < autopsy> ancient_dog, systemd is way better I've read. 19:10 < za1b1tsu> uplime, thank you 19:11 < uplime> np 19:13 < edge_hay> I'm trying to find a way to stay on topic while telling uplime I don't like them. I wonder how I can do that in a bash script 19:14 < edge_hay> echo 'uplime: I don't like you' 19:14 < edge_hay> Never mind I found how to do it. 19:14 < sauvin> It would be nice if you'd find some *other* channel to do that kind of thing in, though. 19:15 < edge_hay> sauvin may I pm you? Because I actually don't want to get banned from this chan 19:15 < sauvin> We can talk about it in ##linux-ops, if you want. 19:18 < mawk> can a process in a sandbox attack me by stealing all the data from /dev/random or /dev/urandom ? 19:18 < mawk> like by doing it when the system is generating a private key 19:22 < spammcoin> they could bring /dev/random to a crawl urandom doesn't seem affected through /proc/sys/kernel/random/entropy_avail 19:22 < AnAverageHuman> mawk: /dev/random blocks until it has enough entropy to continue. So your key might take longer than usual to create, but should still be secure. 19:23 < mawk> I see 19:23 < Dagmar> That's because you should use /dev/urandom. 19:24 < mawk> so no need to limit the read speed from /dev/random 19:24 < mawk> no security need at least 19:24 < mawk> if I wanted to apply a limit, what would be a sane value ? 19:24 < uplime> who even needs security? 19:24 < Dagmar> Anyone who actually has a problem with /dev/random blocking should use /dev/urandom, because clearly you don't know enough to know why it's perfectly okay that /dev/random blocks, and why it's perfectly reasonable to use the random data from /dev/urandom in pretty much _all_ cases 19:24 < JimBuntu> "security" -> That feeling when you don't who what you don't know. 19:25 < uplime> when you don't who? 19:25 < spammcoin> make an entropy namespace 19:25 < JimBuntu> s/who/know 19:25 < mawk> that doesn't exist spammcoin 19:25 < mawk> yet 19:26 < ayecee> ominous thunder 19:26 < fr0xk> uplime: syntax error in english? 19:26 < fr0xk> :D 19:27 < uplime> fr0xk: hehe 19:27 < sauvin> If you are what you eat, right now, I'm a bowl of guacamole. 19:28 < uplime> im a hot pocket i guess 19:28 < uplime> i always knew i was attractive though 19:28 < fr0xk> LOL 19:28 < sauvin> I've never met a hot pocket I didn't like. 19:28 < uplime> exactly 19:28 < sauvin> NOM 19:28 < edge_hay> uplime please turn off your Pedantic Lock. On your keyboard look for your brain. You must have left it somewhere. 19:29 < edge_hay> I'd be beer I guess 19:29 < sauvin> edge_hay, that's not the kind of comment we generally regard as helping maintain and promote a civil atmosphere. 19:29 < ayecee> it's not even funny 19:29 < sauvin> That one about the "pedantic lock". 19:29 < ayecee> it should be at least one of the two 19:30 < mawk> why are you attacking him edge_hay 19:30 < mawk> you have something you need to compense for ? 19:30 < ayecee> says the white knight 19:30 < edge_hay> mawk lets take that to pm if you want to know 19:30 < mawk> lol 19:30 < autopsy> Wow you guys are on one. 19:31 < sauvin> We're all on something, that seems clear. Not too sure where these mushrooms came from. 19:31 < uplime> those were some funny tasting brownies 19:31 < fendur> I just ate a four leaf clover. Will that do anything? 19:31 < sauvin> Expect gastrointestinal irritation and possibly some fever. 19:31 < edge_hay> good luck with that 19:32 < fr0xk> Stop it, I am hungry already! 19:32 < uplime> yeah but it will be a lucky fever 19:32 < edge_hay> ah 19:32 < fendur> uplime: :> 19:32 < compdoc> I think clover.is poisonous unless youre a rabbit 19:32 < fendur> compdoc: then we're good! 19:32 < edge_hay> Isn't it the other way around? D: 19:32 * sauvin wikis 19:33 < uplime> rabbits are poisonous to clovers? 19:33 < compdoc> but also think youd have to eat more than one to harm yourself 19:33 < edge_hay> Yes uplime 19:33 < edge_hay> That's clearly what I meant 19:33 < BustyLoli-Chan> How hard is it to run windows games on linux in the present day? 19:33 < edge_hay> apt-get wine? 19:33 < fendur> BustyLoli-Chan: much easier than going back in time/to the future to do it! 19:33 < autopsy> BustyLoli-Chan, winehq says Tom Clancy Ghost Recon 2 has a platinum rating. 19:33 < emberquill> Depends on the game, to be honest. Lots of games run well in Wine, lots don't. 19:33 < fendur> sorry. It's the 4LC talking. 19:34 < BustyLoli-Chan> or.. is there a variant of linux that is extremely good at running windows VMs alongside a linux machine. 19:34 < fr0xk> BustyLoli-Chan: I guess its very easy, just click that the autorun.exe and the game will run at 300FPS? I think 19:34 < fr0xk> :) 19:34 < autopsy> BustyLoli-Chan, I think you can get VirtualBox which is pretty good i've read. 19:34 < uplime> last time I used a linux desktop it was a pain 19:34 < BustyLoli-Chan> I always want to switch to linux, and I like using it, but I always hate that there are limited gaming options or some new game my friends will play comes out and then I can't play it 19:34 < uplime> but that was years ago 19:34 < emberquill> I'd say check winehq to see if the game works well in Wine. If not, you can run it in a VM but for graphics-intensive games you'll definitely need to do GPU passthrough of some kind. 19:34 < solidfox> I just remmebered that the reason I stopped using xfce (on xubuntu) was because I don't like that all the windows appear in the exact middle of the screen 19:34 * sauvin runs across a reference to clover honey but not finding anything on clover toxicity 19:35 < solidfox> I wonder if there was some other issue I'm forgetting 19:35 < fr0xk> I am not sure if COD Ghost is in steam 19:35 < fendur> BustyLoli-Chan: Dual boot? 19:35 < edge_hay> I don't really like virtual box :/ I have it installed and it seems to have some input delay for some reasons 19:35 < edge_hay> I do have a bad setup tho 19:35 < edge_hay> System Information: Model: MacBook Pro (13-inch, Retina, Late 2013) • OS: macOS High Sierra (Version 10.13.4, Build 17E199) 19:35 < BustyLoli-Chan> fendur it's always an option... but it's like even more of a pain to switch OS entirely 19:36 < fendur> BustyLoli-Chan: I don't follow "more of a pain" to dual boot? Or to single boot linux? 19:36 < edge_hay> It doesn't show but that's an intel i5 and 8gb of ram :/ 19:36 < BustyLoli-Chan> more of a pain to dual boot 19:36 < edge_hay> Agree with that 19:36 < autopsy> edge_hay, that is exactly what I have too on a laptop. 19:36 < edge_hay> autopsy :c 19:36 < fendur> BustyLoli-Chan: Now I really don't follow. How's that a pain? Unless splitting HD resources is problematic. 19:37 < emberquill> Honestly, just having to use Windows is a pain lol. I wouldn't mind dual booting if my Windows installation booted as fast as my Linux one. 19:37 < edge_hay> wine works great tho but I need windows for class 19:37 < autopsy> BustyLoli-Chan, it's no harder to dual boot than by selecting which operating system you want from the boot loader. 19:37 < edge_hay> doing batch scripting >:c 19:37 < fendur> BustyLoli-Chan: wouldn't you rather have to merely reboot to play a game than to give up on your attempt to use linux exclusively and therefore need to reinstall windows? 19:37 < edge_hay> autopsy it means you need to restart often 19:38 < fr0xk> BustyLoli-Chan: Well, honestly saying, I still don't think that linux is ready enough to that windows gaming laptops 19:38 < fr0xk> replace* 19:38 < fendur> fr0xk BustyLoli-Chan: +1 fr0xk 19:38 < autopsy> edge_hay, not if I don't play a game too often. I like Windows 10 after hating windows for ever. 19:38 < edge_hay> and either have a copy of most things or live without it from one os to another… Bash being the main thing that makes me stay away from a dual boot 19:38 < sauvin> Heh... apparently only *some* clovers are toxic (containing cyanides); most are edible. 19:38 < fendur> sauvin: phew 19:38 < uplime> which one can't eat rabbits? 19:39 < sauvin> There's a sweet clover that grows some kind of mold, I think, that contains something toxic to horses. 19:39 < edge_hay> the non edible ones uplime 19:39 < sauvin> Dunno about rabbits. 19:39 < BustyLoli-Chan> I guess... but with tools like cygwin and such do I really need linux? 19:39 < autopsy> Clovers and rabbits that's definitely off-topic. 19:39 < fendur> BustyLoli-Chan: perhaps not! 19:39 < autopsy> BustyLoli-Chan, cygwin is no Linux replacement. 19:39 < BustyLoli-Chan> also unlike windows linux runs great in a VM which means it sort of gets the short end of the stick 19:39 < edge_hay> cygwin isn't bash :/ 19:39 < uplime> yeah sauvin stop going off topic 19:39 < fendur> BustyLoli-Chan: I'd say if you have to ask, and games are important, then you are likely to do just great without linux. 19:40 < compdoc> windows runs great in a vm 19:40 < BustyLoli-Chan> It's easier to run a vm of linux and have nothing go wrong imo 19:40 < autopsy> cygwin is no bash replacement that's right. 19:40 < edge_hay> Yeah but linux always runs well 19:40 < emberquill> If you think you don't need Linux, then you're probably right. 19:40 < edge_hay> it's stable 19:40 < fendur> BustyLoli-Chan: I believe you have solved it. 19:40 < autopsy> Exactly. 19:40 < BustyLoli-Chan> sadly... I love linux, but yeah 19:40 < sauvin> Windows runs great in a VM for many applications (my heavy duty mechanical engineering stuff, for example), but I don't think a VM is a great place for games. 19:40 < edge_hay> I don't know what vm you're using 19:40 < autopsy> Games use too much GPU. 19:40 < edge_hay> what vm are you using? 19:41 < revel> VirtualBox. 19:41 < sauvin> Cygwin isn't a good substitute for actual Linux because there are a few things Linux does easily that Windows can almost not do at ALL. 19:41 < dgurney> indeed, gaming in a vm will require doing complex stuff like PCIe passthrough 19:41 < BustyLoli-Chan> I do have one question though... is there a super nice virtual mixer for linux? One with virtual audio inputs and outputs and stuff? 19:41 < edge_hay> Cause apart from wine I have troubles with windows 19:41 < edge_hay> Most likely 19:41 < emberquill> When Windows 10 turned on the auto-update-and-reboot setting again after the third time I turned it off, I decided to switch fully to Linux lol 19:41 < BustyLoli-Chan> I've been using voicemeter banana to manage my audio lately, but it'd be cool if linux had a tool that does this. I imagine it'd be way better than windows since the windows audio system is sort of infected deep into the core of windows 19:42 < sauvin> I zapped my Windows partition because I hadn't used it in a couple of years and was running out of disk space. 19:42 < BustyLoli-Chan> https://www.vb-audio.com/Voicemeeter/VoicemeeterBanana.jpg 19:42 < retran> Cygwin is good at giving you a posix with glibc on in a winders box 19:43 < BustyLoli-Chan> also can I send individual program's audio to specific audio devices in linux? That would be another cool feature I'm currently missing on windows 19:43 < sauvin> There were too many things I tried to build on cygwin that just wouldn't go, found out later it's because Windows can't fork. 19:43 < qman__> linux audio is kind of a cluster 19:43 < sauvin> qman__, don't like. It's not a cluster. It's a fustercluck. 19:43 < qman__> there are a lot of options with their own shortcomings and incompatibility 19:43 < sauvin> s/like/lie/; 19:44 < BustyLoli-Chan> damn :/ I do a lot of live streaming so I would find awesome value in like a real time audio mixing capable OS 19:44 < retran> that sounds like a really spealicialized need 19:44 < qman__> look into jack 19:44 < retran> you need some specialized hardware for that 19:44 < sauvin> Also look into a low latency kernel. 19:44 < BustyLoli-Chan> I mean right now windows can do it... so I imagine linux can do it better 19:44 < autopsy> sauvin, lol Windows can't fork that's insane. 19:44 < retran> really, at the point of content creation, most people use a Mac 19:44 < sauvin> Heard that, autopsy. 19:45 < edge_hay> retran um? 19:45 < autopsy> Who can't fork. 19:45 < retran> edge_hay, um? 19:45 * edge_hay puts on sunglasses 19:45 < edge_hay> B) 19:45 * nevodka takes off sunglasses 19:45 < nevodka> :) 19:45 < edge_hay> no, noting 19:45 < sauvin> It's not necessarily true that if Windows or Mac can do it, then Linux can do it better. Mac OS X has a FreeBSD core, and in some ways will be more performant. Morover, apps built for Macs tend to have lots of money and developer time behind them. 19:45 < edge_hay> lol nevodka 19:45 < BustyLoli-Chan> hmmmm... I don't want to pirate mac OS or make a hackintosh 19:46 < qman__> windows also can't force close file handles 19:46 < retran> sauvin, right. Mac's have market inertia and lots of development "sunk costs" 19:46 < BustyLoli-Chan> also mac UI is pretty bleh in my opinion... but I do see fully where you are coming from 19:46 < qman__> if a process won't let go of a file, have to reboot, no other option 19:46 < sauvin> The forking issue forked me up royally. I decided "fork THIS!" 19:46 < autopsy> qman__, so it can lead to memory exhaustion. 19:46 < edge_hay> BustyLoli-Chan what you intended to buy windows? 19:46 < retran> for content creation applications and related hardware periphrials 19:46 < fendur> linux encourages the use of open source software. That seems key to me. :) 19:47 < sauvin> Right now, I might be tempted to snag a Mac for one reason: getting tired of kdenlive and openshot crashing. 19:47 < BustyLoli-Chan> I mean... I bought it forever ago and I've been upgraded to win10 for free so... it's been the pat of least resistance 19:47 < autopsy> Yeah where it's free that is cool. 19:47 < retran> Linux encourages the *consumption* of open source software, certainly 19:47 < lightstream> I'm having one of those moments where I've either made a daft little typo somewhere, or I have gone completely and utterly insane 19:47 < lightstream> can't work out which ATM 19:48 < fendur> retran: not the use? 19:48 < edge_hay> Bought a mac like 5 years ago 19:48 < sauvin> lightstream, so, what's the what? 19:48 < qman__> you've been "upgraded" for "free" 19:48 < fendur> retran: what is the special meaning of "consumption" here? 19:48 < retran> fendur, yes. consumption, in this case, is use 19:48 < lightstream> sauvin: i've just put a new SSD in my PC, and still have the old one mounted 19:48 < autopsy> lightstream, yeah what is going on? 19:48 < fendur> retran: I'm curious why you emphasized it. 19:48 < sauvin> lightstream, and...? 19:49 < lightstream> I want to symlink my steamapps/common folder from the new drive to the old one 19:49 < edge_hay> ln -s? 19:49 < lightstream> and i'm somehow ending up with a symlink that refers to the common folder on the new drive 19:49 < lightstream> so i mounted the old drive at /mnt 19:49 < sauvin> So, what are the device nodes to your old and new drives? 19:49 < BustyLoli-Chan> https://www.reddit.com/r/linuxaudio/comments/3wj1qf/linux_alternative_for_voicemeeter_banana/ this looks promising :O 19:50 < autopsy> lightstream, sounds like you have the source and target links backwards. 19:50 < lightstream> and use this to create the symlink: ln -s /mnt/home/compton/.steam/steam/steamapps/common /home/compton/.steam/steam/steamapps/ 19:50 < lightstream> target is first right? 19:50 < BustyLoli-Chan> So does anyone know of a linux distro aimed at audio production? 19:50 < autopsy> lightstream, target is last. 19:50 < lightstream> really?? 19:50 < retran> fendur, just a point I was thinking for some reason about use vs. development 19:50 < sauvin> Really. 19:50 < edge_hay> BustyLoli-Chan macos 19:50 < autopsy> lightstream, ln -s source target 19:50 < edge_hay> :/ 19:50 < BustyLoli-Chan> >_> 19:50 < edge_hay> Not really open source 19:50 < lightstream> oh ok there's my silly little typo lol 19:50 < BustyLoli-Chan> okay... okay... fair enough.. but yeah 19:51 < sauvin> Congratulations, lightstream, and welcome to the club. You've just been initiated. We've ALL done that. :D 19:51 < autopsy> lightstream, yeah you just have them switched around that's all change the order. 19:51 < edge_hay> arch is lightweight 19:51 < fendur> retran: I see. That's what I thought. I would guess it also, perhaps more subtly, encourages development. But I don't know about that. 19:51 < lightstream> heh ok so not going insane just yet :D 19:51 < sauvin> o/~ welcome to my nightmare... o/~ 19:51 < BustyLoli-Chan> I've used slackware before... I love it, but it's kind of a pain and not easy... I'd be interested to try Arch 19:51 < BustyLoli-Chan> I heard they are similar in many ways, but Arch is probably easier and has sexier package managers and such 19:52 < retran> fendur, it sounded better in my mind before I actually tried articulating it :p 19:52 < edge_hay> I like cinnamon as a gui 19:52 < edge_hay> (Please don't hurt me) 19:52 < autopsy> I was using Slackware 1.2.1 kernel when I started using Linux we had to connect our dial up modems using minicom and pppd 19:52 < RonWhoCares> How do I run "ls" without it using the predfined alias 19:52 < edge_hay> kde is nice too 19:52 < sauvin> edge_hay, I'm a KDE fanatic. People who "ugh" at me, I just tell them "Hey, you don't like it, don't look!" 19:52 < BustyLoli-Chan> looks like... windows 8D 19:52 < emberquill> RonWhoCares: "command ls" 19:52 < ksk> eh, ln its syntax is "target linkname"!? im confused :D 19:52 < BustyLoli-Chan> I actually like KDE 19:52 < AnAverageHuman> RonWhoCares: /bin/ls 19:53 < emberquill> Or that ^ 19:53 < BustyLoli-Chan> I wanted to try gnome, but never found a distro that was naturally capable of it 19:53 < AnAverageHuman> or `$(which ls)` 19:53 < retran> BustyLoli-Chan, is that a joke/troll? 19:53 < autopsy> ksk, no ln -s source target 19:53 < edge_hay> I mean I used to have gnome on ubuntu but that's when they still mailed you free cds 19:53 < edge_hay> 08.04 I think 19:53 < ksk> autopsy: show me the manpage 19:53 < uplime> AnAverageHuman: gross 19:54 < BustyLoli-Chan> no?... slackware never had a gnome package and I never could figure out how to get it working 19:54 < emberquill> Ubuntu's switching back to gnome though (if they haven't already, can't remember if that's a 17.10 or 18.04 change) 19:54 < autopsy> ksk, that's what it is I'm telling you. 19:54 < uplime> RonWhoCares: \ls 19:54 < autopsy> ksk, man ln 19:54 < ksk> I have the manpage open, it says " ln [OPTION]... [-T] TARGET LINK_NAME (1st form) 19:54 < fendur> whoa. people still use slackware? 19:54 < ksk> " 19:54 < azarus> fendur: yes 19:54 < retran> yep. ubuntu is all "done" with their Unity experiment 19:54 < autopsy> ksk, yes link is last. 19:54 < azarus> slackware is kinda cool 19:54 < edge_hay> What's slackware? 19:54 < autopsy> ksk, I said target I meant link. 19:55 < BustyLoli-Chan> slackware teaches you lots, but makes you want to cry imo 19:55 < fendur> azarus: indeed, it will always have a cool factor for me. I started with slackware. 19:55 < azarus> edge_hay: a distro 19:55 < ksk> okay, a play on words :P 19:55 < retran> Slackware is an old nearly-dead distro 19:55 < BustyLoli-Chan> *makes me want to cry imo 19:55 < edge_hay> oh 19:55 < autopsy> azarus, I used Slackware when you had to install from 44 floppy disks. 19:55 < azarus> one of the oldest distros around 19:55 < autopsy> It was a long process. 19:55 < RonWhoCares> uplime: What is the ls syntax that would output /home/rpiggott/filename.txt (I don't want the date, permissions or size to be listed) 19:55 < sauvin> ksk, last time I issued an ln -s, I supplied only ONE argument, the source: ln -s /maison/sauvin/autrui/charges/flics 19:55 < azarus> hehe, I wasn't alive when floppies were around 19:55 < sauvin> ksk, it created a link to that directory in the current directory. 19:55 < uplime> RonWhoCares: \ls /home/rpiggott/filename.txt 19:55 < azarus> and yet I still use them 19:56 < fendur> autopsy: thankfully I didn't start using slackwaare until 1995 when CDs were a thing. 19:56 < retran> if you want a bare-bones Linux, perhaps try Alpine? 19:56 < retran> you can feel all hard-core 19:56 < edge_hay> sauvin you speak French? :o 19:56 < sauvin> Once upon a time, yes. 19:56 < retran> Alpine is nice because it at least has package management of sorts 19:56 < autopsy> fendur, I used Slackware with kernel 1.2.1 in 1995 I think also. 19:56 < edge_hay> I like debian 19:57 < edge_hay> Never tried it with a gui tho 19:57 < autopsy> retran, Fedora has package management. 19:57 < fendur> 1.2.8 or 1.2.13 sounds more like it. But I can't recall. Maybe it was 96. 19:57 < sauvin> I use debian for servers. I like debian more than anything else I've ever tried, but when slapping something onto a NEW computer, I'll often use kubuntu because better OOTB drivers. 19:57 < BustyLoli-Chan> lol fedora 19:57 < azarus> autopsy: fedora has completely different goals than alpine 19:57 < azarus> fedora is *huge* 19:57 < sauvin> Ain't tiny! 19:57 < BustyLoli-Chan> our school used to use red hat on all their machines 19:57 < autopsy> fendur, wow that's crazy I was using usbnet over a USB ethernet cord to get internet on my machine. 19:57 < azarus> a base alpine install will run you like ~130MB 19:57 < BustyLoli-Chan> it was just like... why 19:57 < revel> Alpine has really old kernel versions... 19:58 < azarus> revel: nope 19:58 < revel> azarus: No? 19:58 < dgurney> I use Gentoo basically everywhere atm 19:58 < azarus> revel: use linux-vanilla on edgte 19:58 < azarus> you get the latest long-term kernel 19:58 < revel> azarus: When was the current linux-vanilla kernel built? 19:58 < autopsy> What is a long term kernel? 19:58 < revel> Please, tell me it wasn't in 2017. 19:58 < azarus> revel: a couple of days ago 19:58 < edge_hay> :o 19:58 < revel> Oh, good. 19:58 < sauvin> BustyLoli-Chan, there's an argument for red hat in schools because many major businesses use RHEL. 19:58 < revel> The one before it? 19:59 < retran> autopsy, and so do a billion others distros. Alpine is an example of really small bare-bones distro (fork of Busybox) 19:59 < lightstream> huh i see ksk said what i was about to 19:59 < BustyLoli-Chan> https://upload.wikimedia.org/wikipedia/commons/thumb/2/2d/RedHatHeadquartersRaleigh.jpg/1024px-RedHatHeadquartersRaleigh.jpg :O 19:59 < azarus> retran: they don't have a fork of busybox 19:59 < lightstream> target is first.. right? 19:59 < azarus> they just have a couple of patches 19:59 < fendur> autopsy: yeah. old school. diald with a 14.4k :) 19:59 < autopsy> lightstream, ln -s target symlink 19:59 < edge_hay> Same as cp or mv 20:00 < sauvin> The word "target" confuses me, so I tend to think in terms of "source" and "destination". 20:00 < edge_hay> btw is there like cpdir? :/ 20:00 < autopsy> fendur, that was some good old fashioned work there. Dialup is hard to keep alive with bad telephone lines. 20:00 < ksk> edge_hay: cp -R? 20:00 < retran> azarus, say what? 20:00 < ayecee> i get confused when i think "source" and "destination", so i tend to think in terms of "target" 20:00 < edge_hay> ksk yeah I assumed 20:00 < azarus> retran: busybox is alpine's userland 20:00 < azarus> that's all it is 20:00 < azarus> no fork 20:00 < retran> k 20:01 < autopsy> busybox is small 130 MB install is small. 20:01 < retran> yep 20:01 < autopsy> Yeah what is a fork of busybox anyways? 20:01 < azarus> autopsy: a deviation of busyboxes codebase 20:01 < autopsy> Alpine sounds like trees man. 20:01 < autopsy> azarus, I see. 20:01 < azarus> autopsy: also you asked what a long-term kernel is? 20:02 < azarus> autopsy: go on kernel.org and see "long-term"? 20:02 < dgurney> it's a kernel that is supported for several years 20:02 < autopsy> azarus, yeah never heard that term. 20:02 < dgurney> that's it really 20:02 < autopsy> So it deals with support. 20:02 < lightstream> ok so I may yet have gone insane. From ~/.steam/steam/steamapps, I run "ln -s /mnt/home/compton/.steam/steam/steamapps/common /home/compton/.steam/steam/steamapps/" 20:02 < azarus> 4.14 will be supported for 6 years, i believe 20:02 < revel> azarus: Where'd you get the "a couple of days ago" date? 20:02 < lightstream> and when I check the symlink, I see: "common -> /mnt/home/compton/.steam/steam/steamapps/common" 20:02 < azarus> revel: pkgs.alpinelinux.org 20:02 < revel> The mirror seems to say the linux-vanilla package was last touched on "2017-Nov-24 19:12:34" 20:03 < azarus> your mirro is terrible then 20:03 < lightstream> but when I do "ls common" I get "ls: cannot access 'common': Too many levels of symbolic links" 20:03 < azarus> revel: https://pkgs.alpinelinux.org/package/edge/main/x86_64/linux-vanilla 20:03 < azarus> 2018-04-02 07:48:25 20:03 < revel> azarus: The official one...? 20:03 < revel> https://nl.alpinelinux.org/alpine/latest-stable/main/x86_64/ 20:03 < azarus> i said edge 20:03 < azarus> not stable... 20:03 < RonWhoCares> I don't get why the output of ls to the screen and piping to the file isn't the same. When it pipes the output to the file it includes the file details. I don't want this 20:03 < revel> azarus: When? 20:04 < revel> Oh, "edgte" 20:04 < azarus> whoops 20:04 < autopsy> The actual symlink target goes last. 20:04 < ayecee> RonWhoCares: how are you piping the output to a file 20:04 < delt> getsubopt () is defined in stdlib.h right? 20:04 < delt> on termux i'm getting a linker fail on that symbol 20:04 < autopsy> I keep saying target it's a symbolic link to a directory or file. 20:04 < RonWhoCares> ayecee: command \ls /home/rpiggott/Videos/youTube/*/*.mp4 >> /var/tmp/youTube_mp4_exclude 20:05 < ayecee> RonWhoCares: what output do you get 20:05 < delt> i'd ask in #c but their channel is locked or something 20:05 < azarus> delt: ##c 20:05 < lightstream> The old SSD is definitely mounted at /mnt - "/dev/sda2 on /mnt type ext4 (rw,relatime,data=ordered)" 20:05 < RonWhoCares> ayecee: Please look at the pastebin https://pastebin.com/DLL0B5u6 20:05 < sauvin> Where's the new SSD mountd? 20:05 < RonWhoCares> ayecee: - I've included the results 20:05 < ayecee> RonWhoCares: could you delete the file and try again 20:05 < delt> azarus: oh :) #c was redirecting to #namespace which gives me "cannot send to channel" 20:06 < delt> anyway thanks 20:06 < RonWhoCares> Weird 20:06 < RonWhoCares> ayecee: Why didn't the file get over wrote? 20:06 < ayecee> because >> appends 20:06 < RonWhoCares> So I just didn't scroll down 20:07 < RonWhoCares> Bother! 20:07 < ayecee> :) 20:07 < RonWhoCares> Can I over write the file? 20:07 < autopsy> RonWhoCares, to overwrite the file use only one > this time. 20:07 < RonWhoCares> Bother! 20:07 < RonWhoCares> autopsy: Do you perform counseling? 20:07 < RonWhoCares> :) 20:07 < azarus> this is like, standard UNIX stuff 20:07 < autopsy> RonWhoCares, yeah >> appends to the last line > will create a new file. Or overwrite whichever you think of. 20:08 < autopsy> No I have no interest in counseling command output redirection isn't so hard to understand though. 20:09 < autopsy> RonWhoCares, you are writing a script or what? 20:09 < lightstream> ok it has to be something to do with this higher level symlink: "/mnt/home/compton/.steam/steam -> /home/compton/.local/share/Steam" 20:09 < autopsy> RonWhoCares, cause you could just put it in a bash script. 20:09 < retran> "bash script" or not, it's the same thing 20:09 < lightstream> right, i need to use the resolved symlink as the target, not the symlinked one i guess 20:10 < lightstream> phew, insanity averted 20:11 < autopsy> lightstream, test making symlinks in /tmp symlink a file in temp to another name in the same directory you'll see what you are doing. 20:11 < bray90820> is there a way to do something like du -sh for the sizr of a drive instead of a directory? 20:11 < azarus> bray90820: df -h 20:11 < autopsy> df -h disk free 20:11 < bray90820> Thanks 20:11 < autopsy> Free is not used. 20:11 < autopsy> But df might get you what you need. 20:12 < autopsy> lightstream, did you fix it? 20:12 < lightstream> yep i've sussed it now autopsy - it was because my target was /mnt/home/compton/.steam/steam/steamapps/common but /mnt/home/compton/.steam/steam is a symlink to /home/compton/.local/share/Steam 20:12 < lightstream> so i need to use /mnt/home/compton/.local/share/Steam in my target instead 20:13 < autopsy> lightstream, sounds like you had them reversed at first remember to rm the symbolic link though before you try to link it again. 20:14 < lightstream> actually i did have them the right way round to begin with 20:14 < lightstream> it was just the symlink in the target meant i was creating a symlink to the new drive 20:15 < RonWhoCares> Thank you one and all. It worked. (Not to mention I appreciate the mohawk you gave me in the process :p ) 20:15 < autopsy> lightstream, you were creating a symlink to a symlink that is why it gave you too many levels error. 20:16 < autopsy> RonWhoCares, just remember command ls or \ls and > or >> to append more files to the file. 20:16 < lightstream> yeah, but only because /mnt/home/compton/.steam/steam/steamapps/common resolves to the new drive 20:16 < lightstream> "realpath /mnt/home/compton/.steam/steam/steamapps/common" ---> "/home/compton/.local/share/Steam/steamapps/common" 20:17 < autopsy> I don't know what steamapps is. 20:17 < lightstream> it's the folder where steam stores downloaded games 20:18 < lightstream> i want this symlink so i don't have to re-download them 20:18 < lightstream> (of course i could just move them, but don't want to just yet) 20:19 < lightstream> main thing is the output of realpath above doesn't contain /mnt at the beginning 20:19 < autopsy> lightstream, it redownloads them if they don't exist I guess. 20:19 < autopsy> lightstream, /mnt is where you have the old drive mounted right? 20:20 < lightstream> yep 20:20 < autopsy> Hm. 20:20 < lightstream> (yep to both) 20:20 < amelliaa> Oi, can MPV mono a video with a hotkey or something like that? 20:24 < phogg> amelliaa: not so easily, but you can likely apply a filter to do that 20:30 < mawk> what's a sane value to use when throttling block I/O to a disk ? 20:30 < BCMM> amelliaa: mono like monochrome, or mono like single-track audio? 20:30 < mawk> for a sandboxed process and its children 20:30 < ksk> mawk: that depends? ;) 20:30 < mawk> so for a single process 20:30 < amelliaa> BCMM: single-track audio 20:30 < mawk> yeah I know, that's why it'd be the default value 20:30 < ksk> like, on the throughput of your disk, and on your priorities 20:30 < mawk> the sandboxed process doesn't have to do anything with the disk normally 20:30 < ksk> there is no sane default, and there can be no sane default imho. 20:31 < mawk> but it could maybe load a couple of shared libraries, or write small files to the disk 20:31 < ksk> so dont limit it then? 20:31 < BCMM> amelliaa: -audio-channels=downmix does that, but i'm not sure if you can do it after the video has started playing 20:31 < mawk> yeah but then it allows malicious uses 20:31 < ksk> like, I like cgroups and stuff, but if disk usage is never an issue, why bother? 20:31 < mawk> the sandboxed process can steal all the bandwidth 20:31 < amelliaa> BCMM: Alright, thanks. 20:31 < BCMM> amelliaa: (presumably, downmix means "mix both channels together, and put the result on both speakers". i think there are options for if you want *only* the right or left channel) 20:31 < mawk> the goal is to prevent the process from doing harm 20:32 < ksk> yeah okay, if you want to protect your hardware vs. the VM/box/whatever 20:32 < ksk> it then depends on "how much do I want to give a single box" - in contrast to what your hardware offers 20:32 < mawk> yeah 20:32 < mawk> that would be to find an upper limit 20:33 < ksk> I would assume these kind of settings are more like big company secrets, and there are no defaults around. 20:34 < ksk> also heavily depends on the programms running in the limited context, like you run a wordpress blog and say "uh, I monitored it and apply these settings, they are fine" 20:34 < mawk> yeah 20:34 < mawk> it would be for a programming language sandbox 20:34 < mawk> so really anything could run in it 20:35 < BCMM> amelliaa: actually, looks like that option *might* be aimed at downmixing surround to stereo. an example in the man page recommends --af=pan=1:[0.5,0.5] 20:35 < ksk> also, AFAIK its not just "disk IO" its split up into different things like "read/write over a timeframe" or "number of files craeted" or whatever low-level speech applies here 20:35 < ksk> but its an intersting approach 20:35 < mawk> here I can control either iops or bytes per second 20:35 < ksk> some guy on #lxd fought with lxd to archive something like this and had some problems (like cgroups not recent enough etc) 20:36 < mawk> or use a more complex approach using equal time shares or something 20:36 < mawk> cgroups not recent enough ? lol 20:36 < mawk> cgroups v2 maybe 20:36 < ksk> yeah, that sounds like it. 20:37 < amelliaa> BCMM: My mpv didn't seem to have downmix option available, but --audio-channels=mono or --af=pan=1:[0.5,0.5] do what I was looking for. No need to search for this anymore and thanks for help 20:38 < mawk> but blkio is in cgroups v1 since 2009 20:38 < mawk> that's recent enough for me 20:39 < hasdf> hi. Is there any good tutorial for how to use (GNUPG-) Smartcards? (e.g.: Enigmail, Login via PAM, LUKS-Encryption, etc) 20:40 < hasdf> or one of the examples 20:42 < mawk> let's take a statistical approach, the glibc is around 2 MiB, I want the startup time to be less than 100ms for instance, that would make 20 MiB/s of read speed 20:42 < mawk> that seems high 20:42 < mawk> but it's a dumb estimate anyway, glibc is already loaded for anyone to use 20:43 < za1b1tsu_> anyone running linux on vmware fusion, what is the distro you had no problems with? I tried lubuntu and it crashes a lot 20:43 < mawk> crashes when doing what ? 20:44 * sauvin is running kubuntu on bare metal and gets fried out at openshot and kdenlive constantly crashing 20:44 < revel> "Oh, you know, writing random data to /dev/kmem" 20:44 < mawk> lol 20:44 < ayecee> za1b1tsu_: i haven't had a problem with *ubuntu on vmware. maybe examine the crashes to understand what's happening. 20:44 < phogg> if you're crashing doing video editing that's sadly normal 20:45 < za1b1tsu_> mawk, updating sofware etc 20:45 < sauvin> phogg, some reading around suggests you're 100% spot on. 20:45 < sauvin> za1b1tsu_, how are you doing the "updating"? 20:45 < phogg> sauvin: I keep trying the various editors out there and it keeps being true, except for Blender which I just don't understand 20:45 < za1b1tsu_> sauvin: after installation, I was asked to update the os 20:46 < ayecee> saw an error and ran away 20:46 < za1b1tsu_> software update center?! 20:46 < sauvin> I came to Blender. I saw Blender. Blender's user interface kicked my ass. 20:46 < ayecee> it's broken. make it go so we can be strong. 20:47 < phogg> sauvin: after 30 minutes being unable to do anything I tried a video tutorial, but it was for a slightly different version or something and the UI was entirely different. After another 30 minutes I went back to mpgtx and mencoder for my simple slicing needs. Maybe next year. 20:47 < revel> sauvin: Whoah, now, how do you get that excited for video editing software? 20:47 < sauvin> Blender is "video editing software"? 20:47 < phogg> sauvin: amongst other things 20:47 < sauvin> And no, next time I try taking on Blender, I'm wearing body armour. 20:48 < revel> sauvin: Well, either way, how'd you come to it? 20:48 < Psi-Jack> Oh yay.. Nick change flooding. 20:48 < phogg> blender must, as part of its job, be able to do what amounts to video editing. I would not say it's designed for it 20:48 < mawk> sauvin tripped on freenode's cable 20:48 < revel> Psi-Jack: Netsplit. 20:48 < Psi-Jack> Ahh. Been a while. :0 20:48 < revel> Oh, that too. 20:48 < phogg> at least it's not another matrix sync 20:48 < Psi-Jack> sauvin: Stop driving a forklift through freenode's data centers. :) 20:48 < sauvin> revel, somebody asked something about video editing software, so I loaded up two or three things to poke at to try to help. Wound up "accidentally" plopping some Alice Cooper onto a romantic scene and laughed my ass off. 20:49 < sauvin> Psi-Jack, wasn't me, but some of the guys I share a warehouse with are guys I would NOT want to share a highway with. 20:49 < Psi-Jack> hehehehe 20:50 < revel> sauvin: Yeah, right. What, was it your cat? 20:50 < sauvin> I presume somebody at a DC somewhere farted, and there weren't any open windows. 20:50 * Acheron-a loveS fartS * 20:50 < sauvin> Either that, or Freenode has something running on a Windows machine. 20:51 < revel> Acheron-a: I'd personally prefer it if you kept that to yourself... 20:52 * Acheron-a haS farted a few times behind server racks and left the scene * 20:52 < Psi-Jack> No blaming Garfield this time. :) 20:53 < bipul> When we change the passphrase of RSA private key. Could we able to see the changes inside the file? i.e id_rsa ? 20:54 < Psi-Jack> bipul: Of course. 20:55 < bipul> Psi-Jack, I have tried to figure it out, at head id_rsa but nothing new i have seen. 20:55 < Psi-Jack> Please... English. 20:55 < bipul> It was for me? 20:56 < ayecee> probably yes 20:56 < revel> Considering there's nobody else chatting? Yes. 20:56 < bipul> Ah, okay i will try my best. :) 20:56 < Psi-Jack> "at head id_rsa" makes no sense. 20:57 < bipul> I mean to say here is that, i used the command "head id_rsa" to check the changes after resetting the passphrase. 20:57 < Psi-Jack> ... Why? 20:58 < Psi-Jack> cp id_rsa id_rsa.orig; change passphrase; diff id_rsa id_rsa.orig 21:02 < bipul> I have two VM's VM1 and VM2. At VM1 i have copied my ssh keys too VM2, and then i have reset the passphrase of my private key on VM1. For just curiosity, i thought i should check the changes inside private key file i.e id_rsa. But i found nothing new in it. 21:04 < Psi-Jack> And yet. Did you do what I said? 21:04 < Psi-Jack> Sounds like. No. 21:04 < revel> How'd you check for changes? 21:04 < ayecee> he stopped at the first question 21:04 < Psi-Jack> revel: Somehow, with "head" 21:04 < Psi-Jack> heh 21:10 < bipul> Psi-Jack, I thought DEK-Info will be changed. 21:11 < sine0> when you mkfifo NAME do you have to kill that name when your done 21:12 < revel> sine0: No, you can leave it there forever if you want. 21:12 < revel> It'll take up an inode, I think, and maybe clutter up a directory slightly, but other than that, you can reuse them to your heart's content. 21:13 < sine0> but for the sake of ocd, I mean aspergus, I mean tidyness, how do i remove it 21:13 < revel> rm 21:13 < revel> Or unlink, maybe? 21:14 < Psi-Jack> ... 21:14 < Psi-Jack> sine0: Bad example. :p 21:14 < sine0> Psi-Jack: ? 21:15 < Psi-Jack> Insulting people with actual OCD, or Aspergers, is not very nice. 21:15 < Psi-Jack> So, bad example. :p 21:15 < Manis> Hi. When I create a file "dd if=/dev/zero of=file bs=1M count=100" and then try to extend it (incorrectly) "dd if=/dev/zero of=file bs=1M count=10 skip=100", is the rest of the file still on the disk? 21:16 < Essadon> Hello, I have a relatively simply question. I have this regex https://regex101.com/r/K0Bqtl/2 and it does what it's supposed to do, except that the part after Participants: [NAME] ends with

tag instead of a

-tag. I want both tests to match 21:17 < sine0> Sigh-Jack 21:18 < kurahaupo> Manis: looks like a sparse file - 110MB apparent size, but only occupying 20MB. If you later write something to the middle 90MB its space usage will increase 21:18 < kurahaupo> Manis: if you read from the sparse portion it will return NUL bytes 21:19 < Manis> kurahaupo: Are you sure? skip seeks in the _input_, not in the output. 21:19 < Manis> Because what you describe would actually have been what I wanted. 21:19 < Manis> Essadon: Participants: .+?(?= Oh, right. I read seek=100, sorry. No it just overwrites the existing 10MB 21:20 < Essadon> Manis: Does that match either 'div' or 'p'? Or a random combination? 21:21 < Manis> Essadon: random combination? 21:21 < Essadon> Like dp, dip, dvp 21:22 < Manis> Essadon: Ah, yeah, true. Put the div in quotes ("div"). 21:23 < bipul> If i have to remove the passphrase from the private key. How would i do that? Do i need to use "openssl rsa" ? Is it mandatory to use? 21:23 < hexnewbie> Essadon: You may try DOM and xpath instead 21:23 < Manis> Ehrm, my brain is broken. in () not in "". 21:24 < Essadon> Manis: What in " "? 21:24 < Manis> kurahaupo: But then the rest 90MB should still be there, no? 21:24 < Essadon> group the div? 21:25 < Essadon> Manis: Like Participants: .+?(?=<(div)|(p)) ? 21:25 < Manis> Essadon: Participants: .+?(?=<(div)|<(p)) 21:25 < Essadon> lol 21:25 < Essadon> yeah 21:25 < Manis> yeah, like that, but don't forget the < before the (p). 21:25 < Essadon> Thanks 21:26 < prussian> bipul: openssl rsa -passin something -in your/key 21:26 < kurahaupo> Essadon: why do you use look-ahead in your regex? 21:27 < prussian> where passin can take soemthing like env:envvar, file:file or file:<(something) 21:27 < prussian> etc 21:27 < bipul> Without OpenSSL is it not possible? if not then why? :) 21:27 < kurahaupo> Manis: the extra brackets around (div) and (p) don't serve a useful purpose 21:28 < prussian> it's possible? just a lot of implementation details that would have to implemented by someone... 21:28 < Manis> kurahaupo: Well they do around div, so that all three characters need to match 21:28 < prussian> not a fan of openssl's command line ux either. 21:29 < bipul> Okay thank you prussian . 21:29 < kurahaupo> Essadon: /Participants:\s*(.*?)<(?:div|p)\b/ 21:30 < noway96> I'm getting a segmentation fault for QtWebKit 21:30 < noway96> The debugging procedure should be 1. gereate core file 2. look at it 21:30 < noway96> yes? 21:31 < kurahaupo> Manis: if it's PCRE, which it certainly appears to be, then no. (…) creates a match group, […] creates a match class 21:31 < Manis> kurahaupo: Don't you use a match group in your regex, too? (apart from it being non-capture. 21:33 < Manis> In my dd tests after the failed "skip", the file only has the size 21:34 < kurahaupo> Manis: I write (div|p) not ((div)|(p)) 21:34 < Manis> of the "extension" (the amount I tried to add). But in my data loss case, the file has the size of the used part of the file. How can that be? 21:34 < prussian> fallocate or truncate. why are you growing/making NUL byte files with dd? 21:34 < Manis> kurahaupo: Ah, that you meant. Yeah, correct. 21:34 < kurahaupo> Manis: dd does not truncate the target file 21:35 < Manis> prussian: Because I've always done it this way. There is no real reason. 21:35 < kurahaupo> prussian: there are many uses for such 21:35 < Manis> kurahaupo: What do you mean with "not truncate"? 21:36 < kurahaupo> Manis: it's not the same as the shell >| redirection 21:36 < xz> is anybody using slack here? I have $50 deal and wondering if it's worth to purchase yearly membership ($80 - $50 = $30 at checkout) 21:36 < xz> I'm new to slack, haven't explored it much so far 21:36 < hexnewbie> Slackware is offering yearly membership? 21:36 < xz> slack communicator, ha 21:36 < Psi-Jack> hexnewbie: LOL 21:36 < JimBuntu> xz, Yes, it's worth it 21:37 < Psi-Jack> xz: Got a Linux quesation? 21:37 < kurahaupo> Manis: the shell passes O_TRUNC when it calls open(), dd does not. Read man 2 open 21:37 < Psi-Jack> xz: Rocket.Chat and MatterMost, you can run on your own server, for free. 21:37 < xz> I would imagine many of #linux are using slack, couldn't find better place on freenode really 21:37 < Psi-Jack> Slack though has /nothing/ 0% to do with Linux. 21:38 < xz> Psi-Jack, you are 100% right sir 21:38 < Essadon> kurahaupo: Because I want to do that 21:38 < JimBuntu> Well Psi-Jack They do have a native Linux app (I think it's simply electron) 21:38 < Psi-Jack> JimBuntu: They have an Electron crappy app, sure. 21:41 < kurahaupo> Essadon: I mean instead of using a simple (non-capturing) match group. Typically one would want the content without the "participants" label, so apply a capture group directly to the "content" 21:42 < kurahaupo> Essadon: as I did in my suggested version 21:43 < Manis> kurahaupo: I don't really understand what the pipe does in >|, but then I don't really understand this: https://pastebin.com/Nrs7TVr8 21:44 < noway96> To have unlimited core dump size and change the core path, I'm thinking of calling executables inside rc.local. It's systemV. Objections? 21:45 < Dagmar> Yes. 21:46 < kurahaupo> Manis: oh, it's the same as > but overrides the noclobber option 21:46 < hexnewbie> noway96: Where shall I submit a formal complaint? :) 21:46 < Dagmar> You can additionally change your desktop icons, because they *also* have nothing to do with core files 21:46 < noway96> it'll be ulimit -c inlimited; sysctl -w "kernel.core_pattern="; 21:47 < hexnewbie> ulimit is not an executable 21:47 < Dagmar> noway96: If you just set the things for that in sysctl.conf and reboot you'll be done 21:47 < noway96> oh what is it? 21:47 < kurahaupo> noway96: ulimit is a shell built-in 21:48 < hexnewbie> noway96: A shell built-in. Changes the limits for the current shell and all children (nothing would be a child of rc.local's shell, except for the stuff you launch from there) 21:48 < Dagmar> Where core files land is exclusively the domain of the kernel, man. All you need to do is set them in /etc/sysctl.conf 21:48 < notadrop> Why doesn't this work: "su $current_user vi" 21:48 < Dagmar> Trying to change them in rc.local is silly 21:48 < notadrop> /usr/bin/vi: /usr/bin/vi: cannot execute binary file 21:49 < mawk> because you don't use su like that 21:49 < hexnewbie> noway96: You may look at /etc/security/limits.conf for changing the limits 21:49 < mawk> su -cvi 21:49 < notadrop> maybe I should RTFM 21:49 < noway96> yeah I guess I have no real linux education. How do I get that, they don't teach it in university 21:49 < notadrop> Linux+ cert course? 21:50 < Dagmar> noway96: Reading man pages and docs 21:50 < notadrop> and that^ 21:50 < noway96> I like books 21:50 < nnull> Books like you more 21:50 < hexnewbie> noway96: Lucky you. They taught me that at university. 21:50 < noway96> Also apparently my custom linux image doesn't have sysctl.conf 21:51 < noway96> But it does have sysctl 21:51 < Dagmar> noway96: Well, you might want to fix that then. Most distros use that filename and feed it to sysctl at some point early in the boot process 21:51 < Manis> kurahaupo: I still don't get it? If dd does not truncate, how can the file be smaller afterwards? 21:51 < Dagmar> It might be that it's just not there because there's no point in even having a file if you're not setting in anything in it 21:51 < hexnewbie> You can call ulimit at top of the food^Wcall chain 21:52 < hexnewbie> Or would that be bottom? 21:52 < Dagmar> I wouldn't even bother with trying to set the core size limit unless you think it's ridiculously small or something 21:53 < Dagmar> You kinda want an upper bounds on that one 21:53 < kurahaupo> Manis: ok now I'm confused. Did you use dd … of=file or dd … >file ? 21:53 < Manis> kurahaupo: of=. 21:53 < Dagmar> Manis: THat's almost never going to be a good idea 21:53 < Manis> kurahaupo: Did I ever write I used >? 21:53 < kurahaupo> Manis: what was the size afterwards? 21:54 < kurahaupo> Manis: no, but I've been following 4 conversations so I forget, sorry 21:54 < noway96> hexnewbie what? 21:54 < Manis> kurahaupo: In the example? 10MB. In my real case the combined size of all LVs. 21:54 < Dagmar> dd isn't going to append to an output file, and pointing it at the current directory is a real bad idea 21:54 < noway96> Dagmar, ok just some blogs said to set it to unlimited 21:54 < kurahaupo> Manis: wasn't it already 10MB? 21:55 < Dagmar> noway96: Under certain circumstances it can be _disk consumingly_ large 21:55 < Manis> Dagmar: Oh no, I didn't really type "of=." The . was just a full stop. 21:55 < Dagmar> Ah okay 21:55 < Dagmar> *whew* 21:55 < hexnewbie> noway96: Are you trying to debug daemons, user-started programs (UI), programs started in command-line, or? 21:55 * kurahaupo understood that 21:55 < Manis> kurahaupo: No. First run is 100MB, then I add another 10MB 21:56 < Dagmar> noway96: What is this distro you're working with? 21:56 < noway96> Dagmar, it's custom image using yocto 21:56 < Manis> kurahaupo: Check this example case out: https://pastebin.com/Nrs7TVr8 21:56 < Dagmar> noway96: So... sysv-based? 21:56 < noway96> yep 21:57 < noway96> hewnewbie, it's a program started on bootup by an x11 init script 21:57 < noway96> using the exec call 21:57 < hexnewbie> noway96: Then set the ulimit in that x11 init script 21:57 < noway96> oh 21:57 < hexnewbie> before calling that program 21:57 < Dagmar> noway96; So... You may find it illuminating if you just go into /etc/rc.d and do a `grep -R sysctl` 21:58 < Dagmar> er... with an * at the end 21:58 < Dagmar> noway96: THen you'll see where/if it's pulling in /etc/sysctl.conf 21:58 < Dagmar> It *should* be 21:58 < kurahaupo> Manis: ok, clearly it's my turn to go back and learn something. That's not what I expected. 21:58 < Manis> Me neither. 21:58 < Dagmar> That's kind of a no-brainer for a distro *ahem* with a kenrel 21:59 < hexnewbie> noway96: Hm, I am sorry. That won't work. You wouldn't be allowed to, as your core size would already be limited at login. :/ You need limits.conf in some capacity supported when your user session is started. 21:59 < kurahaupo> Manis: if it's a block device rather than a file, the length is implicitly fixed so O_TRUNC or not makes no difference 21:59 < kurahaupo> Which is what I'm more used to doing with 21:59 < Manis> kurahaupo: Ah. That does make a lot of sense. 22:00 < noway96> hexnewbie, I'll try it anyway 22:00 < Manis> kurahaupo: So dd truncates, too? 22:00 < kurahaupo> Manis: gotta dash. Bb later 22:00 < Dagmar> Not unless you have really gone out of your way with options 22:00 < Dagmar> dd's operation is _very simple_ 22:01 < Dagmar> Without being told otherwise, it'll open files with a simple open() call, write to them with simple write calls, and simply close them when it's done 22:01 < Manis> Yeah, sure. But that just doesn't add up in my head. 22:01 < Dagmar> It won't seek or truncate or sync or anything unless you tell it to explicitly 22:02 < kurahaupo> Dagmar: that's what I expected, so what accounts for Manis' paste? 22:02 < Dagmar> No idea. Lemme scroll up and se 22:03 < hexnewbie> dd has conv=notrunc and conv=nocreat, as well as flag=append, so I assume O_CREAT|O_TRUNC is the default 22:03 < koala_man> Manis: what is unexpected about it? 22:03 < Dagmar> Are we talking about this paste? -> https://pastebin.com/Nrs7TVr8 22:03 < hexnewbie> Although I'm much confused by the existence of separate conf=notrunc flag=append which you're supposed to use together 22:03 < Dagmar> I dont' see anything unusual there 22:04 < Manis> Hmm. Maybe I'm too used to working with block devices and dd. With block devices it's possible to overwrite a part of the device with data. 22:04 < Manis> But with files it apparantly overwrites the whole file each file. 22:04 < koala_man> you're not overwriting a part of the device, you're overwriting the device with parts of the data. maybe you wanted seek=100 instead 22:04 < Manis> s/file$/time/ 22:05 < kurahaupo> hexnewbie: O_APPEND doesn't just mean start writing from the end, it means every write is at the end 22:05 < kurahaupo> koala_man: that was the original question 22:05 < Dagmar> Yeah, "skip" is something that only affects the input 22:05 < Manis> Yes, I did want to seek=, which is the very reason I destroyed my data ;-) 22:05 < hexnewbie> kurahaupo: Ah, right 22:05 < koala_man> cool 22:05 < Dagmar> Manis: Well, that's the other thing "dd" stands for... "data destroyer" 22:06 < Dagmar> *do-do do-doooo* The more you know... 22:06 < Manis> -.- 22:06 < dgurney> am I strange for never having messed up with dd? 22:06 < Dagmar> dgurney; Nope 22:06 < JimBuntu> dgurney, yes 22:06 < koala_man> most people have never messed up with dd 22:06 < koala_man> but then most people have never used it 22:06 < Dagmar> koala_man: I'm not so sure about that 22:06 < hexnewbie> dgurney: One shouldn't have to. But for some reason, sooner or later, you do. 22:06 < JimBuntu> dgurney, Oh, I misread... never messing up, no, not strange. I thought you had said 'never messing with' 22:06 < Dagmar> Okay, if the set is all people rather than just the people who've tried to use dd 22:06 < Manis> OK. It's really unfortunate that seek and skip have such a similar name. 22:06 < koala_man> I like being a pedantic jerk 22:07 < za1b1tsu> can you have accureate temperature monitors if you are running linux in a virtual machine? 22:07 < Acheron-a> https://stratechery.com/2018/the-end-of-windows/ 22:07 < bobdobbs> just out of curiousity... does anyone here use linux video editing tools? I've just started playing around with flowblade. 22:07 < JimBuntu> za1b1tsu, sure. No idea what temp you are measuring or with what... but yes, you can have accuracy in a VM. 22:08 < Dagmar> bobdobbs: That depends. How fluent are you in obscenity? 22:08 < koala_man> ten years ago, kdenlive was the editor with the best features-to-crash ratio 22:08 < hexnewbie> Ah, I also misread that. Yes, if one reads it as ‘data destroyer’ every time they use it, gigantic failures are usually averted. You don't dance around a chainsaw. 22:08 < Dagmar> ^-- note the use of the phrase "features-to-crash" 22:08 < koala_man> it had plenty of features and only crashed every 15 minutes or so 22:08 < za1b1tsu> JimBuntu: I'm using the temperature monitor in the lxpanel (lubuntu, lxde desktop). Do you have a recommandation? 22:08 < Dagmar> I used to literally boot a freakin' rastafarian distro just because it had a stable build of cinelerra 22:09 < Dagmar> Video editing under basically anything you didn't pay a lot of money for is a PITA 22:09 < Manis> Since we seem to have so many dd experts here, let me ask the thing that I really don't understand anymore. So let's say the original file was 32GB (lvm2 PV with ~16GB LVs on it), I tried to extend it after 800MB I notice that I used the wrong option. Why is my file now 16GB? 22:09 < bobdobbs> Dagmar: oh, I use linux. I'm deeply fluent in obscenity. positively shakespearian. 22:09 < Manis> Is it because the VG was still active, then something wrote, extending the file to 16GB? 22:09 < JimBuntu> za1b1tsu, I take it that you are wanting to know MB/CPU/RAM type temps? I don't have any suggestions for those as I don't normally look at them graphically. 22:09 < Dagmar> bobdobbs: Youre going to need it 22:10 < bobdobbs> Dagmar: mind you, one could replace 'linux' in my last statement with any OS really 22:10 < Dagmar> Manis: You tried to extend your PV with _dd_? 22:10 < Manis> Dagmar: Yeah, PV is a loopback device. 22:11 < za1b1tsu> JimBuntu: yeah, cpu temperature, memory usage. I see thank you 22:11 < JimBuntu> Manis, depends on the command you used. I'm betting that you overwrote the file, so it was basically negated to 0 bytes at the start... after a while, before you noticed, it had grown to 16GB. 22:11 < Dagmar> Manis: Probably using dd at all there was needless. All you needed to do was append some zeros. You could have done that with cat 22:11 < bobdobbs> Dagmar: I've also got a mac, upon which I've installed Final Cut Pro. And I'll learn that eventually. But I wanted to play around with something on my linux machine, cos I use it more often and enjoy it more. And I wanted to experiment with something lighter in features. 22:12 < Manis> Dagmar: Or dd if=/dev/zero >> 22:12 < bobdobbs> Dagmar: I've experimented with Openshot in the past. But I found it impossible to line up video and audio tracks with any precision 22:12 < xse_> Hi, I have a deamon launched with seccomp filters, i'm guessing that every process this deamon run gets the same restrictions, what happens if this deamon launches another program using seccomp ? like to my understanding every syscalls made by the deamon are filter and examined, but does the syscalls made by the process launched by the deamon gets filtered twice ? 22:12 < Manis> JimBuntu: But why 16GB? 22:12 < Dagmar> bobdobbs: Well, kdenlive isn't bad. Cinelerra isn't bad. Avidemux isn't bad. No wait... It's kind of horrible, but it's only marginally more horrible than everything else 22:12 < Manis> Also there's still data on the PV. 22:13 < Dagmar> bobdobbs: The big stinky is that ffmpeg is a life-sucking desert of shifting sands that many things depend on and the thing being constantly in flux doesn't help 22:13 < JimBuntu> Manis, I don't know why 16GB... when using dd, the file is growing over time, it depends on when you stopped it and what command you used 22:13 < spammcoin> xse_: yeah it will check both filters, unless the first filter prevents additional filters 22:14 < bobdobbs> Dagmar: I see 22:14 < Manis> JimBuntu: Command is "dd if=/dev/zero of=/var/pv status=progress bs=1M skip=32768 count=10240", I stopped it at 850MB. 22:14 < Dagmar> bobdobbs: I suggest having mulitple options installed and figure out through pain and suffering which one is the least unreliable for whatever editing task you are currently attempting 22:14 < xse_> spammcoin: thanks a lot ! 22:14 < hexnewbie> Blender crashes? 22:15 < Dagmar> Manis: You _really_ need to read the man page bits for _every_ option you attempt to use with dd 22:16 < Dagmar> Manis: skip makes it skip that many bytes forward in the file it's reading _from_ 22:16 < Manis> Dagmar: True that. 22:16 < Manis> Dagmar: I know that by now :-( The question is whether the rest of the file is still there. I could live without the first 900MB. 22:17 < Dagmar> hexnewbie: Not often, but I can see why a few hours with some of these video editing tools might inspire someone to think that simply drawing, keyframing, and compisiting the entire thing over in a virtual CAD system would be easier 22:17 < JimBuntu> Manis, I would have expected that to create a 10GB file. 22:17 < autopsy> Am I banned? 22:17 < JimBuntu> autopsy, from where? 22:18 < dgurney> you are banned from life 22:18 < Manis> JimBuntu: Now I would expect so too. 22:18 < autopsy> JimBuntu, don't know said my nickname was liveuser after a disconnect from freenode then I got disconnected again after I identified said I was banned cannot send to channel. 22:18 < Manis> But the file is neither 10GB not 850MB. 22:19 < bobdobbs> Dagmar: are there any packages that you'd recommend in particular ? 22:19 < Dagmar> bobdobbs: We've mentioned them actually 22:19 < bls> autopsy: if your client auto-reconnects using a made up username, you get locked out of channels that require registration until you switch back and identify 22:19 < Dagmar> bobdobbs: There's fewer than five you'd ever want to deal with right now 22:20 < autopsy> bls yeah it reconnected me but with a wrong username maybe autopsy was still in use on Freenode I don't know how that is happenning. 22:20 < bobdobbs> heh... I'm glad the cinelerra website is honest: "There is a fairly obselete user manual..." 22:20 < autopsy> It only happenned once though. 22:20 < hexnewbie> autopsy: I think that Freenode's SASL identifies you with your real nick and let's you even if you're using your unregistered alt (I could be wrong) 22:20 < bls> autopsy: #freenode would know more about the mechanisms though 22:21 < hexnewbie> autopsy: So if your client supports that, it might be the better option 22:21 < autopsy> hexnewbie, hexchat is taking too long to identify for me when it reconnects. 22:21 < Dagmar> Use the SASL option to connect 22:22 < autopsy> hexnewbie, so I get forwarded to another channel and then it identifies 22:22 < autopsy> I don't know weak wireless signal I guess. 22:22 < sammyg> autopsy, hi! did you manage to make that bootable windows install media? 22:22 < hexnewbie> autopsy: Yes, which is why I (and Dagmar) recommended SASL option, which identifies you during the connection and before you have joined any channels. It will work as long as services are online (even with your alt nick) 22:23 < autopsy> sammyg, no I lost all my partitions all I have is a LiveDVD of Fedora 27 running now. 22:23 < sammyg> oh no :o 22:23 < sammyg> i hope you had backup 22:24 < uplime> autopsy: http://freenode.net/kb/answer/sasl 22:24 < autopsy> hexnewbie, it is not working with my alternative nickname as liveuser apparently but I am using SASL. 22:24 < uplime> autopsy: if you were using sasl you would be authenticated no matter what nick you use 22:24 < autopsy> sammyg, nah I had no backups. 22:24 < autopsy> uplime, maybe I'm confused then I can identify using a wrong nickname? 22:24 < sammyg> OH NO :O 22:24 < sammyg> :) 22:25 < sammyg> sorry about that man 22:25 < uplime> autopsy: yes. the nickname you use is separate from yousr username 22:25 < autopsy> uplime, ahhh you mean cause it asks for username and password in the SASL dialog. 22:25 < uplime> exactly 22:26 < uplime> for example, in znc my nickname is uplime, but my username is nchambers (well it would be if I use sasl plain) 22:26 < autopsy> Then something with a latent connection is keeping me connected on Freenode's connection list while I'm disconnected then. 22:27 < autopsy> Cause it switched it to liveuser which I am not. 22:28 < autopsy> uplime, yeah I think I'm confused then. 22:28 < uplime> hi confused im uplime 22:28 < autopsy> It has never happenned before though just once right now when I asked if I was banned. 22:29 < CashDash123> How would Fallout New Vegas preform under wine 2.0.4? 22:29 < CashDash123> modded hopefully 22:30 < ayecee> try it and see, i guess 22:30 < ayecee> maybe check the winedb 22:34 < dunnousernamefn> Is there a tool to compile an OpenCL kernel, maybe load some stuff into memory, execute and show the results? I don't want to debug my loading code yet 22:47 < VjdfMQ> hey all 22:47 < VjdfMQ> Is it possible to run tcpflow to get only output or input data ? 22:47 < VjdfMQ> To separate them 22:52 < noway96> So it worked hexnewbie 22:53 < noway96> Are microcenter USBs better for bootable USB uses? I don't get the segfault if I use microcenter USBs but if I use Cruzer or some custom USB it crashes 22:56 < cahoots> hi, if i'm running "myVar = true; sem_wait(...);", and i want to guarantee that myVar will be set before sem_wait is run, do i need both asm volatile("": : :"memory") as well as __sync_synchronize? 22:56 < cahoots> in C 22:57 < autopsy> noway96, I had a Cruzer SanDisk that was booting Fedora 27 recently till I erased it. 22:57 < koala_man> cahoots: it's not off topic, but you may have more luck in ##c. we're better at questions like "is xubuntu better than lubuntu?" 22:57 < noway96> autopsy, it boots, but some program in it segfaults 22:58 < Bashing-om> noway96: "check disk for defects' passes ? 22:58 < autopsy> noway96, it seg faults when during boot strap? During start up? 22:58 < cahoots> koala_man, true 22:58 < noway96> no during normal operation 22:58 < noway96> after login 22:59 < GautamS> Hey all 22:59 < autopsy> noway96, you can use strace program to do a system trace and see where it is segmentation faulting on the terminal. 22:59 < GautamS> I'm looking for a linux distribution that executes programs like DOS did. I know it's kinda difficult since all programs use the X window system in some way or another, but I'm hoping something like this has been done before. 22:59 < autopsy> noway96, it seg faults but you don't know what it is? 23:00 < lupine> > all programs use the X window system in some way or another 23:00 < lupine> well, they don't 23:00 < GautamS> I meant to say graphical applications, sorry 23:00 < lupine> just take a random distro, systemctl stop gdm3 (or whatever), go 23:00 < GautamS> For example, the virtual console provided under Debian distributions fail to execute something like glxgears 23:00 < hexnewbie> GautamS: You won't graphical, non-X programs? Why? 23:00 < uplime> autopsy: my program was segfaulting the other day if I passed it a string with a length of 20, but every other size was fine :C 23:00 < lupine> consider elinks with svgalib for instance 23:00 < uplime> GautamS: not all programs use the x window system 23:01 < uplime> you could be using wayland 23:01 < koala_man> GautamS: you want to run graphical programs without X? 23:01 < GautamS> Like under DOS, when you execute something like DOOM, it completely takes over the screen 23:01 < hexnewbie> GautamS: You want fullscreen programs? 23:01 < GautamS> Yes, fullscreen graphical programs executed from a terminal 23:01 < autopsy> uplime, I would use strace to trace it's execution. 23:01 < lupine> GautamS: right, you want svgalib, sdl + directfb, etc 23:01 < lupine> it's totally a thing 23:01 < hexnewbie> GautamS: Programs support F11, Alt+F11, Ctrl+Shift+F and other shortcuts to go fullscreen. 23:01 < uplime> autopsy: it won't do any good 23:02 < uplime> (and strace isn't a segfault tool) 23:02 < koala_man> GautamS: you can do that with directfb and svgalib. SDL based games can often be made to use it 23:02 < autopsy> uplime, why is that? 23:02 < GautamS> hexnewbie: I don't want to do that. I just want a minimal system being able to run a few games 23:02 < uplime> autopsy: well A, I already figured it out, and B, I already knew where it was segfaulting, and C, gdb is better for that 23:02 < GautamS> I have difficulty trying to explain this, but.. 23:03 < lupine> just like DOS, it won't be well-accelerated and you're responsible for every pixel 23:03 < uplime> strace wouldn't tell me why it was segfaulting 23:03 < autopsy> uplime, I am not sure how to use gdb 23:03 < hexnewbie> GautamS: Which specific games? 23:03 < noway96> Bashing-om, I used Gparted to set a new partition table and then a new partition with format 23:03 < hexnewbie> lupine: You could have Glide in DOS. Or anything, really, since the program could ship drivers for anything. 23:03 < GautamS> hexnewbie: Nothing specific, stuff that uses SDL/GL 23:03 < GautamS> Most games would do that 23:03 < noway96> autopsy, I know which program segfaults. It's a browser that takes over an X11 session. But I'm not sure why it segfaults. 23:03 < hexnewbie> GautamS: Do the games work without X, and why is not having X a requirement really? 23:04 < GautamS> I should clarify that. I don't want to remove X, I just don't want a window system 23:04 < koala_man> for nostalgia purposes? 23:04 < GautamS> yess 23:04 < GautamS> you got it :D 23:04 < noway96> I just got a core dump though so let's see what I find 23:04 < hexnewbie> GautamS: I'm not sure you'd achieve fullscreen without a window manager 23:04 < GautamS> Like I want to type in DOOM, or glxgears and see my entire screen filled with that 23:05 < koala_man> GautamS: you definitely can. you can also run dosbox or something in fullscreen to play real dos games 23:05 < hexnewbie> GautamS: matchbox-window-manager may do what you want, though 23:05 < koala_man> yes, there are also X11 wms that will show every program fullscreen 23:05 < autopsy> noway96, a browser that takes over an X11 session? What is that? 23:05 < noway96> Qt5WebKit 23:06 < GautamS> hexnewbie: Yes, matchbox is what I want 23:06 < GautamS> Just if I could now have a terminal + matchbox 23:06 < autopsy> noway96, maybe the disk is corrupted. Small amounts of corruption could do that. 23:06 < hexnewbie> GautamS: Matchbox can only display a single program fullscreen (or nearly fullscreen if a panel exists). From what I can remember from when I used it on my older phone 23:09 * Psi-Jack hands GautamS a box of matches. 23:10 < GautamS> man, I could have a lightweight linux distro and make it run dosbox fullscreen on startup 23:10 < GautamS> but that.. wouldn't be satisfying 23:11 < Psi-Jack> I have a virtual machine host server running Linux, running a virtual machine of Linux, running my BBS software, which runs DOSEMU to run door games. 23:11 < lupine> for a while I ran a VM solely for df 23:11 < lupine> I'd connect to it over SSH to play 23:11 < koala_man> GautamS: you can add a grub entry that boots real DOS 23:12 < Psi-Jack> lupine: df? Disk Free? Seems boring. 23:12 < lupine> we all need a hoppy 23:12 < lupine> hobby* 23:13 < Psi-Jack> Could've set that up over a socket, xinetd or systemd sockets, so all you'd need to do is netcat it. :) 23:14 < jim> a hoppy hobby?! 23:18 < rockdarko> I can get evolution to launch no problem, but the second I install evolution-ews, it won't launch anymore and crashes with a seg fault. Info pulled from journalctl | grep evolution: https://darko.vision/pub/evolution01 23:20 < WalterPeck> what do you guys use for wireframing ideas? 23:20 < phogg> bubblegum and bailing wire in meatspace, like a real man 23:20 < Psi-Jack> rockdarko: it's trying to give you a sign that exchange si bad. 23:20 < phogg> s/bailing/baling/ 23:20 < Psi-Jack> WalterPeck: I use wire. 23:21 < rockdarko> Psi-Jack: I already know that, but that's work e-mail. and oddly enough it works on my personal laptop setup. 23:21 < jim> WalterPeck, in your world, what form does a wireframe take? 23:22 < WalterPeck> Psi-Jack: thanks! 23:22 < MrElendig> rockdarko: drop the grep 23:22 < MrElendig> rockdarko: also you want -b1 23:22 < phogg> WalterPeck: In all seriousness I use a whiteboard, or kolourpaint. But I don't need anything permanent or for sharing with people outside of one room. 23:23 < rockdarko> MrElendig: I'm getting used to journalctl - please bare with me, thanks for the help :) 23:24 < Psi-Jack> rockdarko: Fun.. sig11 segfaulting. 23:24 < WalterPeck> phogg: thank you 23:24 < MrElendig> rockdarko: the "don't grep" is universal, grepping often hides the real interesting bits 23:24 < rockdarko> Psi-Jack: Does it mean I have herpes? 23:24 < Psi-Jack> rockdarko: There is a possible chance of such meanings. 23:25 < rockdarko> MrElendig: Yeah, I guess my output omits all the lines that are relevant and just don't have evolution in em, thats dumb. I just dont know better - now I do thx. 23:26 < rockdarko> so sorry to come so unprepared, but that -b1 thing - it's an option to journalctl? 23:26 < Psi-Jack> rockdarko: We'll have to submit you to quarantine for further studies, anal probing for the aliums, and such. 23:28 < rockdarko> Psi-Jack: My 4 anuses are eagerly waiting! haha. But seriously, sig11, does that point to anything at all? 23:29 < Psi-Jack> rockdarko: Corruption of sorts, often times. Could even mean hardware issues. 23:29 < Psi-Jack> rockdarko: What distro is this? 23:29 < Psi-Jack> And, can't your work just provide you IMAP like sane people? ;) 23:30 * Psi-Jack pets his company for actually running their own postfix+dovecot servers. :) 23:31 < rockdarko> Psi-Jack: Mnajaro, all up to date. 23:31 * ayecee is glad to have outsourced that headache :) 23:31 < rockdarko> I have the same install and the same packages on 3 systems and it's the only one giving me issues. 23:32 * Psi-Jack cringes. 23:33 < Psi-Jack> rockdarko: Well, there's your problem. A crap distro. ;) 23:34 < rockdarko> Let's not get into that distro war now :P 23:34 < ayecee> Psi-Jack _is_ the distro war >:) 23:34 < morf> :))) and don't mention manajaro in front of him :) 23:34 * Psi-Jack gets out some napalm. 23:34 < Li> what is the user's primary group and how does usermod -g groupname username differ from usermod -a -G groupname username 23:34 * rockdarko retreats in panic. 23:35 < Li> my lanugage skills fucks when it comes to man pages 23:35 < ayecee> Li: their primary group is the group they're in according to /etc/passwd. it's the group used when creating files. 23:35 < Psi-Jack> Li: Kindly mind the language, please. 23:35 < MrElendig> Li: primary group is the one that gets used when you eg create a new file or whatever 23:35 < ayecee> you type too slow 23:35 < rockdarko> I'm also using KDE, just in case you really want to hate me, here's all the ammo you need. 23:36 * rockdarko is a KDE fan boi. 23:36 < Psi-Jack> rockdarko: I used KDE since pre KDE 1.0.0. So can't fault you there. 23:36 < ayecee> pfft. KDE is too banal to bash. 23:36 < Psi-Jack> Only that current KDE sucks. :) 23:36 < jim> Li, so, man pages are hard to understand? 23:36 < rockdarko> Psi-Jack: what's your biggest beef with today's KDE? curious here. 23:37 * rockdarko looks up what KDE 1.0.0 looked like. 23:37 < Psi-Jack> rockdarko: Instability, broken widgets constantly, and "activities" vs "desktops". 23:37 < phogg> rockdarko: It looked pretty great for the time. 23:37 < jonascj> I'm fooling around with diceware / passphrases from wordlists and for the fun of it thinking how simple one could generate such passphrases from a shell script. So far "od -v -An -N1 -tu1 < /dev/urandom" can give me 8bit unsigned integers uniformly distributed (0-255), but what I need is either uniform between 1-7776 or 1-6. 23:38 < jonascj> Any better ideas than transforming the 0-255 integers to the range I need using floating point arithmetics with 'bc'? 23:38 < Psi-Jack> When taskbar buttons just randomly start blinking and worse, flickering, and becoming a huge nuisance, and the only resolution to it is logging out and back in. That's unacceptabel. 23:38 < Li> thanks ayecee MrElendig ... sure Psi-Jack 23:38 < Li> please ecuse my irc fart 23:39 < Psi-Jack> wut? 23:39 < jim> do the kde people put their stuff under git? 23:39 < Li> nm 23:39 < Psi-Jack> NetworkManager? 23:39 < hexnewbie> Dunno, with the new KDE I have whole *windows* flickering 23:39 < Psi-Jack> jim: I believe they run their own git server, yes. 23:39 < Psi-Jack> hexnewbie: Yep. That too. 23:40 < Psi-Jack> Not as frequently as the taskbar icons/buttons flickering out of control, but it happens. 23:40 < jim> ok... so I guess people can post problem reports to them? or do they keep them private? 23:40 < Psi-Jack> Ever since I swtched to XFCE4, replacing xfwm4 with i3, I've never had stability issues, and barely having to restart anything. 23:41 < Psi-Jack> jim: They have a bug reporting system. 23:41 < jim> ok 23:41 < hexnewbie> jim: I have reported like a dozen KDE bugs 10 years ago, 2 got fixed after 10 years, 8 I still receive emails with no fix 23:41 < Psi-Jack> Yeah. 23:42 < Psi-Jack> They seem to be more focused on new things, instead of fixing the broken things. 23:42 < jim> hmm, I guess it could be they;re not paying attention 23:42 < Psi-Jack> Of course, not paying attention is really bad for such a huge project. :) 23:42 < rockdarko> phogg: it really did look good actually (kde 1.0.0). I started with early KDE 2.something. That's like 18 years ago. Time flies, holy crap. 23:42 < phogg> I will once again mention e16: stable and lightweight as only software that has not changed noticeably in over 15 years can be. 23:43 < Psi-Jack> Heh. KDE 1.x, and KDE 2.x were freaking awesome. 23:43 < phogg> rockdarko: My first KDE was 1.2, as I recall. 23:43 < phogg> interestingly 1.0 came about almost exactly one month after win98 23:43 * crabmelt just waits out the showstopping bugs "This was fixed, but regressed when the changes were accidentally dropped after SDDM 0.17 was introduced into the archive." https://lubuntu.me/this-week-in-lubuntu-development-1/ 23:43 < hexnewbie> 3.x was good when it stopped crashing (which was around the time it got discontinued) 23:44 < Psi-Jack> KDE 3.x, IIRC, was a freak of disasters. 23:44 < Psi-Jack> Heh yeah, seems I remember correctly. :) 23:44 < crabmelt> lotsa bugs to wait out with kde tho 23:44 < rockdarko> kde 1.2 I think: http://ymusk.sourceforge.net/screenshots/tempest-1.gif 23:44 < phogg> say what you like against kde4 but at least they finally dropped artsd. 23:44 < Psi-Jack> KDE 4 was a disaster, because they put out 4.0.0, and it was pre-alpha. Version numbering failure. 23:45 < jim> you say you're using i3 as the wm for xfce... what's the useful things about having xfce in addition to i3? 23:45 < phogg> rockdarko: looks about right, but super compressed! 23:45 < Psi-Jack> jim: The xfce4-panel, mostly. Session management is there, some key bindings are managed by XFCE instead of by i3. 23:46 < phogg> xfce has some pretty decent utility apps for configuring things, and they play nicely with software which expects GNOME 23:46 < Psi-Jack> xfce4-session handles my startup apps of system-tray apps,w hile i3's config loads in my pre-layout designed workspaces. 23:46 < crabmelt> thunar, mousepad, ristretto, xfburn, xfce4-screenshooter 23:46 < Psi-Jack> mousepad has some issues at times, so I stopped using it. 23:47 < Psi-Jack> Seems when you run multiple mousepad instances, if one crashes, they all crash together. 23:47 < crabmelt> ive custom compiled my own so i can run it as root without the nag banners 23:47 < Psi-Jack> This did not float well with me. :) 23:47 < phogg> Psi-Jack: tried leafpad? 23:47 < rockdarko> I just remember doing as much in command line as possible. When I discovered you could play music on a computer remotly via ssh (with mocp) and download p2p music (with giFTcurs) it was over for Windows. THEN I discovered DE's and fell in love with KDE. The mix of both is ultimate. 23:47 < Psi-Jack> phogg: Hmm.. May've been leafpad that did that, I know one of the two did. :) 23:48 < jim> so i3 is running "inside" the panel? 23:48 < phogg> Psi-Jack: I started using leafpad for gtk2 text editing after gedit got lobotomized and have not had that issue. Maybe it just hates you. 23:49 < jim> li, so, reading man pages is a skill, and you have to practice reading to keep it up 23:50 < Psi-Jack> jim: No, i3wm supports these things called i3-layouts, which are definition files that bind windows to i3s' window-frames. So with this, I load up my browser on workspace 1, email client on workspace 2, in a tiled workspace 8, load up Rambox and Pidgin tabbed together, and below that my Hexchat client. And finally on workspace 6, my gnome-terminal. 23:50 * phogg checks out this mousepad business 23:50 < phogg> oh yeah, it's aggressively single instanced 23:50 < Psi-Jack> phogg: heh. Must've been mousepad that I hated then. :) 23:51 < Psi-Jack> I know when it failed, it was HORRIBLE, and I used it for quick notes, often. Right now, I use gedit for such things. 23:51 < Psi-Jack> Course, back when I Used actual XFCE, I was in a windowed environment, instead of a tiled environment. :) 23:51 < crabmelt> ill admit when i need a 2nd wordwrap editor i fire up beaver. mousepad isnt particularly highperformance 23:52 < Psi-Jack> crabmelt: Does your girlfriend/wife know about this? ;) 23:52 < phogg> when I need a quick buffer for notes it's nano 23:52 < Psi-Jack> phogg: Compsoe email a lot randomly? ;) 23:53 < registeringIsBad> Hey does anyone here know how to fix overscan with xrandr? I've been tinkering with it for a few hours and still cant get it to work. 23:53 < phogg> Psi-Jack: todo lists mostly. I hear you can get dedicated software for that but it's all so clunky. 23:54 < Psi-Jack> phogg: So you use nano to compose email to yourself to take notes and send with alpine? ;) 23:54 < phogg> Psi-Jack: no, I use nano to take notes which remain sitting in its unsaved buffer until I don't need them any more 23:54 < Psi-Jack> But, nano is a mail composer. :0 23:54 < phogg> Psi-Jack: no, pico is a mail composer 23:54 * Psi-Jack grins. 23:54 < Psi-Jack> nano is too. :) 23:55 < phogg> nano is just a pico drop in replacement 23:55 < Psi-Jack> nano is what pico is, differently licensed. 23:55 < phogg> and anyway real men use elm 23:56 < TJ-> do real men suffer Dutch Elm disease too? 23:56 < Psi-Jack> Anyway, all swords sharpened. Time to leave the grind stone for the day. :) 23:56 < phogg> which would be easier if you could still get a working elm 23:58 < registeringIsBad> Does anyone here know how to use xrandr? 23:58 < stefmorino> I do 23:59 < ksk> registeringIsBad: better ask your question, dont ask to ask. 23:59 < registeringIsBad> stefmorino any idea how to fix overscan? I've tried using --scale and --transform, but all that does is zoom in and out, and there is still some things that dont show up on the screen 23:59 < registeringIsBad> ksk i already asked my question, but no one responded, so I asked if anyone knew about the topic to get someone to respond --- Log closed Wed Apr 04 00:00:28 2018