--- Log opened Tue Jun 26 00:00:13 2018
00:02 < bls> might want to check out the LFS book to see if it's got this covered (if that's not how you already got here)
00:04 < publio> How can I debug a systemd oneshot service that just adds an iptables line? The journalctl says it has run, but there's no change
00:04 < xamithan> run it manually and see if it works?
00:05 < MrElendig> post it
00:05 < MrElendig> also remember, systemd units are not shellscripts
00:05 < publio> it works fine manually. Im running it using systemctl start
00:05 < xamithan> Post it then
00:06 < publio> http://paste.debian.net/hidden/da69aaa7/
00:09 < xamithan> Where does that cloudservice start?  After networking?
00:10 < publio> it runs/is running
00:10 < publio> let me post that too
00:11 < publio> http://paste.debian.net/hidden/58c414fa/
00:12 < publio> Its worth noting that due to docker's iptables routing, you don't actually need to open port 80.
00:14 < xamithan> I'm not familiar enough with oneshot but I'd probably just redirect the output and logs to a file on the unit and see what it says
00:15 < xamithan> Or just put the commands in a script and call the script
00:16  * MrElendig mutters something about namespaces and stuff
00:16 < bls> xamithan: sacrilege
00:17 < xamithan> its systemd,  the entire thing is blasphemous
00:17 < bray90820> How would I emulate a keypress from the command line?
00:18 < bls> bray90820: xdotool
00:18 < bray90820> I am connected to a remote computer with SSH and I wanna press the space bar
00:19 < bray90820> bls: I can't install any other software on the remote computer
00:19 < koala_man> bray90820: what is asking for a space?
00:19 < bls> then you're likely not going to be able to do this, because it's going to require software
00:20 < bray90820> I am writing a script to play and pause music
00:20 < bls> then why not use a music player that allows remote control instead of simulating keypresses over ssh?
00:20 < searedvandal> that's too easy
00:21 < bray90820> I might need to do that
00:22 < rascul> maybe mpd is relevant here
00:22 < rascul> https://www.musicpd.org/
00:23 < searedvandal> if you want to press the space bar badly enough, you can probably make it work with /dev/uinput and some C code
00:23 < bray90820> Kinda odd topic here but What I kinda want is something that can be controlled with a web interface and installed on a PI
00:23 < bls> assuming he's already got the headers, a compiler, and the know-how
00:23 < bray90820> I found out the space bar doesn't work anyways
00:23 < bls> bray90820: like kodi?
00:24 < bray90820> KODI doesn't work in my case
00:24 < bray90820> I'll find something
00:25 < searedvandal> like libresonic?
00:25 < twainwek> but why can't you install anything if its your pi?
00:26 < bray90820> I can install stuff I just couldn't install anything on the OS I was using RuneAudio
00:27 < Henry1511> hey guys, I'm trying to build node from source and getting some errors, anybody who can help me figure out what i'm missing? I'm on debian 9 and trying to build node version 0.9.11
00:27 < Henry1511> http://termbin.com/p8w0
00:27 < twainwek> if there really isn't anything out there that with basic support of playing music on a pi through a web interface, i wouldn't mind writing one
00:28 < bray90820> I mean there is stuff but nothing with multi room audio and keyboard shortcuts
00:29 < twainwek> Henry1511: node as in nodejs?
00:30 < Henry1511> yeah. I just hopped into #nodejs to ask there as well.
00:30 < twainwek> node 0.9 is ancient
00:30 < Henry1511> yeah, i know
00:30 < bls> the issue is that kodi is the go to solution for this, but if kodi is off the table
00:30 < bray90820> Well why do you think I should be using KODi
00:30 < Henry1511> i was able to build the latest version but i'm having problems building the version i need
00:30 < twainwek> i'ts so old i wasn't even born when node 0.9 was release
00:30 < bray90820> Is there some feature I am not seeing?
00:31 < bls> I don't understand why you can't use kodi
00:31 < bray90820> What I wanna do is control everything from windows 10 and play it on various different clients around the house
00:31 < bray90820> Kodi doesn't really support that does it?
00:31 < bls> no, kodi isn't going to run on windows 10
00:32 < bray90820> Well it does run perfectly fine on windows 10
00:32 < searedvandal> max2play
00:33 < searedvandal> hifiberry got a quite extensive article on how to set up a multi room rpi setup
00:34 < bray90820> I'm looking at max2play right now
00:34 < bray90820> And the hifiberry doesn't assume you are using windows as the controler
00:34 < bray90820> You can't really do that with there solution there solution
00:35 < bray90820> Max2Play looks good tho
00:35 < searedvandal> windows should be able to use the web interface
00:36 < bray90820> I'm looking at max2play and we will see how that goes
00:36 < searedvandal> bray90820, read this https://support.hifiberry.com/hc/en-us/articles/205699981-How-to-build-a-multiroom-audio-system-based-on-Raspberry-Pi-and-Hifiberry
00:36 < searedvandal> explains in quite good detail how to set it up
00:36 < bray90820> Again I am looking into max2play and we will see how that goes
00:37 < bray90820> Oh that is for max2play my bad
00:37 < bray90820> Thanks
00:41 < lordvadr> If I have a directory of RPM's, is there a clean way to ask, "should I want to install anything in this repository, are all dependencies satisfiable?"
00:42 < superboot> Hi all. Why are there three packages displayed when I run: apt-cache show kdenlive? They have 3 seperate version numbers. Does this mean there are 3 versions of the package available?
00:42 < Psi-Jack> lordvadr: You could make a repository out of them with one simple command.
00:42 < lordvadr> Basically, if I wanted to syn a repo from, say, Red Hat--and this assumes their repo is complete--is there a good way to make sure I have a good copy.
00:42 < superboot> (kdenlive is just a specific example for demo perposes)
00:42 < lordvadr> Psi-Jack: That's not what I asked. I intend to populate a repo out of them. I want to know that I have all the packages.
00:42 < bls> superboot: which distro? they're likely from different stability levels
00:42 < searedvandal> superboot, the three packages, is it kdenlive kdenlive-data and kdenlive-dbg?
00:43 < Psi-Jack> lordvadr: Did you use an rsync source to copy them from? If not, then you did it wrong.
00:43 < superboot> searedvandal: No, they all are named exactly 'kdenlive'
00:43 < bls> lordvadr: rpm repos should have a db file/index that you can parse and check against
00:43 < superboot> bls: mint
00:44 < lordvadr> Psi-Jack: I know how to sync a repo. I work for Red Hat. That's not what I'm asking.
00:44 < superboot> The versions are 18.04.2-1,  0.9.8-1ubuntu2~ubuntu14.04~ppa1,  and  0.9.6-5ubuntu1
00:44 < Psi-Jack> ...
00:45 < Psi-Jack> I know EVERYTHING because I WORK for the company. LOL
00:45 < lordvadr> Say I sync a repo. I want to ask, given "this repo" and other configured repos, can I satisfy all dependencies?
00:45 < lordvadr> Psi-Jack: RHCE. Does that work?
00:45 < Psi-Jack> No.
00:46 < Psi-Jack> But, good try.
00:46 < lordvadr> Fine then, but I know how to sync a repo. That's not the problem. I'm asking about if there's a tool that will build the dependency graph and check it for loose edges.
00:46 < searedvandal> superboot, are they in different repos?
00:46 < superboot> searedvandal: Perhaps. How do I check?
00:46 < lordvadr> Or do I have to invent it.
00:50 < searedvandal> superboot, there should be a line which starts with section or something like that
00:50 < searedvandal> been a while since I've used apt-cache
00:52 < searedvandal> superboot, probably has something to do with mint adding their own repos on top of ubuntu's repos. checked my two ubuntu servers and all they could find was 1 kdenlive package.
00:53 < superboot> searedvandal: Ok, that makes sense. Thanks for checking.
00:55 < lordvadr> Psi-Jack: Apparently repoquery has a --resolve and a --recursive, which I could run on every package, but that's what I'm trying to avoid.
00:58 < TheNH813> Anyone know how to disable changing images by horizonal scrolling in viewnior?
00:58 < TheNH813> Because it's so sensitive it makes zooming by scrolling up or down impossible.
01:00 < Elodin> Hello, i'm looking for a way to somehow run update-grub at every shutdown/restart. What whould be the best way of accomplishing that
01:00 < TheNH813> Crontab.
01:01 < lordvadr> Elodin: Why on shutdown? Why not on installation of a new kernel?
01:01 < TheNH813> You could specify it with a @reboot trigger. Which means at every reboot.
01:01 < TheNH813> I think there's also one for shutdown, but I'd have to check,
01:02 < Elodin> lordvadr: i would rather on installation of new kernels and on installation of new packages as well
01:02 < lordvadr> TheNH813, Elodin: Also, @reboot that only triggers "after" reboot, if that matters.
01:02 < Elodin> the resoning behind is that i started using btrfs and snapper for snapshots, however for the snapshots to appear at grub i need to update-grub
01:03 < searedvandal> make a systemd unit
01:03 < TheNH813> Hmmm... do any of the configuration files for those allow specifying a command to run after a successful snapshot?
01:04 < TheNH813> (I'm still allergic to SystemD so I can't help with unit files)
01:05 < lordvadr> Elodin: I don't know enough about btrfs snapshots, but I'd imagine there's a way to list all snapshots? What package manager are you using?
01:05 < Elodin> TheNH813: hmmm, i'll look for it. I was just trying to mimic opensuse behaviour in this regard. thought i'd ask about it. I think a crontab entry is good enough, but it would be better as you said yourself if i could have a hook after a snapshot. Ill look a little bit more
01:06 < Elodin> lordvadr: There is a way to list all snapshots, yes. I have been using pacman.
01:07 < lordvadr> Elodin: https://wiki.archlinux.org/index.php/User:Allan/Pacman_Hooks ?
01:07 < TheNH813> Elodin: A scheduled task for a script that checks every hour for new snapshots and runs grub-mkconfig could work.
01:08 < TheNH813> Elodin: But yeah if you can find a way to use a hook that's far better.
01:08 < Elodin> Thanks guys
01:08 < xaeB5> i run windows in a virtualbox VM inside ubuntu and some game i'm trying to run says i have missing openGL support (i think it said opengl 2.0) is there any way to get that working in vm?
01:10 < xamithan> Maybe try to turn on 3d acceleration
01:11 < xamithan> I think it only does opengl 2.1
01:11 < lordvadr> xaeB5: ^^ and make sure you have the guest additions installed. I don't know that 3d works from linux as a host though.
01:11 < lordvadr> Elodin: No problem. Good luck.
01:13 < xaeB5> thanks
01:16 < Dan39> xaeB5: are you doing gpu passthrough thing?
01:17 < xaeB5> Dan39: no
01:17 < xaeB5> really old game so it shouldn't need that.
01:17 < Dan39> yea but 3D/opengl might not work at all unless you do
01:17 < Dan39> not sure tbh
01:19 < dviola> xaeB5: which game?
01:21 < xaeB5> revolt
01:23 < dviola> xaeB5: I would run this game in wine instead, or using RetroArch with the beetle-psx core
01:24 < dviola> wine should work
01:25 < xaeB5> ok thanks
01:25 < dviola> getting GL to work in vbox can be a bit tricky
01:26 < dviola> np
01:45 < dunnousernamefn> How hard is it to tell linux to login to wifi on first boot on a headless server? Do you think that's a bad idea and I shouldn't even try to do that?
01:48 < dunnousernamefn> Would it be as simple as editing an init script to run wpa_supplicant in the iso image?
01:48 < Dan39> dunnousernamefn: or NetworkManager?
01:49 < bls> if you're spinning your own ISOs and you can set it up with WPA2, it should be fine
01:49 < Loshki> dunnousernamefn: the main objection is having to leave an unencrypted password lying around in a file somewhere, which is a security hole.
01:49 < dunnousernamefn> Because it's an init script, can't I just make it only readable by sudo?
01:50 < dunnousernamefn> err root
01:56 < bls> that's no different than saving any other WPA passwords
02:00 < Cache_Money> I have a directory with a few files that I want zipped up (zip -r -X lookup.zip lookup/). When I unzip it it produces that directory and the 3 files inside of it. Is it possible to zip ONLY those 4 files and not the directory? I tried `$ zip lookup.zip ./lookup/file1.csv ./lookup/file2.csv ./lookup/file3.csv` but that again produced a directory with 3 files in it
02:01 < bls> why not just cd into the directory then?
02:02 < Cache_Money> bls: what would the zip command look like then?
02:02 < esselfe> Cache_Money: It's an unzip behavior, you could do it with tar. With windows it ends up with a directory containing an identically named directory!
02:03 < Dan39> he must be trolling
02:03 < bls> cd lookup; zip ../lookup.zip file[1-3].csv
02:04 < Cache_Money> bls: that still created a directory /lookup within the zip file
02:05 < Dan39> -_-
02:05 < Dan39> i called it
02:05 < Cache_Money> esselfe: I need a zip file because I’m uploading files to an AWS Lambda function and I have to manually drag all of the files into the root directory of the function
02:11 < esselfe> Cache_Money: have you tried unzip -d .
02:12 < Cache_Money> esselfe: I don’t really have any control in the unzip process, as I upload a zip file to AWS Lambda
02:12 < bls> a valid zip command has been provided
02:13 < esselfe> oh it doesn't work sorry
02:15 < Cache_Money> bls: Do you think that because I’m on a Mac that your command isn’t working for me?
02:15 < bls> hahaha
02:15 < Dan39> called it.
02:16 < Pentode> lol
02:16 < meyou_> but Mac is linux
02:16 < meyou_> my nephew told me so
02:17 < koala_man> have him tested
02:17 < bls> what's even funnier is that I couldn't find a linux system around with zip installed, so I tried it out on my mac
02:18 < Dan39> lmao
02:18 < Dan39> we need a quotes bot in here
02:21 < CarlenWhite> I tried seeing if I could daily a Mac.
02:21 < CarlenWhite> Got a fresh one from the great whiteness that is the Apple store.
02:22 < Dan39> that's the whole story CarlenWhite ?
02:23 < CarlenWhite> But trying to compile on it was...weird. Homebrew was completely weird to me and trying to compile a program I used produced a program rough around the edges.
02:23 < bls> yeah, homebrew is a bit of a necessary nuisance
02:23 < CarlenWhite> The worst part of it was the price of the damn thing and how I felt compelled to prevent dings and scratches.
02:24 < jim> oh what could it be... "to daily a thing"...could it be like a cat chasing a string?
02:24 < CarlenWhite> I think I was optimistic thinking Apple could charm me.
02:25 < bls> screen, external display management, and battery life on mine are top notch. macOS is whatever.
02:26 < CarlenWhite> I returned it after I got a Lenovo ThinkPad E470 and reverted to Windows. Loads cheaper with more under the hood at $678 + some retrograde RAM because Chrome.
02:27 < jim> we put debian on a thinkpad just the other day
02:28 < Dan39> "we"?
02:28 < bls> my personal machine is a Thinkpad X1, better keyboard (although not as nice as previous thinkpads) and trackpoint are plusses
02:28 < CarlenWhite> It must've required the greatest minds to install such a thing on a machine :v
02:28 < bls> but the screen just gets on my nerves after the retina
02:29 < jim> me and this drummer, the thinkpad is his, he's gonna give it to his daughter and she's gonna use it to learn music
02:29 < CarlenWhite> After the retina? Oh the vibrant displays that kinda blind you?
02:29 < jim> dunno
02:30 < bls> I'm mostly using white on black, so it's irritating when the black portions of the screen "glow"
02:30 < CarlenWhite> Oh, and I have Ubuntu installed on this machine too. Set up so I have to specify it in the EFI boot menu.
02:30 < CarlenWhite> There was some jank to get Secure Boot happy.
02:30 < hatp> If my drive isn't mounted, is it still spinning?
02:30 < bls> hatp: depends on the drive
02:31 < CarlenWhite> bls, There's a name for this. Backlight bleeding?
02:31  * CarlenWhite Googles...
02:32 < bls> CarlenWhite: I believe that's it, yes
02:32 < CarlenWhite> Backlight bleeding.
02:33 < syb0rg> wait, are we talking about thinkpads? Because I hated mine, and also mine had backlight bleed. Effing thinkpads.
02:33 < Dan39> cheap laptop they gave me at work is the worst, you cant see the whole monitor properly at any angle
02:33 < Dan39> have to constantly move head around to see colors properly near different sides
02:33 < Cache_Money> bls: It was still creating a directory on my Mac when I unzipped it but I tried you command and uploaded the zip file to AWS and it was unzipped correctly (no directory). Thanks for the help!
02:33 < bls> cheers!
02:34 < im0nde> Hi, my webcam is not detected and not listed in lsusb. What can I do? hwinfo : https://ptpb.pw/Tphg
02:34 < CarlenWhite> syb0rg, I gave ThinkPads a go since I had a general distaste for what every other brand offered. Lenovo's other product lines suffer a bit from this, but so far the ThinkPads are the least bullshit ridden.
02:34 < im0nde> lsusb https://ptpb.pw/tjNR
02:35 < koala_man> im0nde: is it a usb webcam?
02:35 < syb0rg> CarlenWhite, the least bullshit-ridden among bullshit is still bullshit. Did I mention that the motherboard died? Twice?
02:35 < im0nde> koala_man: it's a notebook, it may be connected internaly by usb, but i dont know
02:35 < syb0rg> Because it did.
02:35 < CarlenWhite> Christ. What sacred ground did you piss on?
02:35 < syb0rg> ....twice.
02:36 < syb0rg> no sacred ground, just regular ground. With a dead motherboard
02:36 < bls> the T, W, and X series (and some of the older Es) thinkpads I've had have all been decent. way better than the Dells or Asus
02:36 < CarlenWhite> My sister has a garbage Dell laptop.
02:36 < CarlenWhite> Damn thing freaks out when it's closed.
02:36 < koala_man> im0nde: it probably has a button, switch or key combo to enable and disable it. when disabled, it'll look unplugged
02:36 < syb0rg> sorry, I made it my duty to crap on that brand ever since my motherboard died. Twice.
02:36 < im0nde> koala_man: I can'
02:37 < im0nde> t find such switch
02:37 < koala_man> im0nde: which notebook model is this?
02:37 < CarlenWhite> I can't make sense of it after throwing every update I could think of.
02:37 < syb0rg> CarlenWhite, freaks out how?
02:38 < CarlenWhite> When the screen is closed and reopened, the graphics decides it didn't want to live anymore.
02:38 < im0nde> koala_man: Medion Akoya e2212t
02:38 < CarlenWhite> LSD trip of colors and so on.
02:38 < syb0rg> that is unfortunate.
02:38 < CarlenWhite> You'll have to restart or do some specific ritual to get it back to the login.
02:39 < CarlenWhite> Said ritual is not yet understood by our greatest scientists our generation can offer.
02:39 < egonsen> pstree shows that xfce4-panel is a direct child of the root process "systemd". why is it not a child of lightdm?
02:39 < CarlenWhite> Nor can it be accurately be reproduced.
02:40 < CarlenWhite> Something about a goat being sacrificed we think.
02:40 < Henry1511> i wanna learn things but it's hard
02:40 < CarlenWhite> Welcome to life, Henry1511.
02:40 < Henry1511> hmm i have an extra "1" at the end of my name
02:40 < bls> egonsen: because lightdm is no longer running
02:41 < Henry151> CarlenWhite: It's nice being here with you all.
02:41 < CarlenWhite> Famous last words.
02:41 < egonsen> bls: pstree is still showing lightdm, with Xorg as its first child
02:41 < bls> egonsen: lightdm just handles your login and DE/WM selection. once it's done doing that, it gets out of the way
02:42 < Henry151> So in this roomful of linux enthusiasts, do many of you build your own software from source on any regular basis?
02:42 < im0nde> koala_man: I really cant find any switch or fn- combo that does that
02:42 < bls> Henry151: only when absolutely necessary
02:42 < syb0rg> Henry151, I assume some of us run Gentoo
02:43 < Henry151> I've only built a couple things from source ever before and it is very hard. I wanna get good at it to where I do it all the time and am totally comfortable with it
02:43 < bls> Henry151: gone through the Linux From Scratch book?
02:43 < syb0rg> ooh, you could try linux from scratch Henry151
02:43 < Henry151> syb0rg: in Gentoo must you build ebverything from source?
02:43 < koala_man> im0nde: google suggests it's Fn+F9
02:43 < Henry151> bls: I've played with linux-from-scratch a little but never got all the way through it
02:43 < im0nde> koala_man: that is disable touchpad on my model
02:44 < syb0rg> Henry151, technically yes but I assume it has a package manager that downlaods and compiles source mostly automatically
02:45 < im0nde> koala_man: by the way, I tried all fn combos with lsusb checking if something appereared
02:45 < Henry151> I just... i have these goals, that are way out in space, like "build a trading-robot that will turn $100 into millions" .... and then I have these smaller goals, like "compile nodejs with dtrace probe points" and "learn how to python" and "figure out how to copy and paste from this damn terminal window"
02:46 < syb0rg> Henry151, you could honestly just grab and compile random github projects
02:46 < Henry151> i figured out that last one recently
02:46 < Henry151> syb0rg: yeah right now I'm trying to do two things; compile nodejs from source and bcc-tools
02:46 < masber> good morning
02:47 < Henry151> masber: good morning.. it's 8:47pm here..
02:48 < im0nde> koala_man: maybe this is useful? https://github.com/burzumishi/linux-baytrail-flexx10/issues/19
02:48 < bls> Henry151: http://www.total-knowledge.com/~ilya/mips/ugt.html
02:48 < CarlenWhite> Henry151, East Coast Best Coast.
02:48 < syb0rg> nah you guys meant west coast
02:48 < syb0rg> I have three extra hours in my day
02:49 < Henry151> that is wonderful
02:49 < Henry151> UGT that is
02:49 < Loshki> Henry151: the utter bliss of *buntu is nothing has to be compiled before use, but it's comparatively easy to build userland from source if you wish.
02:50 < CarlenWhite> How about Universal Good Time?
02:50 < CarlenWhite> I janked up the meme. It's not Good Time it's Good Day.
02:50 < syb0rg> g'day CarlenWhite
02:51  * CarlenWhite commits sudoku.
02:51 < CarlenWhite> Wait,
02:51  * CarlenWhite commits sudo rm -rf /
02:52 < bls> pasting destructive commands is bannable in here, FYI
02:52 < syb0rg> pssh, CarlenWhite, you atttentin ho'. If you meant it you would have a --no-preserve-root
02:53 < CarlenWhite> Ah. Only meant it in jest, but understood.
02:53 < CarlenWhite> syb0rg, Also yeah. I recall there being that argument.
02:54 < syb0rg> yeah I did that one time in a vm but forgot I had shared directories, so I'm not about to forget
02:54 < syb0rg> fortunately backups
02:55 < CarlenWhite> Those are fun moments where you just want to just don't want to live for a few moments.
03:04 < gambl0r3> hello?
03:04 < gambl0r3> im having errors trying to mount my sd card.
03:04 < gambl0r3> Error mounting /dev/mmcblk0p1 at /media/ricky/CC23-17B3: Command-line `mount -t "exfat" -o "uhelper=udisks2,nodev,nosuid,uid=1000,gid=1000,iocharset=utf8,namecase=0,errors=remount-ro,umask=0077" "/dev/mmcblk0p1" "/media/ricky/CC23-17B3"' exited with non-zero exit status 2:
03:04 < gambl0r3> stdout: `FUSE exfat 1.0.1
03:04 < gambl0r3> '
03:04 < gambl0r3> stderr: `BUG: failed to read 512 bytes from file at 0.
03:11 < UNIXO> Hi , i have to servers (centos) i sent file from one to the other using scp , the sent is allowed basing on public key between those servers , i noticied that some time scp command works and other time not (the command appears like it's blocked)
03:11 < UNIXO> i want to know if there is a param in sshd that allow only some connxion in X time , or some thing else ?
03:12 < Pentode> gambl0r3,  well you could a: try another version of fuse, b: download, build source, debug with GDB (backtrace would be helpful) or C: copy the data off the card if possible, zero it out and put a fresh file system on it and see what happens..
03:13 < boingolov> UNIXO: what's the network like between those two?  I'd check MTU settings, if you have a large MTU sometimes weird things can happen
03:14 < boingolov> also if you're doing vxlan, that adds some overhead so you'll need to compensate
03:14 < UNIXO> boingolov: it's not a local network
03:14 < UNIXO> each server is from different providers
03:15 < bls> also, run with some -v's to see what it's actually doing, could be a slow handshake
03:16 < gambl0r3> anyone able to help?
03:16 < boingolov> UNIXO: okay, so that makes one variable unlikely but opens up a world of others ;)
03:17 < Pentode> o_O
03:17 < Pentode> gambl0r3, i just told you what your options were....
03:17 < boingolov> so is it possible that upstream is flooded on one or the other connection?  sometimes if upstream is flooded, ack's don't make it back fast enough to keep timeouts at bay
03:18 < gambl0r3> a: i dont know what fuse is, b: didnt understand anything, c: dont know what you mean by zero it out
03:19 < Pentode> gambl0r3,  https://github.com/relan/exfat
03:20 < UNIXO> boingolov: should i increase some params in sshd_config.?
03:20 < Pentode> use dd to write over the card with "empty" data then repartition and format it and see if that helps
03:20 < gambl0r3> Pentode, and options b, c?
03:20 < boingolov> UNIXO: is it possible your actual scp is saturating upstream?
03:20 < boingolov> what kind of pipes are we talking about?
03:20 < Pentode> b appears out of your skill range.
03:21 < gambl0r3> Pentode, very useful info. thanks
03:21 < boingolov> is there any QoS going on?
03:21 < Pentode> gambl0r3, the problem is either a legitimate bug or for some reason fuse-exfat doesn't like whats on the card for some reason
03:21 < Pentode> so freshening up the card may clear it up. it's worth trying.
03:21 < gambl0r3> Pentode, it was working yesterday and the day before and the day before. didnt do anything different. so i dont know why im having problems all of a sudden
03:21 < boingolov> UNIXO: as far as ssh_config / sshd_config, you can crank up logging verbosity, but I suspect there is something going on networking-wise if you're seeing intermittent timeouts
03:22 < boingolov> might try #networking
03:22 < Pentode> otherwise you could try using a _later_ version of fuse-exfat if possible
03:22 < gambl0r3> Pentode, like i said, it was working and notthing has changed
03:22 < Pentode> gambl0r3, something changed. ;)
03:22 < gambl0r3> Pentode, wasnt by me then
03:22 < UNIXO> boingolov: it's not the timed out but sending the same file some time it takes 2 sec and other time the command continu running without answer
03:22 < Pentode> maybe you forgot to un-mount it, or it didn't unmount cleanly for some reason.
03:23 < gambl0r3> Pentode, possibly.
03:23 < Pentode> in any case there is _something_ that fuse-exfat doesnt like. :|
03:23 < boingolov> UNIXO: that really does sound like an intermittent networking issue
03:23 < boingolov> so not a large file?
03:23 < Pentode> do other cards work ok?
03:23 < UNIXO> boingolov: no small one
03:23 < UNIXO> any idea to debug this?
03:23 < gambl0r3> Pentode, do you want me to try?
03:24 < boingolov> UNIXO: it would be good to figure out which end is closing down the connection / timing out
03:24 < boingolov> crank up logging
03:24 < boingolov> start there
03:24 < boingolov> but again, I suspect this has nothing to do with ssh
03:24 < Pentode> gambl0r3, it's up to you. it'll at least tell you if fuse-exfat is kludged or if its indeed something on the card or the card itself
03:25 < boingolov> of course, sometimes if dns is having issues, and ssh is set to try to do a reverse lookup on the hosts, this can cause issues
03:25 < UNIXO> boingolov: Ok i will keep testing
03:25 < Pentode> also alternatively before wiping the card in question you could try running exfatfsck on it
03:25 < UNIXO> thx
03:25 < Pentode> but i dont think thats going to help. but i guess it cant hurt. usually...
03:25 < boingolov> maybe try UseDNS no in the sshd_config
03:25 < UNIXO> boingolov: i use the IP address directly not a domain
03:25 < boingolov> this is the server side setting
03:26 < boingolov> the server by default will try to do a reverse lookup on the IP
03:26 < meyou_> it prevents the server from doing an rdns lookup
03:26 < meyou_> makes logins super slow if rdns doesn't resolve
03:26 < boingolov> if the dns server local to the sshd is having issues, that can cause extreme pauses
03:28 < gambl0r3> Pentode, you still there?
03:28 < Pentode> im here
03:28 < boingolov> Pentode: so...  you into tubes?
03:28 < gambl0r3> Pentode, sd card works on my other computer
03:29 < Pentode> gambl0r3, other computer uses linux also?
03:29 < gambl0r3> Pentode, obviously
03:29 < Pentode> boingolov, among other things, yes.
03:29 < Pentode> gambl0r3, try updating your fuse package
03:29 < boingolov> heh
03:29 < gambl0r3> Pentode, you mind explaining how?
03:29 < Pentode> what distro?
03:29 < bls> this happens regularly with proprietary FS...proprietary OS writes stuff not supported by OSS drivers
03:30 < gambl0r3> ubuntu
03:31 < Pentode> gambl0r3, apt-get install --only-upgrade exfat-fuse
03:31 < gambl0r3> sudo?
03:31 < Pentode> yes
03:31 < gambl0r3> exfat-fuse is already the newest version.
03:32 < Pentode> if it says you have the latest version, use apt to uninstall it and download the source from that git link i gave you earlier, build and install it.
03:32 < gambl0r3> sudo apt-get uninstall exfat-fuse?
03:33 < Pentode> well download and build it first. if successful then you can take it out and install the new version
03:34 < gambl0r3> Pentode, are you talking about this link?
03:34 < gambl0r3> https://github.com/relan/exfat
03:34 < Pentode> also how are you mounting it?
03:34 < gambl0r3> mounting what
03:34 < Pentode> manually at command line? or is some file manager doing it for you?
03:34 < Pentode> the sd card
03:34 < gambl0r3> it usually does it automagically
03:35 < gambl0r3> i dont understand what you want me to do from this page. https://github.com/relan/exfat
03:36 < Pentode> hrmph
03:36 < Pentode> try mounting it manually and see if it helps.
03:36 < Pentode> google: how to mount sd cards in linux
03:37 < Pentode> you've got much to learn. ;)
03:37 < gambl0r3> Pentode, i've tried google already. all the posts are from 2011-2012
03:37 < Pentode> i have to get back to work, i don't have a lot of time to walk you through everything.
03:37 < gambl0r3> https://askubuntu.com/questions/95391/how-do-i-mount-an-sd-card
03:38 < gambl0r3> Pentode, you're working?
03:38 < Pentode> mount -t exfat /dev/sddevice /some/mount/point
03:38 < Pentode> i was supposed to work earlier but im lazy and i put it off so now i have to get this done in the middle of the night..\
03:39 < gambl0r3> Pentode, ok no worries. sorry to bother you
03:39 < Pentode> your not bothering me, i'm the one who offered to help.
03:39 < Pentode> gambl0r3, all you need google for is to learn how to mount block devices.
03:40 < Pentode> don't google the problem itself. you likely wont get anywhere with that. im suspicious that one of the mount switches that your file manager / desktop environment is using may be why fuse-exfat is bugging out
03:40 < gambl0r3> Pentode, i thought i was starting to annoy you
03:40 < Pentode> thats why i suggested mounting it by hand
03:41 < gambl0r3> gonna try this. https://askubuntu.com/questions/95391/how-do-i-mount-an-sd-card
03:41 < Pentode> no i'll let you know im pretty verbose and blunt
03:42 < gambl0r3> when i do fdisk -l, i dont see my sd card...
03:43 < gambl0r3> wth
03:45 < Pentode> gambl0r3, try using blkid instead
03:45 < Pentode> if it doesn't show up there something is wrong
03:45 < gambl0r3> Pentode, /dev/sr0: LABEL="Angkor Wat" TYPE="udf"
03:45 < Pentode> also, you could: dmesg | tail after inserting it and it should be there as well
03:46 < gambl0r3> no idea whats that
03:46 < bls> sr0 should be a cd drive
03:46 < bls> or dvd drive
03:46 < gambl0r3> ok i dont see my 64gb card
03:47 < gambl0r3> do you tihnk rebooting my computer will help?
03:47 < triceratux> gambl0r3: if you can still see it in another system back it up
03:47 < Pentode> you can try it
03:48 < gambl0r3> triceratux, and then do what?
03:48 < gambl0r3> cute_korean_girl, hi
03:49 < gambl0r3> ok im gonna reboot but my laptop is messed up so theres a chance it may no boot back up but im gonna try anyways
03:49 < gambl0r3> see you
03:50 < gambl0re> currently rebooting
03:51 < gambl0re> Pentode, yes, rebooting fixed it
03:52 < Pentode> cool
03:52 < gambl0re> i dont know why it does that
03:52 < Pentode> some bug apparently
03:52 < gambl0re> like i was telling you before. i changed nada.
03:53 < Pentode> if you want to be helpful and do your part post a detailed bug report to fuse
03:53 < Pentode> many many people don't bother submitting reports and that sucks because it's important.
03:53 < gambl0re> i dont know what i would even write.
03:54 < gambl0re> couldnt mount. rebooted computer. fixed?
03:54 < Pentode> post the error and explain the circumstances that arise when it wont mount. also mention a reboot appears to clear it up.
03:56 < gambl0re> Pentode, will do. thanks for your help!
03:56 < Pentode> anytime
04:01 < elim_garak> whats a good page that will help me understand command concatenation between & && and | ||
04:02 < ayecee> the bash manpage, or possibly the dash manpage for a simpler overview
04:02 < ayecee> the word "concatenation" doesn't make sense there. avoid ten dollar words.
04:03 < jim> elim_garak, & and | are pretty special... && and || are logical combiners
04:04 < elim_garak> || logical or
04:04 < elim_garak> right
04:10 < CyberManifest> Can anyone shed some light on why the shrug emoji https://emojipedia.org/shrug/ doesn't display correctly for me in GUI apps: https://imgur.com/b3ubP0H
04:10 < learningc> How can I grep through all files except files with .tsu extension ?
04:12 < ayecee> find dir -type f ! -name '*.tsu' -exec grep string {} +
04:12 < boingolov> grep -H even
04:12 < CyberManifest> also there seems to be a bug with `man tar | nc termbin.com 9999` it gives error with multiple "mdoc warning: Empty input line #___"
04:12 < boingolov> because otherwise it will suppress file name
04:12 < boingolov> but otherwise wha ayecee said
04:15 < CyberManifest> https://unix.stackexchange.com/questions/199048/mdoc-warning-empty-input-line-xx
04:16 < phogg> I recommend -not over ! if you have GNU find, because it can't ever clash with shell history expansion.
04:16 < ayecee> good tip
04:17 < ayecee> i wasn't aware of that one
04:20 < phogg> CyberManifest: man -7 tar | whatever # you will get approximately what you expect
04:21 < tjsimmons> hi friends
04:21 < tjsimmons> question.
04:21 < tjsimmons> i'm trying to modify my PATH, but.. i don't have an /etc/env(ironment) file, and no .bashrc in my users folder
04:22 < CyberManifest> phogg: right, but that is a one time solution; it's not working as expected for things such as scripts
04:22 < tjsimmons> and i'm at a loss as to where my PATH is actually being set. This is a linux- based NAS (Asustor)
04:22 < CyberManifest> phogg: that syntax works fine for other man pages
04:22 < tjsimmons> trying to install rclone to a custom location and just need to modify said PATH but.. no idea where.
04:22 < phogg> CyberManifest: what do you expect to happen?
04:22 < bls> tjsimmons: there's also /etc/profile, /etc/bash_profile, /etc/shrc, /etc/bashrc, /etc/profile.d, etc, etc
04:22 < tjsimmons> thanks!
04:23 < tjsimmons> should say, this is the first time i've worked without a bash shell.. it's just sh, so no idea what it actually is
04:23 < bls> likely either dash or busybox-sh
04:23 < CyberManifest> phogg: the second not the first after it pulls up man page as expected: https://imgur.com/td5xq5W
04:23 < tjsimmons> my guess is busybox
04:23 < tjsimmons> hadn't really thought about that. i've used busybox before
04:24 < tjsimmons> This whole "upgrade your DSM and it wipes out / (with the exception of your NAS RAID volume)" is really annoying. lol
04:25 < phogg> CyberManifest: You need to be more specific.
04:25 < CyberManifest> mdoc warning: Empty input line #170
04:25 < CyberManifest> mdoc warning: Empty input line #172
04:25 < CyberManifest> mdoc warning: Empty input line #188
04:25 < CyberManifest> mdoc warning: Empty input line #190
04:25 < CyberManifest> mdoc warning: Empty input line #192
04:25 < CyberManifest> mdoc warning: Empty input line #346
04:25 < phogg> CyberManifest: What I mean to say is: Given that you are sending output to a non-terminal how do you expect man to decide on formatting?
04:25 < CyberManifest> mdoc warning: Empty input line #348
04:26 < CyberManifest> http://termbin.com/m7bt
04:26 < phogg> CyberManifest: Don't spam
04:26 < phogg> I know what you're seeing, I just don't know what you think you should be seeing.
04:26 < CyberManifest> phogg: the same way it does for other man pages
04:26 < CyberManifest> http://termbin.com/bv55
04:27 < CyberManifest> like that same command that spamed gave the above results
04:27 < CyberManifest> for cat
04:28 < phogg> CyberManifest: zcat /usr/share/man/man1/tar.1.gz | sed -ne 162 # notice the problem?
04:28 < phogg> er, 172 I mean
04:30 < CyberManifest> phogg: just gives me another error, so I don't really know
04:30 < CyberManifest> phogg: if I knew I wouldn't be here
04:31 < CyberManifest> phogg: you'd have to really dumb it down for this n00b
04:31 < phogg> CyberManifest: Fix the man page, or just run mandoc directly and tell it not to warn you (-w)
04:32 < CyberManifest> phogg: so it's an error with the man page that needs to be reported to the author?
04:32 < phogg> CyberManifest: yes, it is an error in the man page.
04:32 < phogg> but not an error, a warning
04:32 < CyberManifest> phogg: thank you :)
04:33 < phogg> so the author may not care
04:33 < tjsimmons> now, hopefully my damn binary lives through updates
04:34 < tjsimmons> i literally went 6 months without backing up this NAS to B2 and I had no idea because I never checked. >.<
04:34 < CyberManifest> phogg: any ideas on the other issue: https://imgur.com/b3ubP0H ? It's happening across most/if not all GUI text entry locations.
04:35 < phogg> CyberManifest: what issue?
04:36 < CyberManifest> phogg:  ¯\_(ツ)_/¯ doesn't display correctly for me... please see image: https://imgur.com/b3ubP0H
04:37 < phogg> CyberManifest: you mean reverse solidus does not display correctly
04:37 < CyberManifest> phogg: I guess, in GUI only as you can see it's fine in Terminal
04:37 < phogg> CyberManifest: copy the offending character, open gnome-character-map (or your favorite char map) and search for it. WHat does it show?
04:40 < tjsimmons> oh shit. the hell?
04:40 < phogg> if it's really U+005C and just displaying like that then something has gone wrong somewhere in the GTK stack. I'll bet Qt applications don't have the same issue. It would be worth testing that.
04:40 < tjsimmons> this particular NAS doesn't have cron installed? At least.. crontab is around.
04:40 < phogg> if it's not really U+005C then it's something wrong with your IME
04:40 < CyberManifest> phogg: it's worth noting that it doesn't happen in certain situations, like if I have: ¯\_( )_/¯ or it's preceded by a letter like: x ¯\_(ツ)_/¯ or like: ¯\_()_/¯
04:41 < CyberManifest> or if it's highlighted
04:41 < phogg> tjsimmons: "embedded" systems often have some insane configurations
04:41 < tjsimmons> yeah i'm gathering that
04:41 < tjsimmons> and it's not like it has a package manager either
04:41 < phogg> CyberManifest: then it's in the rendering layer, which means your gtk or a lib it depends on has gone sideways
04:41 < tjsimmons> lord have mercy. i just want to schedule regular backups to B2
04:41 < phogg> CyberManifest: Complain to your distro (or rebuild gtk and all of its deps from known-stable versions)
04:42 < phogg> CyberManifest: as I was saying, try a Qt app. If it has the same problem you can probably blame harfbuzz or something else on that layer
04:42 < phogg> do people still use pango?
04:43 < tjsimmons> hm, according to Backblaze they've partnered with Synology for B2 backup.. but I don't see it.
04:43 < CyberManifest> phogg: here is a better description of my issue: https://github.com/signalapp/Signal-Desktop/issues/1690
04:44 < tjsimmons> oh cool. the actual app on the NAS is Cloud Sync. so much easier than what I'm doign on my own NAS. hollllaaaa screw all this "lol this thing doesn't have cron screw you" stuff
04:45 < CyberManifest> phogg: sorry just now saw, I'll look into GTK
04:45 < tjsimmons> literally the extent i have with "stripped down" linux is what I've done for pen testing.. and my linux sysadmin skills are fledgling at bestl
04:45 < CyberManifest> phogg: thanks for all your help
04:46 < jim> CyberManifest, do you see the right thing here? -> \_(ツ)_/
04:46 < phogg> CyberManifest: did you try checking the character against a char map like I said?
04:46 < jim> CyberManifest, do you see the right thing here? -> \_(ツ)_/
04:47 < phogg> CyberManifest: it's important to know if it's doing an input-time conversion or if it's a render-time bug. The linked report does not say
04:47 < phogg> I'm leaning on input-time conversion.
04:47 < phogg> makes more sense that way
04:47 < phogg> CyberManifest: do you use any input method editor?
04:47 < jim> looks like he's nonresponsive
04:48 < phogg> people do that some times
04:48 < phogg> I'm about to go nonresponsive myself for about 8 hours
04:49 < tjsimmons> lol
04:49 < jim> also, whether he's seeing the right thing or not here in irc, that will tell us if it's global
04:49 < CyberManifest> jim: yes it shows correct there
04:49 < jim> ok, so not global
04:50 < maxcell_> hey, i got an Celeron 1600mhz and it's really slow, like, it takes a while to open an youtube page. I'm running an debian-xfce on it. There is another option that improves the performance by A LOT?
04:50 < CyberManifest> phogg: no input method editors that I know of; it happens if I copy and paste or if I type it out manually
04:51 < strive> maxcell_: I'd try a window manager.
04:51 < CyberManifest> jim: only happens in GUI environments; not in Terminal
04:51 < strive> maxcell_: I've had luck in window managers requiring less resources.
04:51 < maxcell_> ok i will try that strive
04:51 < CyberManifest> sorry about not responding was in another chan responding
04:51 < maxcell_> thx
04:52 < strive> Personally, I use i3wm. There are others you could experiment with :)
04:53 < CyberManifest> jim: you can see it happen in it's various places from this: https://imgur.com/b3ubP0H it's hard to make out but it happens in mousepad too.
04:53 < phogg> maxcell_: e16 has not changed in terms of performance in 19 years. Lots of 90s WMs are like that.
04:54 < jim> CyberManifest, yeah, again, if it doesn't happen all the time (incl terminal, irc, whichever), then we can say it's not a global thing
04:54 < phogg> I am betting it's GTK only
04:54 < phogg> and only at input time
04:55 < CyberManifest> jim: it is a global thing graphically all GUI text entries only not happening in Term window
04:55 < jim> (remember that if you're running an xterm or gnome terminal or something like that, that's gui too
04:55 < maxcell_> ok worth try it out
04:55 < jim> CyberManifest, so, you would go forward by finding out what's the difference
04:56 < CyberManifest> phogg: then what about when I type it out... it seems to be a render thing of mixing foreign characters with standard characters as described in that link I provided.
04:56 < phogg> even a GTK-based terminal doesn't do the same sort of input processing as a GTK text input widget
04:56 < CyberManifest> jim: I think phogg may be right about the GTK thing
04:56 < phogg> CyberManifest: it's not a "mix of foreign characters" it's a single Yen sumbol
04:56 < phogg> symbol*
04:57 < phogg> CyberManifest: what is your locale set to?
04:57 < CyberManifest> I must have a missing library or borked config somewhere cause others that use my distro don't have same problem
04:57 < phogg> if it's really trripping into an encoding-detection mode a tweak to locale might prevent that. Maybe.
04:58 < tjsimmons> Also, someone explain to me how the hell this makes sense.. Backblaze B2 requires GLOBALLY (across all their customers) unique bucket names.
04:58 < CyberManifest> phogg: US.UTF-8
04:59 < CyberManifest> phogg: I've already rebuilt local cache a few different times
04:59 < CyberManifest> phogg: sorry, en_US.UTF-8
05:01 < CyberManifest> locale = http://dpaste.com/0789KHW
05:06 < CyberManifest> phogg: ツ is the foreign character I was referring to that triggers it.
05:07 < CyberManifest> without ツ no issue
05:10 < granttrec> Hey all, I want to used named pipes to communicate between two process, how can I set up named pipes so that in one bash script I can read/write to a pair of named pipes, I know a pair will be needed since each will be needed to relay messages, but not sure how to have a read/write in the same bash script for mkfifo?
05:15 < CyberManifest> granttrec: forgive me cause I'm new around here but have you asked in the #bash channel?
05:16 < granttrec> CyberManifest: didn't know there was a bash, though I think named pipes would be fair game for linux?
05:16 < granttrec> #bash
05:16 < CyberManifest> granttrec: not sure, but I know they usually talk about a lot of script stuff over at #bash
05:17 < granttrec> good point
05:23 < jim> granttrec, you're welcome to ask (no guarantees about results)
05:42 < req> right now i have command1; kill <pid>
05:42 < req> so I want a way so that the command1 runs only for 1 hour
05:42 < bls> req: have you read up on the timeout command?
05:43 < CyberManifest> Linux manual entry for timeout, acquired from terminal command: 'man timeout': http://dpaste.com/0SP2YC0
05:44 < req> bls: could you throw me an example
05:45 < bls> timeout 1h command1
05:50 < req> bls: after 60seconds, the job says error
05:50 < req> any idea why
05:51 < req> can we redirect the error using /dev/null
05:52 < req> bls: also does the command1 gets killed on its own after 1h?
05:53 < automatt> Hello
05:54 < automatt> @req hello
05:54 < bls> the behavior of the command is documented in its manual
06:04 < Ekman> Ho
06:04 < Ekman> Ho
06:04 < Ekman> Hi
06:05 < Raed> For a minute there I thought it was Christmas or soemthing
06:05 < Raed> something*
06:05 < storge> learn to spell or die
06:05 < ayecee> por que no los dos
06:06 < Raed> storge: Learn to capitalize and punctuate or die?
06:06 < storge> que lastima, mi tocasdiscos estan descompuesto. too bad, my recordplayer is broken.
06:07  * storge marks this one [SOLVED]
06:17 < maxcell_> do you guys know any linux distro that isn't on beta test versions and worked fine with nvidia?
06:17 < sacules> MX Linux usually does
06:18 < sacules> even has a little program that does install the privative drivers
06:21 < maxcell_> sacules, thank you man
06:22 < maxcell_> seems to be a good distro
06:22 < maxcell_> it isn't ubuntu based right
06:23 < xinming> How do I disable partition detection for specific disk please?
06:23 < xinming> Any ideas? Thanks
06:25 < bls> "disable partition detection"?
06:26 < xinming> bls: Yes,
06:27 < xinming> for now, /dev/sdd will have /dev/sdd1 /dev/sdd2
06:27 < xinming> What I mean is, leave the /dev/sdd as it is
06:27 < jainam> my ubuntu distro is very slow
06:27 < jainam> please help
06:27 < bls> you could try to blacklist the device via udev rules, but what are you actually trying to accomplish by doing this?
06:30 < xinming> bls: some iscsi disk will be used in multipath, So, I don't want them to have partition detected until they are combined as multipath device
06:32 < bls> ah, never had issues with that
06:32 < storge> jainam: define slow
06:33 < sacules> maxcell_: Debian based, pretty light and stable
06:33 < sacules> jainam: get a lighter distro
06:34 < delt> hello everyone
06:34 < delt> quick question: how do i force a server/daemon to write to a new log file, the old one having been "rotated" to a different filename and probably gzipped?
06:35 < bls> delt: the daemon has to implement that functionality
06:35 < delt> on 2 of my server i have /var/log/auth.log and /var/log/syslog at 0 bytes and not moving
06:35 < delt> despite running logrotate manually and even rebooting
06:36 < Triffid_Hunter> delt: most daemons accept sighup or sigusr1 to reopen their logs
06:36 < Triffid_Hunter> delt: and most os-provided logrotate scripts will already have stuff inside to do that
06:36 < delt> Triffid_Hunter: yep, that's what i see in the logrotate config directory
06:36 < Triffid_Hunter> delt: perhaps your server had a config change and your logging daemon is no longer writing to those files?
06:37 < bls> if it's still not working after reboot, it sounds like you've either broken syslogd or you've upgraded to systemd
06:37 < Triffid_Hunter> delt: mine writes to daemon.log and messages here, not syslog
06:38 < delt> Triffid_Hunter: "here" you meant this channel?
06:39 < Triffid_Hunter> delt: no, /var/log/messages
06:39 < Triffid_Hunter> delt: it would be a terribly rude thing to hook one's syslog up to a public channel :P
06:39 < delt> Triffid_Hunter: that's what i was telling myself :O
06:40 < delt> bls: OS is ubuntu 14.04, has always been running systemd ...and i don't remember doing anything that would break syslogd
06:40 < bls> systemd doesn't log to /var/log
06:42 < delt> anyway.... i'll take care of this tomorrow. bed time for now
06:42 < delt> thanks for the assistance guys
06:43 < storge> hey no problem we accept PayPal
06:43 < rexdaemonia> i like money
06:44 < ansraliant> "If you are good at something, never do it for free" - The Joker
06:45 < alpha_> Hello guys.. might not be particular linux question...Is there anyway to set supervisor task based on environment variable? For eg. if env SUPERVISOR=TRUE...run certain task, else none...Thnks
06:45 < supernov1h> how do I figure out what the implicit integer type is when I write a hexadecimal number, with my compiler
06:46 < supernov1h> Oh it's based on the rules of promotion and precedence huh...
06:46 < bls> alpha_: there's a lot of context missing from that question
06:48 < lnnb> -Wconversion
06:49 < supernov1h> someone wrote -0x8000, and I'm trying to figure out what it is, the environment is 16 bit, so I'd assume it looks like in a 32bit gcc, -((uint16_t)(0x8000)) which is 32768, but that doesn't exist, so what is it actually
06:50 < supernov1h> because 32768 is -32768 always
06:52 < bls> purposefully overflowing unsigned ints is a very strange way to set constants
06:52 < lnnb> might be undefined behavior too?
06:52 < bls> yeah, I'd expect some squawking from a compiler about that
06:54 < supernov1h> it's not a gcc compiler unforunately
06:54 < supernov1h> it's some priorietary c89 derivative
06:54 < supernov1h> proprietary*
06:54 < lnnb> what arch is it
06:55 < supernov1h> msp
06:55 < supernov1h> doesn't matter, it's odd code and it was erroneous, but I wanted to characterise what was happening
06:56 < supernov1h> it was supposed to be used in: uint32_t result; result = something; if (result < -0x8000 || result > 0x7FFF) { do something }
07:01 < bls> only seens stuff like that from people that liked to use their "clever tricks" all over the place
07:03 < Dagmar> purposefully overflowing unsigned ints is never a good idea, particularly when setting constants
07:03 < Dagmar> If someone's trying to "test" whether or not they're on a little- or big-endian platform, they should do a *test* and use a macro.
07:03 < Dagmar> Not just make a second error and hope the two cancel out
07:04 < Dagmar> That stuff will _break_ eventually
07:04 < Dagmar> ...and then some poor schlub has to try and figure out what mixture of amphetamines and psychedelics the original coder was on at the time.
07:05 < lnnb> how much ram you got in that thing?
07:06 < Dagmar> alpha_: That would be almost hilariously insecure so, no.
07:07 < Dagmar> PRETEND_I_AM_ROOT=pretty_please passwd root
07:14 < saderror256> hi, is it possible that if i split up my hard drive into multiple partitions, i could dd and put an iso image on one and boot from it?
07:14 < saderror256> i want to install debian, but i dont have a usb or cd/dvd, i only have an sd card  but i cant boot from it
07:15 < saderror256> would this work?
07:23 < horseface> i have 16gb ram on this new laptop and i am only using ~2gb of it
07:26 < TabMasher> Does anyone know of a good utility for optimizing the filling of DVD/BluRay disks?  Like, being able to give it a target capacity size, and having it sort directories of media optimally for that capacity size?
07:28 < alexey-nemovff> TabMasher: other than doing it manually.. I don't know any utility
07:30 < TabMasher> alexey-nemovff: I'm amazed a utility like that would have escapted the linux commuity.  Over a decade I have a program on windows called "BTTB" (Burm to the brim) that does that but I can't get it to work right in WINE using linux directories.  :'(
07:33 < Dagmar> That's because over time it's utility is limited by your ability to figure out which #@$@#$ disc a particular file went on
07:33 < Dagmar> ...so you can potentially save 40 cents.
07:37 < alexey-nemovff> horseface: cool! start copying big files or execute a torrent client.. it will get full rapidly
07:40 < alexey-nemovff> saderror256: splitting your HDD into several partitions and then dd ISO images into them won't clearly work
07:41 < saderror256> aww
07:41 < Pentode> the closest you can get to doing something like that would be to write the image to a hard disk then plug it into a sata/usb adapter. much space would also be wasted, unless you modify the iso to contain a bunch of free space.
07:42 < alexey-nemovff> yeah
08:14 < hexnewbie> TabMasher: Get hard drives instead of DVD/BluRay disks. It's cheaper per byte.
08:24 < mini0n> hoi!
08:38 < Dagmar> hexnewbie: You have a point, but hard drives aren't as likely to survive sitting in a box for a decade
08:38 < Dagmar> I'm backing stuff up to blueray m-discs these days
08:42 < TabMasher> hexnewbie: It's about archiving important data and having them stored in multiple places without a risk of nedia failure.  The per-disk/storage is reasonable and it is very easy to distribute to third parties.
08:43 < TabMasher> Dagmar: Also, exactly my situation and points.
08:47 < ldlework> https://www.youtube.com/watch?v=YpphKzmDiJM
08:53 < notmike> Don't reach, young blood
08:54 < tdn> I am trying to configure xrdp with xfce4. I have put "xfce4-session" in my ~/.xsession and installed xrdp and xfce4 on the server. I then connect from the client via rdesktop and I do get a login screen. When I log in with my user and password, it does seem to authenticate but I get a blank teal screen instead of an actual xfce4 desktop. How do I make the actual xfce4 desktop appear?
08:59 < Dagmar> Make sure the screen resolutions match
08:59 < Dagmar> You might just be looking at the upper-left 80% of the display because it's not going to bother rescaling unless you tell it to
08:59 < Dagmar> Also, RDP sucks
09:10 < no_gravity> Good Morning!
09:10 < no_gravity> How can I make 'find . -iname *hello*' output the modification times of the files?
09:12 < Dagmar> -exec stat {} \; with possibly some arguments passed to stat
09:12 < no_gravity> Yeah!
09:12 < no_gravity> I just arrived at this via pure guess coding: find . -iname *hello* -exec ls -l {} \;
09:13 < searedvandal> -printf %c or %Ck
09:14 < searedvandal> no, it's %t
09:16 < no_gravity> Not sure what that means.
09:16 < no_gravity> But the thing I wrote above works for me.
09:16 < no_gravity> Thanks!
09:17 < searedvandal> according to man find "%t     File's last modification time in the format returned by the C `ctime' function."
09:17 < searedvandal> oh, he quit
09:29 < rukusza> morning o/
09:29 < little> is there a way to always download a latest artefact from the nexus repository?
09:30 < garylabronz> what repo in nexus? pretty sure its dependant on the package api
09:31 < garylabronz> eg python eggs, ruby gems, docker images etc
09:31 < garylabronz> dont think there is a rando global "always get latest" api that nexus has sneaky backdoor like
09:34 < n3xuz> Hello
09:34 < n3xuz> I have a file on my system which I am unable to delete even when I am root
09:34 < garylabronz> give up then
09:34 < garylabronz> chattr -i <file>
09:34 < n3xuz> When I run lsattr it says no such file or directory
09:35 < n3xuz> while reading flags on blabla.mkv
09:35 < garylabronz> vlc blabla.mkv
09:35 < n3xuz> file blabla.mkv says regular file, no read permission
09:36 < n3xuz> I have terminal only on this machine garylabronz
09:36 < garylabronz> du blabla.mkv
09:36 < Triffid_Hunter> n3xuz: fuse mounts can do that
09:37 < n3xuz> garylabronz: sayys 2.5g
09:37 < garylabronz> aws s3 cp blabla.mkv s3://n3xus_personal_videos/
09:37 < garylabronz> oh and --acl public
09:37 < n3xuz> Triffid_Hunter: Its a file on my owncloud server. I have deleted it everywhere but it still floats around in there
09:37 < garylabronz> then give us the link
09:38 < n3xuz> garylabronz: sure
09:38 < n3xuz> Triffid_Hunter: So its a webdav mount
09:41 < Triffid_Hunter> n3xuz: webdav? hmm.. heh, filesystem over http? seems like whatever server you've got it linked to says you can't delete or read the file but it exists
09:42 < M3rd> ..
09:42 < Triffid_Hunter> n3xuz: so maybe you need to unmount, then go direct to whatever server it's on and sort it out there?
09:43 < n3xuz> Triffid_Hunter: I actually deleted it from the owncloud server data folder, and it disappeared from all my sync clients except the webdav mount
09:43 < n3xuz> Triffid_Hunter: I am sshd to the server when I get the error. It seems like its there, but when I try to mv it I get input output, and all attribute checks says it does not exist.
09:44 < Triffid_Hunter> n3xuz: might be cached or something.. seems a bit glitchy for it to fail to recognise a file disappearing on the remote though
09:44 < k_sze[work]> If a process has called putenv to change or set a new env var, is there a way I can see it from outside the process? Assuming I have root access.
09:44 < k_sze[work]> "see" as in "inspect"
09:44 < Triffid_Hunter> k_sze[work]: /proc/pid/environ
09:45 < k_sze[work]> Triffid_Hunter: I just tried. It looks like the new env var is not there.
09:45 < n3xuz> Triffid_Hunter: Yes.. I tried mount / dismount too, but its still there for some reason. Ill try to do a file system check on the mount. I thnik its a bug in davfs or something
09:47 < rascul> /proc/<pid>/environ is only the initial environment when the program starts
09:49 < Triffid_Hunter> k_sze[work]: https://unix.stackexchange.com/a/70636
09:49 < Triffid_Hunter> k_sze[work]: need to hack on it with a debugger apparently
09:49 < k_sze[work]> I see.
09:49 < k_sze[work]> Thanks. :)
09:50 < Triffid_Hunter> because /proc/pid/environ is a fixed-size memory block handed to it by the kernel, so can't be extended.. libc makes a copy in the process's heap for modifications, so need to fish out the dynamic one if you want to see the changes
10:02 < n3xuz> Triffid_Hunter: I thought maybe you want to know what I did to fix this buggy file. I dismounted the webdavfs, deleted the mount folder. Then I removed the parent directory on the server from the apache folder. Made a new folder to mount the dir, resynced all clients, remounted and poof its gone.
10:04 < Triffid_Hunter> n3xuz: heh sounds like a complicated way to delete a file ;)
10:05 < n3xuz> Triffid_Hunter: Indeed :p
10:07 < TabMasher> Dagmar: https://sourceforge.net/projects/filesfitter/
10:20 < tdn> Dagmar, how do I set/see screen resolution? When I start rdesktop on the client, I do not provide any resolution info. It just defaults to something that looks like 1024x768
10:20 < tdn> Dagmar, maybe RDP sucks, but surely not as much as pure X forwarding over SSH? That is practically unusable over WAN
10:21 < trae32566[w]> IMO RDP is a significantly better solution if available
10:21 < trae32566[w]> esp. with RemoteFX
10:26 < madsj> I get network unreachable with "ssh -6 hostname", but "netstat -tulpn" has "tcp6       0      0 :::22                   :::*                    LISTEN      -"
10:27 < MrElendig> that doesn't mean that your dns or routing is correct
10:27 < MrElendig> also, are you trying to loopback?
10:27 < MrElendig> also, firewall
10:28 < MrElendig> network unreacable means that you have no route to whatever hostname is pointing at
10:28 < MrElendig> would be a different error if the port was closed or nothing was listening on it
10:31 < madsj> I uncommented the ListenAddress in the config. I also created A and AAAA records.
10:31 < madsj> MrElendig: 22 (v6)                    ALLOW       Anywhere (v6)
10:31 < MrElendig> traceroute
10:33 < madsj> traceroute6 says "network unreachable"
10:39 < MrElendig> as I said, your network or dns is the issue
10:39 < MrElendig> traceroute the ip
10:40 < MrElendig> also, does it work with ipv4?
10:41 < Raed> madsj: If you dig the hostname does it return the correct address?
10:42 < madsj> Raed: dig only responds with the A record.
10:42 < Raed> madsj: dig AAAA hostname
10:43 < madsj> The gateway in the inet6 section is incorrect. That might be a source of the error.
10:43 < Raed> madsj: Well, that will make a difference too lol
10:44 < dilema> Hi! does anybody tried to setup elk stack to work with exim logs? How easy/difficult it was to setup? Anybody knows a better alternative?
10:45 < silent_ack> hey
10:45 < silent_ack> Really very happy to configure the new desktop envirenment :))
10:48 < ren0v0> Hey, i'm trying to connect to a NAS device on a network outside of the one i'm currently on, just wondering if someone can point me in the right direction?
10:48 < ren0v0> I've setup an SSH tunnel to the other network, that works fine
10:48 < ren0v0> But the device i'm trying to connect to there is on a different IP, how do i mount an NFS there?
10:48 < Raed> ren0v0: What exactly is the problem?
10:49 < Dagmar> These two statements do not compute
10:49 < Raed> lol
10:49 < ren0v0> Hi Raed , I'm not sure how to format the mount command
10:49 < Dagmar> You need a _route_ for IP traffic to get to/from the remote NFS server
10:49 < tdn> trae32566[w], that is what I hope it will be. I have installed xrdp on the server and rdesktop on the client. But when I connect and log in, I just get a blank teal screen.
10:50 < MrElendig> ren0v0: you set up a route
10:50 < ren0v0> So, the tunnel i created goes to a certain IP on the other network, so i can't just use "sudo mount - nfs -o port=<port> localhost:/path"
10:50 < Raed> ren0v0: mount 1.2.3.4:/directory mount/point
10:50 < Raed> oh you want to mount through the tunnel
10:50 < Dagmar> You could, except NFS uses UDP and a number of different ports
10:50 < Dagmar> Just run a VPN link over SSH
10:50 < ren0v0> Raed: yea i guess :D
10:50 < Raed> so mount 127.0.0.1:1234:/directory /mount/point
10:51 < Raed> or whatever the port you are using for the tunnel is.
10:51 < ren0v0> Raed: the tunnel goes to a different device on the network, not the NAS
10:51 < ren0v0> so, i can connect to the NAS from browser, using this tunnel
10:51 < ren0v0> but not sure how to mount it as NFS
10:52 < Raed> regardless, you point the mount command at the tunnel instead of the address
10:52 < Raed> and the tunnel will push it to the NAS server
10:52 < madsj> Raed: dig still only responds with an A record.
10:53 < tdn> ren0v0, not having read the entire backlog: Are you attempting to mount NFS share over SSH tunnel? Any reason not to just use sshfs?
10:53 < Raed> and you need mount -o port=#### 1.2.3.4:/directory /mount/point
10:53 < Raed> madsj: Do you mind sharing the hostname?
10:53 < madsj> worldinfilms.dk
10:53 < ren0v0> Raed: the tunnel is a different IP completely, if i mount it like you're suggesting it'll try to mount the server not the NAS ?
10:53 < madsj> DNS is configured through DigitalOcean.
10:54 < ren0v0> tdn: I'll check out SSHFS now, maybe that's easier, but again does that matter if the NAS itself isn't open? There is a reason i only have this one server open that i can proxy through
10:54 < Raed> ren0v0: If you tell mount to connect to the tunnel, the tunnel is going to forward whatever you send to it
10:54 < ren0v0> Raed: it'll be mounting a completely different drive
10:54 < tdn> madsj, is wagtail good?
10:54 < ren0v0> So that isn't useful ?
10:54 < hans_> is there something like cat but prints the filename before starting on a new file?
10:54 < ren0v0> I need to mount a specific IP on the other networkk
10:54 < ren0v0> that isn't the same as the tunnel
10:55 < tdn> ren0v0, not sure what it means that NAS isnt open. Can you not SSH to the NAS?
10:55 < bartmon> ren0v0, ssh tunneling https://gist.github.com/proudlygeek/5721498
10:55 < ren0v0> tdn: nope, not from externally
10:55 < Raed> madsj: Try dig @1.1.1.1 AAAA worldinfilms.dk
10:55 < ren0v0> i don't wish to open that
10:55 < ren0v0> I only have a single entry point to the other network
10:55 < tdn> ren0v0, frmo eternally what does that mean? You are blocked by NAT?
10:55 < Raed> madsj: Because I am getting a AAAA record, so maybe the DNS server you are using isnt supplying ipv6 records for some reason
10:55 < ngomes> hi there! i'm reviewing this hack, where you can merge a image file to other file ( cat file.jpg compressed.zip >> new.jpg ). the binary "file" only outputs the discovered jpg file and not the zip file. is there a way to list all files with binary "file" or other ?
10:55 < ren0v0> tdn: i only have a single IP and port open from outside of the network
10:55 < hans_> huh, apparently you can trick tail into doing it, tail -n +1 file1.txt file2.txt file3.txt
10:56 < hans_> (`+1` is apparently a magic code for `print the entire file` )
10:56 < tdn> ren0v0, if so, you can either use 1) port forwarding in NAT, 2) ssh reverse tunnels to a server outside of NAT or 3) Set up SSH over a TOR Hidden Service / Onion Service.
10:56 < ren0v0> excuse some of my terminology as i'm not a network expert! clearly.
10:56 < hans_> ren0v0, i'll never forgive you
10:56 < MrElendig> ngomes: sure, parse the entire thing instead of just looking at the first few bytes like `file` does
10:57 < Raed> ren0v0: What command are you using to tunnel to the remote host?
10:57 < tdn> ren0v0, if you are not going to move huge amounts of data and latency/bw is not important for you, I would definately go for #3.
10:57 < ren0v0> tdn: So i have an SSH tunnel to the network, this part is fine, and i can access all devices through browser etc etc, but I'm looking to mount a specific share now
10:57 < bartmon> ngomes, file only looks at the magic bytes at the beginning of a file. if you want to have a reversible way of packing multiple files into one, use tar
10:57 < adac> "sudo smartctl -H /dev/sda" does this always show the current status? Or do I have to run something before this?
10:57 < adac> *current disk status
10:57 < Raed> ren0v0: So mount the share through the tunnel, like I showed you.
10:57 < tdn> ren0v0, you will probably have a bad time mounting NFS through SSH tunnels.
10:57 < Raed> ^^
10:57 < ngomes> bartmon, not sure how i will do that
10:57 < tdn> ren0v0, use #3 above and sshfs instead
10:57 < Raed> It isnt going to work very well.
10:57 < MrElendig> nfs over the internet is a bad thing period
10:58 < tdn> MrElendig, agree
10:58 < ren0v0> Raed: ii'm using this >    ssh -D <port> -f -C -N <domain> -p <port>
10:58 < ren0v0> It's only for a few text files and stuff
10:58 < Raed> Just scp the files then.
10:58 < ren0v0> meh  :D
10:58 < tdn> ren0v0, follow this howto: https://nurdletech.com/linux-notes/ssh/hidden-service.html  then after that, just mount via sshfs and you are good
10:59 < Raed> You are making way more complicated than it needs to be lol
10:59 < ren0v0> would rather use nautilus!
10:59 < dnanib> ren0v0: Why are you doing contortions like NFS over SSH? I do sshfs, works just fine.
10:59 < Raed> Then mount the ssh server in nautilus
10:59 < MrElendig> or use something like owncloud or whatever
10:59 < Raed> You dont even need sshfs for this.
10:59 < tdn> ren0v0, with sshfs you can use whatever file manager you want. It will look like a proper folder.
10:59 < ren0v0> dnanib: i'll look into this, but doesn't that mean i need direct SSH to the device in question ?
10:59 < Raed> Just use connect to server in nautilus, then point it at the tunnel port
11:00 < ren0v0> Raed: the tunnel:port isn't the device i'm trying to reach, this is why the mount command you gave doesn't make sense to me
11:00 < ngomes> bartmon, any example ? I don't see any option in file
11:00 < bartmon> ngomes, if you control this process if joining files, refactor to use tar instead. if you are gien such a file, MrElendig's advice stands: you need to have knowledge of file formats to know where a format's data ends and another format begins. a much more complicated process than just dealing with the tar archive format.
11:00 < ren0v0> If i point it to the tunnel, it'll mount that servers folders, which isn't correct
11:00 < Raed> You are ssh'd into the box you want to mount nfs over correct, or no?
11:00 < ren0v0> Raed: no
11:01 < Raed> Ah ok
11:01 < ren0v0> but that box has access
11:01 < ren0v0> to the NAS
11:01 < MrElendig> ngomes: what are you actually trying to do?
11:01 < ren0v0> ok well i'll just have to SCP, just thought there must bee a solution :D
11:02 < ren0v0> "graphical solution"
11:02 < MrElendig> sidenote: exif has a field you can dump arbitary data into
11:02 < MrElendig> ren0v0: your fancy gui file manager will do sshfs
11:02 < Raed> So you can just make the tunnel a bit differently. ssh -R ####:whateverhostthenfsison:#### whateverhostthesshserverison
11:02 < Raed> Then mount the first port you use
11:02 < dnanib> ren0v0: Joined a bit late to the party. You have a NAS device at the other end? So its volumes must be mounted on something there?
11:02 < MrElendig> ren0v0: so will things like filezilla
11:03 < madsj> tdn: I like it. It's quite easy to configure different components and snippets (structured information you can use, e.g., people, films etc.)
11:03 < Kingsy> question, if I have something like this "* * * * * www-data /path/to/script.sh" in my crontab, why wouyld the output of the cron be www-data not found, its like its not treating www-data as a username, which it should when its placed there correct?
11:03 < ren0v0> MrElendig: dnanib I don't have direct SSH access to the device in question, only through another device
11:03 < MrElendig> ren0v0: can still do it
11:03 < ngomes> bartmon, MrElendig , as I use file, i don't need much understanding about magic bytes , in order to make it prompt for me. MrElendig i want a proof-of-concept that those hidden files can be easy unconvered , that's why I was trying "file", too bad it does not read all the file to seek other filetypes in the same file
11:03 < ren0v0> So i'm connecting to 192.168.1.10 for example, but the NAS is on 192.168.1.20
11:03 < djph> ren0v0: o...kay?
11:04 < ren0v0> MrElendig: if you have a sexy mount command for me please share!
11:04 < MrElendig> ngomes: you don't need a proof of concept for this, it have been researched to death already
11:04 < MrElendig> ngomes: a lot of the existing forensic tools will check for things like this
11:04 < ngomes> MrElendig, can you name 1 ? that's what i'm looking for
11:04 < Raed> ren0v0: https://www.ssh.com/ssh/tunneling/example
11:04 < tdn> madsj, I am looking for a good CMS. Will look into it. Thanks.
11:04 < djph> Kingsy: missing the '-u'
11:05 < tdn> madsj, what is the security track record for wagtail?
11:05 < djph> Kingsy: * * * * * -u www-data /path/to/script
11:05 < MrElendig> ngomes: see the defcon talk about auditing fails
11:05 < MrElendig> talks even
11:05 < ren0v0> Raed: I don't want to do it system wide, am i now being too difficult ?
11:06 < madsj> tdn: it's based on django, which gets updates. I'd say it's reasonable safe.
11:06 < madsj> Raed: AAAA 	
11:06 < madsj> worldinfilms.dk Copy
11:06 < madsj> 	directs to
11:06 < tdn> madsj, ok
11:06 < ren0v0> I want to use this local network also, they have similar ports etc that would clash
11:06 < MrElendig> also the recent stenography talk
11:06 < Raed> madsj: Its returning 2a03:b0c0:3:d0::b3d:8001 for me
11:06 < dnanib> ren0v0: NAS => the box exports filesystems that you need to mount using some network protocol (NFS or SMB, usually) on hosts that need the storage.
11:06 < Raed> ren0v0: What do you mean system wide?
11:06 < heptagon> Is there a way to prevent user programms from accessing the ouput input devices (after X11 remaping), or even better masking that?
11:07 < dnanib> So 1.20 is your NAS which exports filesystems, other nodes in 1.x can NFS mount those exports
11:07 < ren0v0> dnanib: yes, and those are exported to devices on that network, including the one i am tunnelled to
11:07 < Raed> ren0v0: You can choose whatever port for the tunnel
11:07 < dnanib> So you need to login to 1.10, mount the NAS's export there, and then access that mountpoint via an Internet filesystem like sshfs
11:07 < ren0v0> dnanib: yes, i could do that also i guess
11:08 < Raed> dnanib: He is using an SSH box as the middle man to the NFS server
11:08 < ren0v0> i understand that method
11:08 < ren0v0> yes, ideally i would like to just hop straight to the NAS
11:08 < Raed> So he needs to tunnel
11:08 < dnanib> Which is a roundabout way, but if the NAS box doesn't provide any other access perhaps that is your best option. Check the NAS's config to see if it doesn't allow SCP/SFTP/FTPS etc...
11:08 < ren0v0> mounting on this server is also an option though
11:08 < Raed> ren0v0: Did you look at the page I sent you? It explains remote forwarding
11:09 < ren0v0> Raed: i did, but i thought i was already doing that with the command i pasted a minute ago, as i say i can already access the NAS through browser
11:09 < ren0v0> just setup proxy in chrome and it works
11:09 < ren0v0> not sure how to use this proxy for mount command, or nautilus etc
11:09 < Raed> I told you. Lol
11:09 < ren0v0> ha, i know i'm just clearly missing it
11:10 < Raed> mount -o port=#### )whatever the proxy port is) nfshostaddress:/directory /mount/point
11:10 < ren0v0> ok lets see
11:10 < Raed> Wait a minute lol
11:10 < Kingsy> djph: ahhhh! damn thanks
11:10 < Raed> dont use nfs host address, use 127.0.0.1
11:11 < Raed> you are doing it in a strange way in the first place
11:11 < ren0v0> if i do that, there is nothing there to tell mount where the NAS is...
11:11 < Raed> Let me draw this out lol
11:11 < ren0v0> :D
11:11 < Raed> Because now I am confusing myself
11:12 < ren0v0> dnanib: just to clarify, the NAS does allow various connection methods, but out of choice i've closed the whole network apart from this single entry
11:12 < Raed> ren0v0: So the NAS only allowys SSH in right?
11:13 < ren0v0> i dont trust all of these cloud apps and other things, and would rather have only one place to maintain :D
11:13 < Raed> allows*
11:13 < ren0v0> Raed: it has SMB and NFS protocols open only
11:13 < dnanib> What are the other protocol options?
11:13 < Raed> So tell me again why you cant just directly connect to it, and you need to go through the tunnel
11:14 < Raed> From the machine you are trying to mount the NAS on
11:14 < ren0v0> Because its in a different part of the country :D
11:14 < Raed> If the ports are open it wouldnt matter where it is..
11:14 < ren0v0> it doesn't have an external IP, all ports are closed
11:14 < ren0v0> i dont want it open
11:14 < Raed> So the other SSH host is in the network that the NAS is in?
11:15 < ren0v0> yes
11:15 < Raed> Oooohhhhhhhhhh
11:15 < Raed> Got it
11:15 < ren0v0> !
11:15 < Raed> So
11:15 < madsj> Raed: that dig AAAA command also returns the correct ip6 address. I could try to disable the firewall and poke around.
11:16 < Raed> So you want ssh -R #####:nasaddress:nasport sshhost
11:16 < Raed> and the first set of numbers you pick for a port are going to be your local proxy port
11:16 < Raed> Then you just mount -o port=#### 127.0.0.1:/NFSdir /mount/point
11:17 < ren0v0> ok, and the "sshhost" there, should that be pointed to the tunnel i've already setup? so its two tunnels ?
11:17 < Raed> madsj: Try ssh -6 -vvv host
11:17 < Raed> madsj: And see what it tells you.
11:18 < Raed> ren0v0: The host is that server you are sshed into to get to the NAS
11:18 < Yamakaja> Is there any way that the ping binary could be abused to execute commands or modify files?
11:18 < Raed> You dont need 2 tunnels, just the command I sent you.
11:18 < ren0v0> Ok let me try
11:18 < Yamakaja> I'm wondering how hard i need to filter input to ping / which options i can allow access to
11:18 < Raed> ren0v0: You could keep the second tunnel open if you wanted to maintain your web access, but it isnt necessary
11:19 < ren0v0> Raed: there is no "nasport", so does it matter when i set there?
11:19 < Raed> ren0v0: Is it just an NFS share?
11:20 < ren0v0> yea
11:20 < Raed> Whatever port NFS listens on
11:20 < ren0v0> well no there are other things running on this, its a netgear product
11:20 < ren0v0> ok will check that
11:21 < Raed> It looks like you might need a second option, because NFS needs 2 ports
11:22 < madsj> Raed: https://dpaste.de/hMvu
11:23 < ren0v0> Raed: that just connected me to the SSHhost
11:23 < Raed> madsj: Does the box you are trying to connect from, have a valid ipv6 address and is it actually able to access the internet over ipv6?
11:23 < ren0v0> the port i used for local, no process is open on that looking at netstat
11:23 < Raed> ren0v0: After you run that command then you pass the mount command
11:24 < ren0v0> but "netstat -tupln" shows 6666 (the port i used) as nothing running there
11:24 < ren0v0> unlike my other tunnel which shows SSH open on it
11:24 < djph> sounds more like you're wanting a VPN ...
11:25 < ren0v0> djph: not really because i dont want to run everything through it on the system, just a couple of things
11:25 < ren0v0> i looked  at "proxychains", which works for firefox and other things, curl and that, but not for nautilus
11:25 < Raed> madsj: I think you dont have ipv6 connectivity because I was able to connect over v6 to it
11:25 < znh> what about sshfs
11:25 < djph> ren0v0: you don't have to run "everything" through the VPN
11:25 < ren0v0> lol, everyone keeps saying sshfs but miss the point that i'm needing an extra hop
11:26 < ren0v0> the FS i want to connect to doesn't have SSH open
11:26 < Raed> ren0v0: No, if you run sshfs on the NFS box instead of nfs then you canjust tunnel right to it
11:26 < ren0v0> yes i get that, just not what i wanted to do
11:26 < Raed> ren0v0: Without having to do all of the mounting through the tunnel
11:27 < ren0v0> if i get it working with tunnel it would be preferable and i'll just setup a script to do it when i need access
11:28 < Raed> ren0v0: Ah here you go
11:28 < Raed> ren0v0: http://nfs.sourceforge.net/nfs-howto/ar01s06.html
11:28 < Raed> Scroll to the bottom
11:29 < MrElendig> openvpn
11:29 < Raed> MrElendig: Too easy.
11:32 < djph> Raed: the funny thing is, in some cases "easy" is also "more stable"
11:32 < Raed> djph: I know I was being facetious
11:32 < djph> ^_^
11:34 < ren0v0> well this is giving me a headache, maybe i'll just use the webUI :D
11:34 < Raed> ren0v0: Did you look at the article I sent?
11:34 < ren0v0> Raed: yes but it's very confusing :D
11:35 < Raed> ren0v0: I thought it was straight forward.
11:36 < ren0v0> ssh root@<sshhost> -L 250:<nashost>:2049  -f sleep 60m
11:36 < ren0v0> So i would do this, then mount localhost:250  ?
11:37 < Raed> actually even better you can do it all in one line
11:38 < Raed> ssh -L 2500:nashost:2049 -L 2501:nashost:111 user@sshbox
11:38 < Raed> Then mount, as it is shown in the article.
11:39 < ren0v0> works
11:39 < ren0v0> :D
11:39 < afidegnum> hi, good morning, i woke up this morning to find out Apache2 is eating my CPU, but i don't remember installing it and i m using nginx by default
11:39 < Raed> (Assuming that 2049 and 111 are the NFS ports for your server)
11:39 < afidegnum> how do i find out what's apache2 is runing?
11:39 < afidegnum> Raed: me?
11:40 < ren0v0> Wow, thanks for sticking with me Raed  :D
11:40 < Raed> ren0v0: No worries, glad we got it figured out for you.
11:40 < ren0v0> mounted and showing in nautilus, perfect
11:40 < Raed> afidegnum: ps aux |grep apache2
11:41 < afidegnum> thansk
11:42 < TabMasher> I'm making a BTRFS partition to install Ubuntu. (already asked in #ubuntu) But, how can I make/set all the files to be compressed?... Either during the install, or after the install, I'd like as much of the disk to be compressed?
11:44 < TabMasher> I just need a filesystem that supports compression mainly, and I don't know which other linux supported filesystems supports file compression.
11:45 < djph> TabMasher: why do you think you need disk-level compression of files?
11:45 < TabMasher> djph: Large files with lots of redundant data.
11:46 < djph> are they text?
11:48 < djph> I mean, most media containers (mpeg/h264/mkv/jpg/tif/etc.) are already compressed formats, so you're not going to gain anything there.
11:49 < TabMasher> djph: Yes, some are large log files, for example.  Others are varied data, but compress well.
11:52 < djph> TBH, unless it's "all today's data", I'd look into something like logrotate to compress them as-needed -- especially if they're getting shipped off somewhere for medium-term storage
11:58 < TabMasher> djph: Yeah, I need a solution rather than an alternative work around.  I'm using a laptop and free space on the disk is important.   Compression would work well.  The only alternative might be NTFS of all things since it's pretty solid and supports compression.
11:59 < hans_> TabMasher, when the installation is complete, when it says "do you want to reboot or blabla", go back to the main menu, then go to the terminal, then nano /target/etc/fstab , add compress=<algo you want>  to the / mount options, then run btrfs filesystem defragment -r -v -c <algo> /target/
12:00 < TabMasher> hans_: Thank you.   Also, does any knoe of any good online literature for btrfs and zfs?
12:01 < hans_> dunno, i like https://wiki.archlinux.org/index.php/Btrfs and https://btrfs.wiki.kernel.org
12:02 < revel> Hmm. So, I've been wondering. What does it mean when `file` reports an executable as being dynamically linked, but ldd{,tree} report it as having no interpreter and being statically linked?
12:02 < hans_> TabMasher, 1 last thing, i don't remember the details, but iirc, there may be some issues if you compress the drive, then tell the main menu to reboot, but no issue if you just type "reboot" directly in the terminal.. honestly don't remember what it was tho
12:03 < revel> It can run just fine on both musl and glibc machines which don't have the required libs, so it seems like ldd{,tree} aren't lying to me, at least.
12:04 < hans_> (the installer's main menu, that is)
12:04 < TabMasher> hans_: I tend to put my /boot and a couple other folders in their own EXT3-4 directory so I don't have to deal with them.
12:04 < hans_> amen, but that's not the the kind of issue i'm talking about. personally i put /boot in an ext2 partition
12:06 < TabMasher> hans_: I see.  I wish I knew where on the web to find info about the issue.  Thank you for the online references for btrfs.
12:07 < hans_> it was issues i've experienced personally when doing it, but i didn't write it down, nor did i search online for references to it, just saying.
12:08 < hans_> (there may not exist any, idk)
12:08 < hans_> note that last time was with the Ubuntu 16.04 installer, not the 18.04 installer tho
12:24 < jim> hans_, please spell out idk as I don't know, we think this will help newer english speakers to understand more of what's going on
12:27 < TabMasher> All new English speakers: IDK = "I don't know", and FYI stands "For your information"  :D
12:27 < TabMasher> hans_: Thank you again.
12:29 < stealthii> I'm finding guidance hard on LVM snapshots, and how to take a snapshot of it's existing state *before* restoring from an older snapshot.
12:29 < jim> TabMasher, you'd have to do that constantly as new people join and others leave
12:30 < stealthii> which file do I snapshot? the old lvm file, or the snapshot that I believe is holding the new data?
12:31 < jim> stealthii, I've never snapshotted before :/
12:31 < jim> but, I have seen youtube videos are out there on lvm, maybe you'll find something good
12:34 < runjutsu> Can I execute a suit of software in an enviroment where only a few libs are changed without making a chroot for it?
12:35 < revel> Set LD_LIBRARY_PATH, maybe?
12:40 < hans_> jim, oh cmon, we're at the internet, and `idk` has even long since entered many english dictionaries
12:40 < Kingsy> djph: hm, so I have this line --> * * * * * -u www-data echo "Hello world" >> /var/log/cron.log 2>&1 <-- and it does not run at all, if I remove the -u it says www-data not found
12:40 < runjutsu> revel: I see the problem now, yeah. I need to keep the old one too.  LD..PATH=old:new
12:42 < hans_> jim, https://en.oxforddictionaries.com/definition/idk , https://urbandictionary.com/define.php?term=idk , https://en.wiktionary.org/wiki/IDK
12:45 < supay> in bash, how would you do something like: for i in `docker exec -it t4re ls /home/work/$TASK_TMP_DIR/`;  do echo ${i}; done?
12:45 < supay> i keep getting syntax error: try.sh: line 2: syntax error near unexpected token `$'\r''
12:47 < KekSi> supay: at the very first i would suggest you're using docker wrong
12:47 < supay> KekSi: haha, i feel so too
12:48 < somnus_> is it okay to install package for ubuntu xenial on bionic? i installed 18.04 but k8s recomends install docker 17.03; but docker does not provide package for bionic
12:49 < Raed> somnus_: If it works, then whatever, you could try it, however there is probably a pretty good reason they are suggesting another version.
12:49 < djph> Kingsy: does www-data have write permissions to /var/log/cron.log ?
12:50 < Raed> somnus_: So you might run into other issues
12:51 < somnus_> Read: thanks for advice. i will take risks
12:52 < madsj> Raed: traceroute6 google.com also responds network unreachable, so perhaps my ISP is the issue.
12:53 < runjutsu> in what order are the directories in LD_LIBRARLY_PATH checked?  a:b:c .. is it "a" first?
12:53 < Raed> madsj: Does the system even have an ipv6 address?
12:53 < Raed> madsj: Like, can you even access the itnernet over ipv6? Does your ISP even support it?
12:53 < Raed> internet*
12:54 < madsj> ifconfig just gives me "fe80:....". So I'm not sure.
12:55 < Raed> madsj: That is a link-local address, so you probably don't have ipv6.
12:57 < sssitrar> hi
12:58 < sssitrar> I got an alsa audio device connected to my linux systen
12:58 < sssitrar> *system
12:58 < djph> as do many
12:58 < MrElendig> lies, we all use pc speakers
12:58 < MrElendig> and tweeters
12:59 < sssitrar> And it is recognised by alsa drivers , but there is an issue in recording with alsa utility
12:59 < Raed> MrElendig: Nope, I've got a phonograph attached to my audio port.
12:59 < MrElendig> fuser /dev/snd/*
13:00 < sssitrar> It says "arecord: set_params:1145: Unable to install hw_params"
13:01 < sssitrar> What is going wrong ?
13:38 < madsj> Raed: I figure ipv6 is going to get adopted more, so I wanted to toy around more with it. Anyway, seems not to be worth the effort currently. Thanks for the assistance.
13:39 < Raed> madsj: If you wanted to toy with it, look into some of the ipv6 tunneling services out there, there are lots of free ones and it uses a tunnel to get ipv6 to you.
13:40 < Raed> madsj: I have personally used hurricane electric's service in the past, and it worked well for me.
13:50 < madsj> Raed: Yes, the need for a tunnel was also the conclusion I came to. Hurricane Electric is the first on Wikipedia's list. Thanks again.
13:56 < jken> Hey, I have an odd issue I am having trouble debugging. I run xorg and openbox, using openbox's autostart script to kick off my app on boot. On kernel 4.9 this all works fine, on kernel 4.16 (stretch-backports) there is a ~45second delay between openbox starting and my app starting.. switching the kernel back corrects this. Might anyone have some insight here?
13:58 < Raed> madsj: No worries! Good luck!
14:00 < mub> Hey y'all
14:02 < mub> So, my boss wants us to migrate away from firewalld and back to iptables management.. I've already written at least a thousand lines of puppet manifests to configure our apps and set up firewalld rich rules
14:03 < mub> Is there any real benefit to using iptables over firewalld?
14:03 < ananke> no
14:04 < Raed> mub: But telling your boss that someone on IRC said it wasn't any better or worse isn't going to change their mind lol
14:05 < mub> No.. There is nothing I can do
14:05 < ananke> mub: I'd be interested to see why your boss thinks it's a good idea to repeat the effort
14:05 < mub> we keep a file at /etc/sysconfig/iptables that is well commented
14:05 < mub> He doesn't like how firewall-cmd has no --comment argument
14:05 < Raed> Wow lol
14:05 < Raed> Well then
14:05 < Dagmar> Firewalld is fancy, therefore expensive.  THey want greater value, so firewalld
14:06 < ananke> mub: sounds a bit silly
14:07 < mub> isn't iptables being deprecated in favor of nftables?
14:08 < mub> eventually
14:08 < revel> No, it's being deprecated in favour of BPF.
14:09 < mub> This is now my ammo
14:10 < Dagmar> It barely matters whether it's eventually being deprecated or not
14:30 < biberkopf> not to be meant as flame, but would like an honest opinion. Is Slackware and to what extent a 'pointless exercise' in 2018 and what advantages would one have using it as a daily driver/tool in 2018 compared to other distros? thanks!
14:33 < searedvandal> what makes you say Slackware is a 'pointless exercise
14:33 < searedvandal> ?
14:34 < notmike> Hate
14:34 < notmike> It's infected her heart
14:34 < notmike> Jealousy
14:34 < notmike> Envy because they can't compile slackware and so will never know that Joy
14:35 < oiaohm> biberkopf: I still know of a few embedded systems based around slackware.
14:35 < oiaohm> biberkopf: packaging with slackware ends up compact.
14:35 < searedvandal> probably a reason why it's the oldest linux distro still maintained
14:36 < searedvandal> I think it had first release before debian?
14:36 < biberkopf> as i said, the questions is *honestly* not meant as flame.
14:37 < searedvandal> biberkopf, why only include slackware? why not include all distros aimed at more advanced users?
14:37 < biberkopf> Pointless exercise is meant in quotes, as i read online that you basically have to compile everything, and that you basically get what you want with much more time used
14:37 < searedvandal> so, like gentoo?
14:38 < notmike> Except better
14:38 < notmike> Like the best sex ever
14:38 < biberkopf> i included slackware as i'm looking for a daily driver, and it's the oldest distro, + the main point everyone sold me on was stability and long loooong support
14:38 < biberkopf> + i use xfce so for the naked eye, no change ever happens
14:38 < notmike> Alabama black snake
15:05 < halftroll> Hello
15:05 < halftroll> when I enter into my vps I think something is being executed each time I login with ssh
15:06 < lupine> all sorts of things will be
15:06 < lupine> check ~/.bashrc, ~/.bash_profile, etc
15:09 < halftroll> ok thanks
15:09 < revel> And the global profile+bashrc in /etc/
15:21 < kcn> Hi, I'm trying to pass a large data (~MB) between kernel and user space. I'm considering to use libnl but the max msg size by default is only 8K
15:22 < kcn> so when I pass the data as an attribute, the message becomes too large and couldn't go through
15:23 < kcn> so, I'm thinking about dividing the data into several chunks and pass them as several nested attributes.
15:23 < kcn> would this make a difference?
15:29 < lightslategrey> Hi! Can anyone suggest me about udev and NetworkManager?
15:30 < Psi-Jack> That's rather vague.
15:32 < lightslategrey> Psi-Jack, I want prevent NetworkManager from seeing device before udev finish executing some program
15:33 < Psi-Jack> Eh? What actual problem are you trying to solve?
15:35 < lightslategrey> Psi-Jack, It's long to explain. Apparently thereis bug in NM that makes it unusable when multiple NIC's have same MAC address. I am changing MAC for them from udev. But network manager does observes them with old MACs and then it does not works.
15:36 < Psi-Jack> Eh? No, that's not a bug, that's a complication that's unusual.
15:38 < Psi-Jack> Sounds like it's high time to get a new NIC that doesn't conflict.
15:38 < lightslategrey> Psi-Jack, NM does takes into account subsequently that MACs are changed then (and shows correct ones in future in mncl), but it behaves strangely and have weak proof that it caused by NM seen old MACs once and that the root case.
15:38 < triceratux> lightslategrey: did you try spoofing the macaddr with ip link set rather than udev ? or cycling NM yourself after youre sure your spoof has taken effect ?
15:38 < phinxy> oh no, my $1 USD linux stickers got stuck in customs.  $8 customs fee.  :,<
15:40 < phinxy> werent Unixstickers.com shipping from within EU?
15:41 < Psi-Jack> Better question to ask of them. :)
15:41 < Industrial> Hi. Say I have a .env file. How do iset all it's vars in the current shell session?
15:41 < Psi-Jack> soiurce file
15:41 < Psi-Jack> source*
15:41 < Industrial> oh, k
15:41 < Industrial> ty
15:41 < azarus> or
15:41 < azarus> . file
15:42 < azarus> (it's POSIX)
15:42 < Psi-Jack> He left already.
15:42 < lightslategrey> triceratux, I have to use udev RUN with python script because I need to come up with new pseudo-random MAC, from script I change mac calling ioctl SIOCSIFHWADDR. I don't think other mac chage tools do any other syscall.
15:42 < Psi-Jack> Rotating door. :/
15:42 < azarus> damn
15:43 < triceratux> thats why ya gotta answer those questions *fast*
15:43 < Psi-Jack> lightslategrey: ACTION=="add", SUBSYSTEM=="net", ATTR{address}=="XX:XX:XX:XX:XX:XX", RUN+="/usr/bin/ip link set dev $name address YY:YY:YY:YY:YY:YY"
15:43 < lightslategrey> Psi-Jack, there is no way to change MAC on these devices (CDC USB Ether devices), I read driver code and see there is no way to detect and configure MAC duplicates it in driver.
15:44 < Psi-Jack> My suggestion still stands. get a new NIC. You WILL have constant complications with that as-is, even if you do "solve" this front-facing issue.
15:45 < leftyfb> lightslategrey: you have multiple USB ethernet adapters and they all have the same MAC address?
15:46 < lightslategrey> Psi-Jack, what I am actually doing is just more advanced version of this. I will not gonna work, because NetworkMangler discovers device Before any actions done by udev. It is show in logs in debug verbosity of both udev and NM. NM will then "rediscover" them with new IFnames and MAC it Seems that damage from first exposure us already done.
15:46 < Psi-Jack> So, done. Answer: New NIC.
15:47 < lightslategrey> Psi-Jack, you are right that the normal solution is just fix the devices, you are right.
15:47 < br1s> I read from sched-design-CFS.txt that Linux does not have "timeslices" in the scheduler. However this OpenSUSE document advice on how to tune sched_latency_ns and sched_min_granularity_ns regarding timeslices. https://doc.opensuse.org/documentation/leap/tuning/html/book.sle.tuning/cha.tuning.taskscheduler.html Should I just rely on the kernel's .txt and believe that tuning only /proc/sys/kernel/sched_min_granularity_ns is sufficient?
15:47 < winsoff> Are there any DEs out there that try to follow "flat design" or other modern design trends?
15:47 < leftyfb> lightslategrey: If you have multiple NIC's with the same MAC address from the manufacturer, the problem is 100% the cheap hardware. Invest in hardware that was designed to proper specifications
15:48 < leftyfb> winsoff: try #ubuntu-offtopic
15:48 < Psi-Jack> winsoff: 1980's calling. They want their Windows 3.11 back.
15:48 < leftyfb> br1s: try ##linux
15:48 < winsoff> Psi-Jack, have you even used a DE since 1980? ;P
15:48 < oiaohm> winsoff: define what you call modern.     With the numbers of windows managers that have been done there is very few new ideas.
15:48 < br1s> leftyfb: isn't this ##Linux?
15:48 < winsoff> leftyfb, do you know where you are? LOL
15:49 < Psi-Jack> winsoff: Actually... Yes. :)
15:49 < triceratux> winsoff: flat design is just a theme to many DEs these days. dozens of distros come up flat themed these days in multiple DEs
15:49 < winsoff> triceratux, interesting. What's your favorite style?
15:49 < leftyfb> br1s: another channel dedicated to general linux support. Your question has very little to do with Ubuntu. In fact, you mentioned OpenSUSE more than Ubuntu.
15:49 < leftyfb> br1s: oops, my bad
15:49 < oiaohm> The first flat theme I saw was for mwm
15:49 < leftyfb> wrong channel lol
15:49 < Psi-Jack> leftyfb: This is ##linux. :p
15:49 < br1s> leftyfb: this has nothing to do with OpenSUSE
15:50 < leftyfb> Psi-Jack: yup, just noticed I thought I was in another channel
15:50 < lightslategrey> leftyfb, you right and I also looking for patching device firmware. But the actual cause is also NM. It should not go crasy from that face of once exposure of unchanged MACs. It should wait for some time or at least have an option. The reason as I understand that subscribes to udev hardware events indepently don't cooperate with udev.
15:51 < notmike> If you're not using slack you're wrong
15:51 < winsoff> Did I already ask earlier this week which distros people prefer to put on old hardware?
15:51 < br1s> winsoff I prefer Xubuntu LTS
15:51 < notmike> winsoff: see above
15:51 < winsoff> I have a friend who really likes alpine, but I think he uses that for server-based stuff, not "I need a word processor and firefox"
15:51 < Psi-Jack> One does not /run/ slack. They simply use it.
15:51 < oiaohm> winsoff: really most applications are not that ram friendly.
15:51 < notmike> Psi-Jack: why you hate fun, fam?
15:51 < winsoff> br1s, any advantages over lxle?
15:52 < Psi-Jack> leftyfb: You probably are. just had the wrong one active. :)
15:52 < winsoff> oiaohm, it's true, which is why I'd rather get the OS out of the way as much as possible RAM-wise
15:52 < oiaohm> winsoff:   its the biggest issue with old hardware is not being able to put enough ram in it.
15:52 < notmike> winsoff: why use a DE at all?
15:52 < leftyfb> Psi-Jack: correct
15:52 < br1s> winsoff most of the time when I use old HW I just use it for some simple general purpose task, and almost everything works out-of-the-box in Xubuntu :)
15:52 < winsoff> notmike, not for me; for a pipeline worker.
15:52 < Psi-Jack> When you suggested the ubuntu offtopic channel, I blinked.
15:52 < triceratux> winsoff: in xfce for example it just requires a flat XFWM4 theme & replacing the xfce4-panel with one of the flat docks https://swagarch.gitlab.io/ i still use something with a little more shading & gradients in it. gtk2 stuff which theres a lot of still in xfce
15:53 < oiaohm> winsoff: anything you can get 4g of ram in is fine with xfce for most tasks without running into swap hell.
15:53 < winsoff> Hahaha. "Swagarch." Interesting name.
15:53 < winsoff> oiaohm, makes sense.
15:54 < oiaohm> winsoff: like you can make linux work on 1G and less but you do end up suffering.
15:54 < notmike> winsoff: you getting some pipe out the deal?
15:54 < Psi-Jack> Heh, I can't recommend SwagArch at all. the sole developer on it, is obviously quite inexperienced. :)
15:54 < winsoff> notmike, nay; just common payment.
15:54 < notmike> TempleOS is obvious choice
15:55 < oiaohm> winsoff: I do have a Raspberry Pi 3 with 2GB  and I really do hate the ram issue if you attempt to run libreoffice or the like.
15:55 < oiaohm> winsoff: so its not exaclty old hardware where the ram issue is a pain in the tail either.
15:56 < triceratux> xfce is the only hope for older hardware http://www.freerepublic.com/focus/f-chat/3666090/posts but Lubuntu 18.10 lxqt has been solid for so long i can hardly keep from using it http://cdimage.ubuntu.com/lubuntu/daily-live/current/cosmic-desktop-amd64.iso
15:56 < winsoff> oiaohm, libreoffice on an rpi must be a nightmare, too.
15:56 < winsoff> just in general
15:57 < Psi-Jack> winsoff: Hmmm... Budgie's actually got a rather modern flat look to it.
15:57 < Psi-Jack> In Solus, which is actually a pretty impressive distro.
15:58 < winsoff> Ah yeah, it does.
15:58 < oiaohm> triceratux: lxde is slightly lighter but lacking search on menu is a serous pain in tail.
15:58 < searedvandal> lxqt
15:59 < searedvandal> liking it more each day now
15:59 < oiaohm> searedvandal: does that have a search on the menu.
15:59 < winsoff> Is Solus rolling, Psi-Jack?
15:59 < winsoff> Also, what makes an impressive distro?
15:59 < Psi-Jack> winsoff: Yes it is.
15:59 < oiaohm> searedvandal: Linux with a few thousand applications installed without a search is pain.
16:00 < searedvandal> oiaohm, few thousand applications? but yeah, there is a search in the applications menu
16:01 < Psi-Jack> winsoff: Well, it's solid, fast, new and interestingly well designed, and for some reason, with Solus specifically, I actually have more RAM for applications to run. like right now, after 12 days now, I have 2.88GB of /used/ RAM (excluding caches), and any other distro I'd have already been swapping out my 16GB RAM.
16:01 < triceratux> oiaohm: lxde itself has no future now that lubuntu is doing away with it. lxqt has a search box at the bottom of the panel menu but it seems to only go thru everything in the menu. for something more powerful youll need dmenu or something
16:01 < triceratux> Psi-Jack: have you tried the ubuntu budgie spin ?
16:02 < Psi-Jack> winsoff: They put system stuff where it really belongs, too. Keeping it out of /etc.
16:02 < Psi-Jack> triceratux: You have to ask have I used Ubuntu?
16:02 < azarus> ubahntu
16:02 < Psi-Jack> The answer is always HELL NO!.
16:02  * triceratux is just making sure rofl
16:02 < winsoff> Psi-Jack, that's really interesting. Hmm. What is budgie? is this a newfangled de?
16:02 < dbolser> hello
16:03 < Psi-Jack> winsoff: It's based off, and even still uses parts of Gnome 3, but yes, custom newfangled DE with some pretty impressive simple pleasures.
16:03 < oiaohm> triceratux: I will have to give it a try.    Also it will have to check if you can tell between flatpak and normal applications as well.   Basically my two 100 percent sure I will not use that de is 1 no menu search.  2 not being able to pick between distribution libreoffice and flatpak libreoffice.
16:03 < winsoff> Ah, interesting.
16:03 < winsoff> I'll throw solus on this laptop first, and if that breaks, then something else.
16:04 < azarus> alpine 3.8 got released today, yay
16:04 < Psi-Jack> winsoff: Also. Solus is specifically designed for Desktop/Laptop use. Not designed for servers. ;)
16:04 < winsoff> Now with that in mind, if I start with a more barebones/"PERFOARMANCE" operating system like alpine, am I going to break something/ruin everything by then slapping on a new de?
16:04 < azarus> no
16:04 < JeffATL> anyone have suggestions as to how i can more easily/readily shell into a cluster edge node from my XQuartz-running macbook and get xosview windows exported to me from each machine in the cluster (edge node is on an independent LAN with the other cluster nodes)?
16:04 < dbolser> I'm behind a firewall that's blocking a tls connection (as far as I can tell). When it (tries to) establish a connection I see the local and remote port
16:04 < azarus> it's never broken if you know how to fix it ;)
16:05 < dbolser> I'd like to proxy the local port over ssh to avoide the fw
16:05 < winsoff> azarus, what makes alpine better for servers than other environments? Sorry to target you; you just seem knowledgeable there
16:05 < dbolser> however, I can't guess the port before staring the thing
16:05 < azarus> winsoff: i don't really use it on servers that much
16:05 < winsoff> To guess which port, dbolser?
16:05 < JeffATL> right now I shell into the edge node three times and in each window shell to each cluster node and run xosview &
16:05 < winsoff> azarus, orly? Desktop only, then?
16:05 < azarus> winsoff: works on everything
16:06 < azarus> it's a general purpose distro
16:06 < winsoff> I'm tempted, but like
16:06 < Psi-Jack> winsoff: standard english please. Not "orly"
16:06 < winsoff> Psi-Jack, orly has been in the lexicon longer than "on fleek."
16:06 < Psi-Jack> winsoff: It's not accepted /here/.
16:06 < winsoff> I am just trying to make you feel old to distract you.
16:06 < Psi-Jack> Seriously. :)
16:06 < Psi-Jack> I know I'm old. Get off my lawn! :)
16:07 < winsoff> You'd think that old people would get over their lawns and put in a garden already.
16:07 < dbolser> winsoff: when I fire it up it says "BROKER=tls:www.xxxxx.co:port=8443" and then hangs. Then "netstat | grep -w 8443" gives me "tcp        0      1 10.16.170.127:40776     bb.bbb.bb.bbb:8443      SYN_SENT"
16:07 < winsoff> dbolser, what program is this?
16:07 < dbolser> I can't guess 40776 in order to proxy it
16:07 < dbolser> winsoff: python twisted.internet
16:08 < winsoff> Also azarus: I don't see the 3.8 release article on their site. What gives?
16:08 < azarus> winsoff: not quite updated yet
16:08 < azarus> ncopa is investigating it
16:08 < azarus> but it is released and on the mirrors
16:09 < winsoff> Does grsec sacrifice performance for security?
16:09 < Psi-Jack> Of course it does.
16:09 < azarus> blanket statement
16:09 < winsoff> Ah; my bad.
16:09 < azarus> but it's dropped anyways for 3.8
16:10 < winsoff> Why so?
16:10 < azarus> winsoff: no longer maintained
16:10 < azarus> grsec has gone private
16:10 < winsoff> RUDE
16:10 < Psi-Jack> That's like asking, "Why do I need a door to my house? It just slows me down from entering my home."
16:10 < azarus> nah
16:10 < winsoff> Psi-Jack, that's a valid question, though.
16:11 < winsoff> Some places really don't need doors.
16:11 < azarus> your mom's house
16:11  * azarus dabs to the next universe
16:11 < azarus> </immaturity>
16:11 < winsoff> See, DAB is a younger term than orly.
16:12 < winsoff> How can alpine be so small?
16:12 < Psi-Jack> winsoff: muscl.
16:12 < winsoff> What takes up most of modern linux distros?
16:12 < DLange> is this a quiz?
16:12 < azarus> Psi-Jack: musl libc*
16:12 < Psi-Jack> Modern software.
16:12 < winsoff> Speaking of, do people have to recompile to target musl libc?
16:13 < azarus> winsoff: yes
16:13 < winsoff> Interesting.
16:13 < azarus> that's also why precompiled stuff doesn't work on alpine
16:13 < winsoff> DLange, have you ever had a conversation with a novice? What is 4+7?
16:14 < DLange> winsoff: mind switching off that talk script. Or whatever it is.
16:14 < dbolser> is it possible to guess the port?
16:15 < winsoff> dbolser, perhaps it is configured in the twisted.internet module
16:15 < winsoff> DLange, the what?
16:15 < dbolser> winsoff: it's different every time... ticking up at a constant ish rate
16:15 < dbolser> but I can't get in front of it
16:16 < azarus> i don't see anything, irssi is prolly ignoring it
16:20 < winsoff> dbolser, you might be able to just change the port somewhere in there.
16:20 < triceratux> winsoff: so basically yeah ubuntu budgie. https://ubuntubudgie.org/
16:21 < winsoff> triceratux, interesting.
16:22 < winsoff> With this in mind, since DE is really interchangeable in many ways, what exactly differentiates distros?
16:22 < triceratux> its flat-themed, its modern buntu
16:22 < triceratux> winsoff: the boottime initrd
16:22 < dbolser> winsoff: I think the OS picks a local port
16:23 < triceratux> winsoff: the main difference in distros is packagemanagement & repomanagement. followed by dev attitude & community attitude
16:23 < akk> And frequency/testing of updates.
16:24 < winsoff> Makes sense. How does ubuntu have such old packages if it is such a popular distro?
16:24 < triceratux> which is a polite way of saying dev attitude
16:24 < Psi-Jack> moony: Not definite anymore?
16:24 < azarus> winsoff: because it's released semi-annually with LTS's every two years
16:26 < triceratux> winsoff: theres a spectrum of package oldness. ubuntu falls favourably on it in terms of currentness. it takes a lot of attention to have packages more current than that. you have to cater to one of these rolling releases
16:26 < winsoff> Good points, triceratux and azarus.
16:26 < revel> Musl(
16:27 < azarus> alpine is nice in that regard, imo. releases circa every half-year, but you can always use edge if inclined
16:27 < revel> Err. Musl.
16:27 < azarus> (which I do, works fantastically)
16:27 < Psi-Jack> musl has no muscles.
16:27 < revel> Was scrolled up, ignore.
16:27 < azarus> musl has a tendency for correctness tough ;)
16:28 < azarus> and as such, sometimes incompatible with glibc.
16:28 < azarus> and that's OK!
16:29 < widp> What exactly is jack/alsa and pulseaudio?
16:29 < azarus> widp: parts of the linux audio jungle
16:30 < Psi-Jack> jack, alsa, and pulseaudio are three completely different things.
16:30 < widp> yeah, I am right in the middle of the jungle.
16:30 < BeforeClick> get a banana then
16:30 < Psi-Jack> ALSA is the Linux Sound System itself. That's the core kernel drivers to audio.
16:30 < revel> Audio APIs.
16:30 < BeforeClick> much more fun than configuring audio on linux
16:30 < widp> Someone told me I need to get jack for audio intensive applications to work.
16:30 < azarus> this explains is pretty well: http://0pointer.de/blog/projects/guide-to-sound-apis.html
16:30 < widp> so I installed it and wrote some incantations as scripts.
16:31 < Psi-Jack> Jack is an audio mixer sound server system. PulseAudio is a sound mixer sound server.
16:31 < widp> Now sound works sometimes and doesn't , other times. I am determined to fix this problem once and for all.
16:31 < revel> 0pointer.de is Poettering's blog (the creator of systemd and PulseAudio), so, might want to take it with a grain of salt.
16:31 < widp> Psi-Jack: What's the difference?
16:31 < MrElendig> widp: you need jack if you want low latency, not for "audio intensive" whatever that means
16:31 < MrElendig> need/WANT
16:31 < MrElendig> note: most people never need low latency audio
16:31 < widp> MrElendig: right, things like DAWs
16:32 < widp> thanks for the link azarus
16:32 < MrElendig> most of the time one want glitch free over low latency
16:32 < azarus> widp: do mind it's the inventor of pulseaudio, which I don't like/use at all :P
16:32 < drzacek> Hello there
16:33 < widp> after half assing my way through a lot of write-ups and docs, I figured I need pulseaudio to make it so that I could use both alsa and jack.
16:33 < akk> You need alsa to use pulseaudio, but not the other way around.
16:33 < azarus> well -- i don't use jack/pulseaudio at all, just alsa, and everything works fine
16:34 < Psi-Jack> widp: Most things use PulseAudio for it is easy and simple to use, and generally the standard. JACKd is for low-latency audio stuff that... Yeah, I don't get it either. And I hate it. jackd is such a PITA.
16:34 < MrElendig> pa tend to make life easier though
16:34 < widp> right, I read alsa is what comes default with my distro(ubuntu).
16:34 < akk> You do need pulseaudio to get sound in firefox, unless you build firefox from source.
16:34 < learningc> I have a command line application that I can see using "ps aux".  How can I get it to show in my terminal again?
16:34 < drzacek> In my company network there is a server that hosts our website - only accessible from within local network. Problem is, this is windows-oriented net and I'm one of the few guys that work on linux, and the host name isn't recognized by dns
16:34 < MrElendig> learningc: depends on what you did to it
16:35 < MrElendig> if you disowned it then you don't, unless you do some really hackish stuff
16:35 < drzacek> I am able to access the website via IP address, but it doesn't resolve the name. Now I'm not sure how the networking magic works, but is there a way to fix it?
16:35 < akk> learningc: Try: jobs. If it's listed there (maybe you backgrounded it) you can get it back with fg.
16:35 < MrElendig> drzacek: fix your dns ?
16:35 < pankaj> What is so special about xterm. I remain hearing about it.
16:36 < drzacek> MrElendig, I checked the resolv.conf, seems okay
16:36 < MrElendig> drzacek: ask the network admin what they are using, maybe they are relying on mdns?
16:36 < drzacek> and if they do?
16:36 < MrElendig> then install and run avahi
16:36 < ansyeb> hello. our zabbix fixates spikes in "context switches p/sec" during times when nginx stops serving requests even tho server resources are by far not depleted, nothing calls oom-killer or whatnot. at some point "show porcesslist" amount of queues begins to grow and stops responsing back to php seemingly. stack is: nginx-php5-mysql(myisam+innodb), myisam doesnt have enough memory to keep indexes there at all times. what could be the cause of that and where
16:36 < ansyeb> to look for the issue?
16:36 < MrElendig> your network admin should know what to do
16:37 < learningc> MrElendig, I was running the command line application in gui terminal. The Ubuntu showed login screen without really logged off.  I went back in and my terminal disappeared, but I can see the application is still alive.
16:37 < learningc> akk, Nothing showed with jobs
16:37 < akk> If the terminal you ran the program from is gone, there's probably no straighforward way to get it back.
16:38 < drzacek> MrElendig, ok thanks
16:38 < Pentode> you _could_ use gdb or something to reattach it
16:38 < Pentode> if you google it yo can find some instructions if its that important..
16:38 < learningc> Pentode, how I can actually do it?
16:39 < Pentode> but like akk said it's not very straightforward
16:39 < learningc> Ok.
16:39 < learningc> I know the pid though
16:40 < Pentode> i dont remember exactly. you have to create a pipe then use gdb -p PID close stdin and call open directed to the pipe
16:40  * Pentode consults google
16:41 < Pentode> https://unix.stackexchange.com/questions/31824/how-to-attach-terminal-to-detached-process
16:41 < Pentode> you can try that
16:41 < WhiteDevil> I am thinking about giving python a try
16:41 < Pentode> think again.. ;)
16:42 < learningc> Pentode, ok thanks
16:42 < MrElendig> learningc: did you happen to click switch user or suspend?
16:42 < MrElendig> suspend/lock
16:43 < learningc> MrElendig, no, it just happen that it goes to lock/login screen randomly when I run this application in desktop terminal. That application is bitbake if you are curious.
16:46 < azarus> akk: alpine compiles firefox so it works with alsa :)
16:46 < akk> azarus: alpine is a distro? I only know the mailer by that name.
16:47 < revel> Yes, Alpine Linux.
16:47 < azarus> alpine linux*
16:47 < azarus> yu
16:47 < azarus> yup*
16:47 < akk> I'm glad to know there's at least one distro that hasn't drunk the pulseaudio koolaid.
16:47 < azarus> (me too, but don't quote me)
16:48 < zzZZzzZz> jackaudio
16:49 < akk> I use use plain alsa. I don't do anything complicated, just occasionally play a youtube video or some mp3 audio through speakers attached to the monitor.
16:50 < akk> I don't really see the point of running a big framework that's designed to control a network of external speakers.
16:50 < azarus> akk: sndio is also cool
16:50 < bb36e> my main annoyance with linux audio is that the only way to get a good systemwide EQ without glitches or delay is using PA -> JACK ;/
16:51 < webstrand> I have some code that gets run by sudo. It sets ruid = euid, but I haven't been able to figure out why. As far as I can tell, the real uid doesn't affect system permissions at all. Is this true?
16:51 < learningc> Pentode, unfortunately it has not worked
16:51 < akk> I've had so many situations over the year where the answer to "why don't I hear any sound?" was apt-get purge pulseaudio.
16:51 < akk> I hear it's less flaky now but it's hard to believe that.
16:52 < Pentode> sorry. at least you tried. ;)
16:52 < azarus> akk: alsa on linux/sndio on OpenBSD have been the most troublefree audio solutions I've used
16:53 < azarus> (you can use sndio even on linux)
17:05 < noway96> what does this mean: attempting task abort! scmd(ffff8810142d1380)
17:06 < noway96> [sg12] tag#10 CDB: opcode=0x28 28 00 3d c9 06 80 00 00 80 00
17:06 < noway96> it's executing a READ(10) command
17:06 < webstrand> noway96, looks to do with your sata controller, are your disks healthy?
17:07 < noway96> webstrand, no
17:07 < noway96> but it shouldn't cause a memory leak yeah?
17:07 < webstrand> nope, where are you seeing this message?
17:07 < noway96> i.e. something that causes totally unrelated programs to crash
17:07 < noway96> in /var/log/messages
17:08 < noway96> can too many zombie processes cause a totally unrelated program to crash?
17:08 < webstrand> pid exhaustion can prevent new processes from starting
17:08 < webstrand> poorly written programs may crash when the cant spawn anything new
17:08 < noway96> ok
17:09 < noway96> I have a QtWebKit crashing because of something in a shared library. I have no idea why though
17:09 < webstrand> same with a failed read, if the program doesn't expect it to fail
17:10 < noway96> Anyway, my quick hack to is to just restart the browser when it crashes hehe
17:10 < webstrand> I'm guessing you're looking at a backtrace?
17:11 < noway96> yeah but it's meaningless. A backtrace behind a bunch of .so files
17:18  * EvilRoey notes precise is also The Chatter Formerly Known as Prence`
17:18 < EvilRoey> *prence is
17:19 < prence> only because they won't let me use that weird little male/female hybrid symbol
17:20 < dreadkopp> hey community. got 10 1-to-1 intel nuc clones in front of me. i set up the first, then booted a live system and dd'ed the internal drive to another ssd. However: booting up the cloned machines they are incredible slow. even simple tasks take long time in the terminal
17:24 < dreadkopp> well, it's especially the sudo commands ...
17:25 < bratchley> that could be any number of things, you just have to kind of poke around and try to see what's broken
17:25 < bratchley> usually that means it's trying to do something and can't
17:26 < bratchley> I'd check networking first
17:36 < jhodrien> dd isn't necessarily the best tool to do clones with, given you often don't want an exact clone.
17:39 < dreadkopp> jhodrien i wanted 1:1 clones though... seems to have been the network though .@ bratchley
17:39 < bratchley> that'll do it
17:39 < jhodrien> Sure, but it's not normally quite 1:1.  You often end up with things like MAC addresses lurking in config files.
17:40 < ahoo> what is the kind of software called where all information about (internal or customer) computers, devices, appliances etc. are being maintained in?
17:40 < jhodrien> inventory ?
17:41 < ahoo> no.
17:41 < jhodrien> CMDB?
17:41 < ahoo> also no.
17:42 < ahoo> these programs often "discover" nodes
17:42 < bratchley> CMDB is pretty much the ITIL term for it
17:42 < bratchley> that's literally what a CMDB is
17:43 < bratchley> are you thinking of the word IPAM ?
17:44 < bratchley> just trying to think of something that does network discovery and might feed into a CMDB system
17:44 < ahoo> thanks, i've used CMDB.
17:44 < jhodrien> Always fun doing other people's homework for them.
17:45  * Psi-Jack cracks the whip glaring at jhodrien. "Well, get to it then, boy!" :)
17:46 < bratchley> CMDB isn't a product, it's the role a piece of software can play within your organization. It's the piece of software that tracks assets and records basic software information for the various systems and services running on your network
17:46 < jhodrien> (I'm an ITIL shop)
17:46 < bratchley> jhodrien: don't be so hard on yourself
17:46  * bratchley not a big fan of ITIL as a whole
17:47 < jhodrien> I'm not a fan of their choice of language in places, and I think you can do horrible things with most frameworks.
17:47 < bratchley> the problem with ITIL is that when you actually do more than a trivial amount of ITIL it gets bureaucratic as hell
17:48 < bratchley> tons of paperwork and niche termininology, lots of meetings with many of them seemingly redundant
17:49 < mawk> what's ITIL
17:49 < bratchley> It's a process management paradigm
17:49 < bratchley> formalizes things like change control and documentation management
17:49 < mawk> like agile stuff ?
17:50 < bratchley> in a similar vein yeah
17:50 < mawk> I have terrible memories of managers forcing us to do agile just for the sake of it
17:51 < Psi-Jack> Heh
17:51 < Psi-Jack> Agile can work great for development, where there's a constant influx of things going on. But agile doesn't work for all.
17:51 < jhodrien> Lots of common language stuff.  Process/Function/Problem/Request/Incident/Change etc..
17:54 < bratchley> yeah
17:56 < bratchley> I'm also a bit jaded with ITIL because two of the places I worked at that stressed ITIL had somewhat toxic management practices. rank and file successes were examples of their excellent leadership while failures were explain by the implementors' lack of adherence to this or that part of ITIL.
17:56 < bratchley> that might be tainting my view of it as well tbh
17:56 < pressure679> Hi, I am shameful to admit, as Linux user for years, that now is my 1st time building and patching the kernel myself. Any tips?
17:57 < bratchley> Like using change control to defer blame onto their subordinates
17:57 < jhodrien> bratchley: Sure.  There's a lot of ITIL that you can look at and see as sane, until you see it being used to beat people with.
17:57 < Psi-Jack> pressure679: Don't. There's generally no need to do so.
17:57 < jhodrien> You're not to blame, you're accountable.
17:57 < jhodrien> ;)
17:57 < bls> I got a similar thing with Agile. the "agile advocates" in the company were all or nothing...you followed all the process, even it if didn't make sense and if there were problems it was because you weren't following agile exactly. then some managers found it was a wonderful tool for micro-management
17:57 < ntd> pressure679, what, you haven't been using grsec? :P
17:58 < Psi-Jack> Heh.
17:58 < bls> pressure679: do it as a learning experience, consider that phase of your education done, then move on to something more useful
17:58 < Psi-Jack> Agile pretty much is micromanagement.
17:59 < bls> accounting for accounting's sake
17:59 < bratchley> exactly, I'm all for change control, but it seems like a lot of ITIL is designed to create management positions that do only the ITIL stuff. Like a lot of project management is only necessary to check all the ITIL boxes.
18:00 < Psi-Jack> lol
18:02 < capcj> For me, all of that shiny methods and words is for a couple of people that don't want to dive too much into technical stuff but keeps trying to live in the IT field
18:03 < capcj> Sry guys, just my opinion
18:03 < Psi-Jack> Standard English, please.
18:03 < bratchley> Somewhat true, I can't count how many people who use server counts to judge change importance or storage capacity to judge the difficulty of a change
18:03 < bratchley> I'm not even kidding about that
18:04 < capcj> Psi-Jack, what I did wrong?
18:04 < Psi-Jack> capcj: "sry"
18:04 < capcj> ok
18:07 < bratchley> heh I remember one change that had to go through change control
18:08 < bratchley> it was for the anti-virus that the ISO department said the Linux systems absolutely had to have (Symantec to be exact)
18:08 < bratchley> I went through the DEV/TEST steps and then submitted the results as part of the change control item to the change control board
18:09 < bratchley> then had to join the meeting on the phone and wait for my item to come up because they supposedly didn't understand it.
18:09 < bratchley> when I explained what it was, they acted like it was a stupid change and I should just go ahead and do it. Then canceled the change.
18:10 < bratchley> as in not "declined" the change just canceled it with the idea that I would I guess be installing anti-virus on production systems without an approved change control
18:11 < kubast2> Hey ,how can I get first column from zram0/mm_stat ?
18:11 < kubast2> or any other one
18:11 < capcj> @bratch
18:11 < kubast2> anything other than pipe awk "{print $1}"
18:12 < capcj> @bratchley, normal stuff
18:12 < lulzake> kubast2: you got problem with awk?
18:12 < bratchley> capcj: for that place yeah unfortunately
18:12 < nixx> 40 billable man hours worth of meetings and chasing signatures so I can move a plug from one circuit to another, on a system that has redundant power.
18:12 < bratchley> setting me up to either silently succeed or get written up for changing production without a change control item
18:13 < capcj> Almost all the places with some fancy stuff kind of management where usually the manager doesn't know a sh*t about how the things works have situations like that
18:13 < kubast2> well sudo su -c 'cat /sys/block/zram0/mm_stat | awk "{pring $1}" ' prints out the whole column lulzake
18:13 < MrElendig> kubast2: drop the cat
18:13 < kubast2> you mean like awk instead of cat
18:13 < kubast2> sure
18:13 < MrElendig> and the sudo su
18:13 < capcj> 200 hours in sprints, 20 really diving into code, weird
18:14 < bratchley> in this case I think they understood it well enough. I think the idea was to get Anti-Virus on the systems since ISO was wanting it but to not have their names on an approved change control in case it broke something
18:14 < bratchley> that way it's my fault somehow (either I would have "misunderstood" something or didn't explain something well enough, etc)
18:14 < kubast2> MrElendig, su ed into root
18:14 < MrElendig> you became root to become root.....
18:15 < MrElendig> i heard you like shells so I put a shell in your shell so you can.
18:15 < kubast2> MrElendig, sudo su ,because sudo isn't enough for /sys afaik
18:15 < MrElendig> :p
18:15 < MrElendig> wrong
18:15 < kubast2> and I often get Lack of permissions
18:15 < kubast2> when I use sudo
18:15 < kubast2> instead of sudo su
18:15 < MrElendig> that is beacuse you don't understand how shells work
18:15 < kubast2> o /sys
18:15 < kubast2> yeh
18:15 < MrElendig> not because sudo doesn't work
18:15 < bls> maybe you need more sudos
18:16 < clemens3_> one root to rule them all
18:16 < MrElendig> sudo echo foo > /sys/bar will run echo as root, but the redirection will be done by your shell running as your user
18:16 < kubast2> ah
18:17 < MrElendig> echo foo | sudo tee /sys/bar will  run echo as your user, but tee as root
18:17 < kubast2> ik that
18:17 < prussian> sysctl is a thing as well
18:17 < bls> or sudo sh -c 'echo foo > bar' is the other pattern for this
18:17 < lulzake> I have /etc/hosts with first line `127.0.0.2 A A` and second line `127.0.0.3 B B`. direct access to 127.0.0.3 is ok but acess with hostname B redirect to 127.0.0.2. any idea?
18:18 < clemens3_> NOPASSWD: ALL also wors wonders
18:18 < clemens3_> wors=works
18:18 < bls> not really related to the problem at hand
18:19 < kubast2> prussian, for some /sys stuff I use it ,but I change directlly when I don't know how to find something I know a path to on /sys
18:19 < kubast2> afaik the convention for sysctl is like folder.option
18:19 < kubast2> =value
18:20 < SirLagz> bleh
18:21 < capcj> pring?
18:21 < kubast2> +I use sudo su -c ' ' except the fact I didn't knew it run redirect as my own user(I did knew pipe and and or does) MrElendig
18:21 < kubast2> because I like the fact it remembers the password
18:22 < kubast2> I could drop into sudo -s
18:22 < kubast2> sure
18:22 < MrElendig> there are no reason to run sudo su ever
18:22 < MrElendig> except if you are on a 20 year old operating system like some of the UNIX™ boxes
18:22 < kubast2> Only because I rather sudo su -c rather than sudo -s :shrug:
18:22 < MrElendig> inb4gettingsuedfortradenameviolation
18:22 < MrElendig> why?
18:23 < MrElendig> also those do different things
18:23 < triceratux> what if su asks for a password & you can forego that via your sudo authorisation ?
18:24 < kubast2> I usually use the same long password for root and my user ,since it's a personal usage machine and nothing of value is in here anyway
18:24 < MrElendig> just use sudo
18:24 < bratchley> MrElendig: IIRC RHEL4 didn't have "sudo -i"
18:24 < bratchley> regarding the "20 year old" thing
18:24 < kubast2> MrElendig, how to redirect the output as root when using sudo ?
18:24 < kubast2> on nonprivledged
18:24 < MrElendig> bratchley: and it is about 200 years old
18:24 < MrElendig> :p
18:25 < MrElendig> bratchley: could just sudo bash though
18:25 < MrElendig> instead of opening an extra shell
18:25 < paulcarroty> https://old.reddit.com/r/linux/comments/8tzu99/til_microsoft_is_a_platinum_sponsor_of_debian_lts/ hahah
18:25 < bratchley> or "sudo su -" which has a purpose on that system
18:26 < MrElendig> which is moot since it haven't been supported for years anyway
18:26 < bratchley> EOL doesn't mean it's gone, the only point I was modifying was "there are no reason to run sudo su ever"
18:27 < kubast2> well besides that I still don't know how to read this column:
18:27 < kubast2> http://termbin.com/nb0e
18:27 < kubast2> how to read a column from this line
18:27 < MrElendig> bratchley: as mentioned, even then sudo bash would in most cases do the job anyway
18:28 < MrElendig> either cut, awk or ex will do just fine
18:28 < bratchley> we're kind of talking about this more than I thought we would but the point is that in that context the lack of "sudo -i" is in fact a reason to do "sudo su -" regardless of whether there are other ways of accomplishing the same thing
18:29 < MrElendig> echo "1368064   367723   745472        0   745472        0        0" | awk '{ print $2 }'
18:29 < mediapifan> hey guys
18:29 < kubast2> idk why did it worked this time MrElendig
18:29 < kubast2> wth
18:29 < mediapifan> first time on an IRC channel looking for help
18:29 < bratchley> kubast2: probably misfortune due to evil acts in a previous life
18:30 < kubast2> nah
18:30 < kubast2> sudo su -c
18:30 < MrElendig> mediapifan: print out and keep under your pillow: http://www.catb.org/esr/faqs/smart-questions.html  :)
18:30 < kubast2> lel
18:30 < ferrets2> keep getting time drift on some of my VMs on vmware
18:31 < bratchley> normal for VM's
18:31 < bratchley> due to time stealing
18:31 < ferrets2> only normal if there is cpu scheduling contention
18:32 < bratchley> which is eventually going to happen
18:32 < bratchley> unless the system is basically completely idle
18:32 < bratchley> that's why people usually recommend configuring NTP on VM's
18:32 < ferrets2> it is only configured with NTP
18:33 < bratchley> so then your problem is that NTP isn't correcting the drift?
18:33 < MrElendig> use ntp on everything unless you have really good reasons not to
18:33 < ferrets2> only have 2 ntp sources in our org
18:33 < MrElendig> or the simpler subset of ntp
18:34 < bratchley> ferrets2: are there errors you're getting when trying to sync the time then?
18:34 < sacules> is there a vim plugin that shows you the definition and docstring of the function or variable the cursor is on?
18:34 < ferrets2> no errors
18:34 < MrElendig> sacules: multiple
18:34 < MrElendig> some general, some language specific
18:35 < sacules> was looking for some general ones
18:35 < bratchley> that seems weird, for NTP to think it's fixing clock skew but to not actually be
18:35 < sacules> like, WingIDE has this very cool source assistant that does exactly that, but only works for python
18:36 < MrElendig> sacules: see ycm etc
18:36 < atmtt> hello i was looking for a list of freenode channels from emacs. How can i list it?
18:36 < bratchley> YouCompleteMe is a popular auto-complete VIM plugin, I don't know if it has documentation tie-ins though
18:36 < MrElendig> optinal when combined with other modules
18:36 < MrElendig> optional*
18:37 < bookworm> atmtt: /msg alis list $query
18:38 < sacules> yeah, wasn't looking for autocompletion, just something like this http://i.imgur.com/6kr0eBW.png
18:39 < sacules> that's when the cursor is over a variable containing a list
18:39 < capcj> Some plugins turn vim into a slow mess of editor. I guess YCM uses a outside layer to bring that info, that's why sometimes is such a burden to me (or maybe it's my old laptop)
18:39 < atmtt> thanks @bookworm
18:39 < MrElendig> vim8 helps due to async
18:39 < MrElendig> most plugins does not take advantage of it though
18:40 < capcj> oh, sacules, you're looking for something related to the language docs and stuff
18:40 < sacules> yeah, sorry it wasn't clear
18:40 < MrElendig> there is a plugin that leverages ycm to show defenition/help/etc for the object under the cursor too, can't remember the name though
18:41 < lightslategrey> Hi! For some reason I can't use any network management manchinery in Linux. Then what roughly are my major steps to get newly plugged in NIC working? Assuming driver created interface successfully and I got notified by udev.
18:41 < MrElendig> also some of the tagbar plugins has the feature too
18:41 < capcj> Me too
18:41 < bratchley> capcj: yeah personally I limit my .vimrc to just the basic stuff like monkeying with the status line
18:41 < MrElendig> lightslategrey: define "newly plugged in"
18:41 < bratchley> if I want autocomplete, documentation, etc I'll use a regular IDE. When I'm in vim it's because I like something functional but fast
18:41 < capcj> bratchley: wise move, I'm almost doing that myself
18:41 < MrElendig> lightslategrey: hotplugged pcie? plugged in and then booted?
18:42 < capcj> when you take a big mess of code the editor almost freezes
18:42 < sacules> I'll look into ycm and related, thanks
18:42 < capcj> sacules: https://vi.stackexchange.com/questions/39/is-it-possible-to-have-vim-auto-complete-function-names-variables-etc-when-us
18:43 < capcj> autocompletion it's a little different
18:43 < lightslategrey> MrElendig: let's say it's many CDC Ethernet devices. Let's assume driver creates and removes ethX entries for them when they come and go.
18:44 < akk> Those vim plugins sounds pretty cool. Is there a way to do that sort of thing in emacs for python? Show current function, docstring, even autocomplete?
18:44 < capcj> sacules: https://github.com/vim-scripts/DoxygenToolkit.vim
18:45 < capcj> akk: Emacs can do everything, even a Supernova
18:45 < MrElendig> lightslategrey: networkmanager works pretty well for usb crap
18:45 < MrElendig> sidenote, why disable predictable naming?
18:45 < akk> capcj: Yes, you're right, and I definitely misspoke. Is there a mode that can already do it, without writing it all from scratch in elisp? :)
18:45 < MrElendig> or is this some silly distro that doesn't use systemd yet? :p
18:45 < lightslategrey> MrElendig, no it doesn't, it has weird bug that makes it useless for my task.
18:45 < MrElendig> what weird bug?
18:45 < MrElendig> and how old is the nm?
18:45 < capcj> akk: Be a Jedi - https://tkf.github.io/emacs-jedi/latest/
18:46 < capcj> I was kidding man, i got what you're speaking, hahaha
18:46 < lightslategrey> MrElendig, no, assume I have modern distro with any tools at disposal
18:47 < akk> capcj: Cool, thanks, I'll check it out. How about displaying the current function name, with or without docstring?
18:47 < lightslategrey> MrElendig, it's tool long to explain and you will not be able to help with that exact root case. Let's suppose it's multitude of devices with duplicate MAC addresses.
18:47 < MrElendig> ah, those broken china dongles :p
18:47 < MrElendig> tip: set up mac randomisation for them
18:48 < capcj> I never deeply used that, akk, shame on me, nowadays I did like Paul Graham and stiked with vim
18:48 < MrElendig> shouldn't break nm though
18:48 < lightslategrey> MrElendig, can you help me with main steps to make interface work manually? Am not expert in networks and network tools unfortunately.
18:48 < capcj> Maybe it's a Heaven's call to come back to the Emacs, I'll try that lol
18:48 < lightslategrey> MrElendig, I am sorry but it's not work
18:49 < lightslategrey> MrElendig, not gonna work to be precise
18:51 < lightslategrey> MrElendig, can you outline the general steps to make interfaces work?
18:51 < sacules> capcj: not exactly what I was looking for, but close enough, thanks
18:52 < triceratux> rule #1: dont use ifconfig https://www.howtoforge.com/linux-ifconfig-command/
18:55 < akk> Unfortunately the steps for configuring a network by hand vary with distro. You'd think there would be a standard way but it seems to vary a lot.
18:55 < akk> I have a script that does it but it's fiddly and often needs to be adjusted.
18:57 < mawk> ip is pretty universal
18:57 < lightslategrey> akk, I don't think there is standard way, but i also dont' beilive that say manually requesting address with dhcp vill vary a lot, and I am sure adding routes commands will behaves almost same way.
18:57 < mawk> for a transient network
18:57 < mawk> yeah
18:58 < lightslategrey> akk, I asked about general steps. BUt it seems people seem beilive they are geniuses but can't really understand question.
18:58 < akk> For manually requesting addresses I use dhclient or dhcpcd (which one is better depends on distro and circumstances).
18:58 < mawk> lightslategrey: you want to add temporary addresses and stuff right ?
18:59 < akk> I tend to keep both installed, so if one fails I can use the other.
19:00 < lightslategrey> akk, I have some isc-dhclient-4.3.6 installed /usr/sbin/dhclient - will try it. THANK YOU for suggestion. Will at assign response to interface automatically or I should grab the answer than parse it and the assign myself?
19:00 < MrElendig> networkd and nm is pretty universally available these days
19:01 < mawk> it will assign the address
19:01 < mawk> and the route
19:01 < lightslategrey> MrElendig, NM have bug it will not work for me. It doesn't seems to be possible without fixing NM itself
19:02 < akk> lightslategrey: I use dhclient -v $interfacename
19:02 < akk> which sets it automatically
19:02 < mawk> I don't think that it's NM that is bugged, lightslategrey
19:02 < mawk> but alright
19:02 < lightslategrey> mawk, dhclient will will also set the default route? great!
19:02 < mawk> yeah
19:02 < mawk> if the dhcp server pushed it
19:03 < mawk> also the very universal thing is to use /bin/ip, to add static addresses and routes
19:03 < mawk> or do add additionnal addresses to a system
19:04 < lightslategrey> mawk, will I need to create any routing tables?
19:04 < mawk> with dhclient ? no
19:04 < lightslategrey> mawk, or should I ensure routing tables are created?
19:04 < mawk> neither
19:05 < mawk> dhclient will use the main routing table anyway, none have to be created
19:05 < mawk> and it will add the rules automatically depending on the info in the DHCP lease
19:05 < lightslategrey> mawk, so it is all will rely on `main` table?
19:05 < mawk> well the kernel adds the route for the LAN, and dhclient adds the route towards the gateway
19:06 < mawk> yeah
19:06 < mawk> it's the one and only routing table, you don't add routing tables yourself
19:06 < mawk> well you can but normal users don't have to
19:07 < mawk> you can add secondary routing tables then make complicated rules to direct traffic in them, you can even group network interfaces together using vrf interfaces that have their dedicated routing tables
19:07 < mawk> lots of neat stuff can be done
19:07 < lightslategrey> mawk, what if don't want to change the default route (and I manually use specific outgoing interface when need to). I still have set up routing in that case?
19:07 < mawk> the default route isn't one specific route
19:08 < mawk> it's just one route that happen to be a route for 0.0.0.0/0
19:08 < mawk> so dhclient will just add a new default route, and you can remove it afterwards
19:08 < mawk> or edit dhclient config to tell it not to add default routes
19:08 < capcj> sacules: You're welcome
19:08 < mawk> or if it's just for quick testing, you can add the static ip yourself, if you already know the IP range
19:08 < mawk> you pick something that's likely to be free, or ping for it first
19:09 < lightslategrey> mawk,  greate! apart from getting IP from DHCP and setting up routing what else I will need?
19:09 < MrElendig> nothing
19:09 < mawk> nothing, you can use the new ip right away
19:09 < MrElendig> (except some sane firewall rules)
19:09 < mawk> and if you don't want a default route you don't need to setup routing
19:09 < mawk> the kernel automatially adds the route to the LAN when you add an ip
19:09 < mawk> see the output of `ip route', these routes are marked "proto kernel"
19:10 < mawk> even if you can cheat and add proto kernel routes yourself
19:10 < lightslategrey> thank you people. not seems more clear. going to go read. at least I have a plan how. thank you!
19:10 < jim> the dhcp should set routing up for you (that's usually how it works and what it's for)
19:10 < mawk> yeah
19:10 < lightslategrey> i am sorry: noW seems more clear
19:10 < jim> no worries...
19:10 < mawk> no need to setup routing in any case, except if you want to remove routes added by dhcpd
19:11 < mawk> in that case just copy the route line and put if after `ip -4 route del'
19:11 < jim> does ip addr show you have addresses on your interfaces after dhcp has completed?
19:11 < mawk> yes
19:12 < mawk> or you mean in his case ? he didn't try yet
19:12 < mawk> but since its dhclient is no different than ours, I can predict the future and tell it will be the case
19:12 < jim> that makes me itly curious as to what's stopping him from trying
19:13 < mawk> lol
19:13 < mawk> yeah actually lightslategrey if you want a secondary ip, you have special steps to take when using dhclient
19:13 < jim> idly even
19:13 < mawk> using dhclient will just tell you again the IP you already have
19:14 < mawk> in that case the easiest thing is to pick a static ip yourself that's unlikely to be taken
19:14 < mawk> like x.x.x.253
19:15 < lightslategrey> I am reading "man dhclient". I want to understand what I am doing. Also the set up is more complex. I have separate routing tables and about 10 ppp interfaces and it's all managed from sotware I write. Now I want try CDC not NM have a bug.
19:15 < mawk> are you sure it's separate routing tables ?
19:16 < mawk> and not just separate routes
19:16 < mawk> they mean different things
19:17 < mawk> different routing tables are used with policy routing, basically a set of rules/commands that get executed by the kernel for each packet to pick and tune the corresponding routing table
19:17 < mawk> and different routes are just what go in a routing table, usually the main routing table and nothing els
19:17 < mawk> e
19:18 < lightslategrey> I can't run this machine now. Yes, I barely remember. It was separate rounting tables. I dinamically created them and they was shown "ip rule show". It was year ago and I forgot all details. At that moment I got full understanding, but lost it now since year.
19:18 < mawk> alright
19:19 < revel> Hmm. Is there any utility that gets input piped from stdin, replaces all newlines with linefeeds (\n with \r) and feeds it all back to the terminal (so that every new line of output replaces the old one)?
19:19 < bratchley> revel: tr
19:19 < revel> Didn't seem to work.
19:19 < mawk> tr does it on the fly ?
19:19 < Dagmar> \r is carriage return.  Newlines *are* linefeeds
19:19 < revel> Yeah, carriage return.
19:20 < mawk> tr seem to be doing buffering, it doesn't work with what revel is asking for
19:20 < jim> maybe you can say something about your overall goal that you have right now
19:20 < eset> had someone issue with repository? I'm adding repo http://repo.zabbix.com/zabbix/3.4/debian/pool/main/z/zabbix-release/zabbix-release_3.4-1%2Bjessie_all.deb using ansible and even manualy from CLI and in source.list.d I don't see any new file :/
19:20 < mawk> I'd write a quick script, but I'm sure you have other ways
19:20 < bratchley> revel: works for me
19:20 < Dagmar> Newsflash, pretty much everything is going to use buffers for this
19:20 < mawk> maybe an expect script will be easy
19:20 < bratchley> revel: https://dpaste.de/owcj
19:20 < revel> bratchley: I mean on the go.
19:21 < revel> Well, basically, I have something like `tail -f | grep $someexpression` and it's outputting something like "[9971/31680]"
19:21 < Trel> Is there a good or generally reccomended user/group TUI?
19:21 < mawk> I'm not an expert in expect scripts but I'm sure it's 4 lines maximum
19:21 < Dagmar> Either way, converting \n to \r alone is probably going to be wrong.
19:21 < mawk> it seems to work theoretically Dagmar
19:22 < mawk> well not, you're right
19:22 < Trel> Dagmar: did any OS use \r alone?  I know Windows is \r\n and Linux is \n
19:22 < mawk> you need to add a few terminal control characters to at least clear the line
19:22 < mawk> not just go back to the beggining, otherwise you'll have overlapping text
19:22 < bratchley> Dagmar: unless I'm missing something the original question was to replace \n with \r
19:22 < Dagmar> Trel: None that I've seen
19:22 < mawk> Trel: old mac os
19:22 < revel> Overlapping text shouldn't happen in this case.
19:22 < mawk> why revel ? lines are padded with lots of spaces ?
19:22 < mawk> it's the only reason that can prevent overlaps here
19:23 < mawk> or if all the lines are the same length
19:23 < Dagmar> bratchley: Yeah but pretty much nothing that should still be in use uses *only* \r/
19:23 < revel> It's in the format of "[9971/31680]"
19:23 < revel> So, yeah, same length.
19:23 < mawk> for safety you can add a few spaces at the end
19:23 < mawk> and it will be perfect
19:23 < mawk> now you'll have a nasty cursor flickering effect
19:23 < Trel> mawk: old as in OS9 and earlier, or old as in boots off 5.5in floppies?
19:23 < mawk> but you can maybe work your way around this
19:24 < mawk> Trel: I don't remember, I think it was the mac os before os x
19:24 < mawk> the whole series
19:25 < bratchley> Dagmar: don't older Mac systems use \r instead of \n ?
19:25 < adrian_1908> hello. does anyone know by which means the openvpn client updates the routing table? Is this done through a script, or a feature within the openvpn binary?
19:25 < Dagmar> Wait... So... you're trying to get a completion ratio to basically sit on the same place in the screen?
19:25 < akk> Mac used to use \r\n IIRC
19:25 < Trel> Ah ok.  I used to use OS9 via Sheepshaver on windows to play an old Mac only game I bought until there was away to play it via the emulator for those old Lucasarts game (the emulator name is escaping me at the moment)
19:25 < bratchley> akk: \r\n is Windows
19:25 < mawk> revel: a quick python script will do the job, you just need to enable line-buffering on stdin, then a loop that replaces the terminal \n by \r
19:25 < akk> bratchley: Oh, you're right, oops.
19:26 < Dagmar> bratchley: Nothin that anyone should still be feeding power to.  That's _very_ old.  Like 90's-relevant old
19:26 < mawk> and maybe some control code trickery to prevent flickering/overlap
19:26 < akk> Then yes, mac was \r.
19:26 < Dagmar> There's multiple easier ways to do this.  THe most efficient of which isn't exactly obvious
19:26 < bratchley> but it seems like converting \n to \r in a pipe is the wrong way to re-write the status, otherwise if the program outputs a newline at the end then you're left with a blank space
19:27 < Dagmar> The lazy way would be to use tput to set the text window size to one line high
19:27 < bratchley> if the goal is to continually overwrite the same line
19:27 < Dagmar> That way each line would simply scroll off and replace the previous one
19:28 < mawk> let me research a bit the right control codes to prevent flickering revel
19:28 < bratchley> might make printing errors more complicated though
19:28 < revel> What errors?
19:29 < bratchley> If the terminal is one line high and you need to display an error rather than a rolling status
19:29 < bratchley> errors are often more than one line and probably wouldn't be on the same line as the status
19:29 < ThePendulum> 'lo
19:29 < MrElendig> can still use stderr for that
19:29 < ThePendulum> is it typical that formatting a 6TB drive to ext4 results in 45GB being used preliminarily? o_o
19:29 < MrElendig> ThePendulum: normal
19:29 < MrElendig> ThePendulum: also don't forget the 5% default reserved
19:29 < revel> I don't think it'd really need any error reporting...
19:29 < ThePendulum> that's quite a lot compared to NTFS' supposed 200MB, but ext4 allows me to use ? :, so eh, worth it over 6TB I guess
19:30 < Dagmar> You can disable the reserved part if this is strictly a user-centric store
19:30 < bratchley> ThePendulum: NTFS relies on defragmentation processes whereas ext3/4 tries to preempt file fragmentation
19:30 < Dagmar> It's mainly only appropriate for system-oriented filesystems
19:30 < ThePendulum> Dagmar: what are some criteria for that?
19:30 < ThePendulum> I see
19:31 < ThePendulum> well it's an external HDD, there isn't going to be an OS on it or anything
19:31 < MrElendig> sad thing is that MS made a better fs, but recently removed support for it from win10 home/pro
19:31 < MrElendig> do "differentiate" their versions
19:31 < Dagmar> Like, if you're mounting a 6Tb volume at /my_mp3s, then perhaps root has zero interest in having guaranteed space
19:31 < Dagmar> ThePendulum: Yeah, set the root reserved amount to zero for that
19:31 < revel> MP3s? Please, 6TB is for pr0n storage.
19:32 < ThePendulum> right, let me see. so system-oriented would be when I installed an OS on it or the like?
19:32 < ThePendulum> revel: get out! who would do such a thing!
19:32 < ThePendulum> got a 64TB bay for that
19:32 < ThePendulum> nah, I wish
19:32 < revel> If it's just for general media, then having 5% of it reserved for root seems pointless to me.
19:33 < Dagmar> ThePendulum: Yes.  If, for example, users fill up /tmp which happens to still be part of /, that means / is full as far as they're concerned, but root still has another 5% of the disk available
19:33 < Dagmar> Otherwise, THIS WOULD BE REALLY BAD
19:33 < ThePendulum> makes sense
19:34 < ThePendulum> so would those 5% show up as Used in (g)parted? since 45GB is less than 5%
19:35 < ThePendulum> rather where is this arranged, when mounting or when formatting?
19:35 < MrElendig> forget about what gparted things
19:35 < MrElendig> thinks*
19:35 < MrElendig> and parted doesn't do filesystems anymore
19:36 < ThePendulum> I'm thinking of fdisk, my bad
19:36 < MrElendig> fdisk doesn't do filesystems either (and never has)
19:36 < ThePendulum> what
19:36 < ThePendulum> what do you mean by "do filesystems" exactly
19:36 < MrElendig> they don't know about nor care about the filesystems, or lack of, on the partitions
19:36 < revel> It's for partitions.
19:37 < MrElendig> sidenote: google "new ext4 partition used space"
19:37 < MrElendig> you will find the question repeated for the past decade or so :)
19:38 < MrElendig> it's not just the reserved space, quite a bit is taken up but the tree and the metadata itself
19:38 < mawk> revel: I'd insert \033[1A\033[2K at the beggining of every line, and just leave \n as it is
19:38 < mawk> that way the script is adaptable to every situation, and there'll be no cursor flickering nor text overlapping
19:38 < MrElendig> there are some flags/options for ext4 that can quite heavily change the amount too
19:38 < mawk> just remember to not output it for the first line and you're good
19:38 < mawk> let me write the script
19:39 < revel> Ooh, thanks.
19:40 < ThePendulum> hm, can I read the current reserved space rather than just setting it
19:40 < mawk> ESC [ 1 A goes back one line up, and ESC [ 2 K clears the whole line
19:41 < ThePendulum> ah df to the rescue
19:41 < mawk> there is ESC [ 1 ` to mimic \r to get back the cursor at position 1 but it doesn't work with my konsole, and I don't want to change the tty output mode to ensure that \r works as expected
19:41 < Dagmar> Set the window size, dude
19:41 < Dagmar> We used to do this to people to screw with them
19:42 < Dagmar> ...and I'm not talking about the xterminal size
19:42 < eset> is there a way to check for stretch and jessie is service is enabled?
19:43 < eset> I mean it has to be the same command
19:43 < Dagmar> One with systemd and one without?  Nope
19:45 < Dr_Coke> triceratux
19:45 < Dr_Coke> rindolf jim Psi-Jack  hi
19:46 < eset> Dagmar: jessie also has systemd ?
19:46 < triceratux> jessie was already systemd https://arstechnica.com/information-technology/2015/05/debian-8-linuxs-most-reliable-distro-makes-its-biggest-change-since-1993/
19:46 < triceratux> ^^
19:46 < eset> triceratux: but they are quite different
19:46 < MrElendig> revel: https://bpaste.net/show/134bf8976ae1
19:46 < eset> I don't see there that service is enabled for example
19:46 < eset> on jessie I can see it
19:47 < eset> on stretch
19:47 < eset> sorry
19:47 < Dr_Coke> Dagmar what's up man
19:47 < revel> lol
19:47 < Dr_Coke> revel my main man
19:48 < Dr_Coke> lol
19:48 < Dagmar> Dr_Code: Just archiving the new NIN album
19:48 < Dr_Coke> oh lovely
19:48 < ThePendulum> is it any good?
19:49 < Dr_Coke> Dagmar I didn't know NIN was still going
19:49 < Dagmar> That happens when you can't be killed
19:49 < Dr_Coke> lol
19:51 < Dr_Coke> Dagmar late July I think we are getting the HTC U12 here in Australia
19:51 < Dr_Coke> a month or more after the usa
19:52 < Dagmar> Well, it takes time for the emu to swim them across
19:53 < Dr_Coke> The emu needs a rocket attached to it
19:55 < Dagmar> That's okay.  It looks like we might not be getting the new old-people Nokia at all
19:56 < mawk> revel: https://paste.suut.in/GeGbwfEn.py
19:56 < mawk> click on "texte brut" to get the downloadable raw text
19:57 < Loshki> "texte brut" is my favorite champagne
20:03 < revel> mawk: Hmm, doesn't seem to work. Gets stuck on "read(0," after like a second according to strace
20:05 < revel> Error message is https://0x0.st/s9Y9.txt . Though I really can't expect you to do everything for me. Thinking of going to sleep now. Bye.
20:06 < mawk> yeah that's when you ^C it revel
20:06 < mawk> it's not an error
20:06 < mawk> even though I could've hidden that
20:07 < mawk> output is on stdout, maybe you expect it elsewhere
20:07 < mawk> also if the input is strange like having \n\n instead of \n nobody will show, but I can fix that
20:08 < mawk> also the script takes input from stdin revel , not from its command line arguments
20:41 < Industrial> How do I do an "inverse" of grep? e.g. `something | filter-out-x`
20:42 < jim> grep -v pattern
20:42 < Industrial> ty
20:42 < jim> check man grep for details
20:43 < jim> welcome and please spell out ty as thank you to help any new english speakers who might be present
20:43 < ldlework> lol
20:44 < RustyJ> ldlework, please spell out lol as laughing out loud
20:44 < ldlework> XD
20:45 < meyou_> i feel like using common abbreviation would help new english speakers more
20:45 < meyou_> so they can learn the abbreviations?
20:46 < mawk> yeah
20:46 < RustyJ> mawk, please spell out 'yeah' as you are correct
20:46 < Dagmar> If someone's going to go to the trouble of typing a long explanation for something you were too lazy to Google, the _least_ you can do is press the ten keys it takes to spell out "thank you"
20:47 < meyou_> my cube neighbor is an arab dude raised in france, i'm teaching him U.S. slang on a daily basis
20:47 < mawk> RustyJ: y e a h
20:47  * stealintv agrees with Dagmar
20:47 < RustyJ> meyou_, whats to teach? the formal use of the F word?
20:47 < meyou_> bounce, plox, imo, etc
20:47 < meyou_> those are the most recent ones
20:48 < RustyJ> plox?
20:48 < meyou_> that's what he said!
20:48 < ldlework> my policy is don't try to regulate other people's speech
20:49 < RustyJ> ldlework, should you be using contractions?
20:49 < ldlework> ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn
20:49 < RustyJ> blah i'm back to idle.... i had my fill of fun
20:49 < ThePendulum> I can see an argument to prevent hatespeech, but what that entails shouldn't start slipping like it does
20:49 < jim> I know some of you are having fun, and I tend to let that happen (but limit the time), and, please drop the admin stuff
20:50 < Dagmar> They probably have both cable TV and multiple forms of terrestrial radio if they want to learn to be stuid
20:50 < Dagmar> s/stuid/stupid/
20:50 < RustyJ> jim, admin stuff?
20:51 < jim> yeah, hate speech is strictly out (bannable)
20:51 < jim> please let the ops do the opping :)
20:51 < RustyJ> ahhhh
20:51 < meyou_> we need to round up all the Mint users and put them in camps
20:51  * meyou_ runs
20:51 < RustyJ> mint hate speech ^^^
20:51 < capcj> haha
20:51 < searedvandal> oh no
20:52 < jim> meyou_, :P
20:53 < RustyJ> i think meyou_ is distro-angry
20:54 < meyou_> i'm just kidding i have some good friends who are Mint users
20:55 < RustyJ> ok, back to lurk mode for real
20:55  * EvilRoey digs RustyJ with a programmer's spoon
20:56 < EvilRoey> *digs out
20:56 < ldlework> I have been using NixOS for a couple months now. I am falling in love with Linux all over again.
20:58 < WhiteDevil> okay guys i got a new wireless adpater the panda express pau09
20:58 < WhiteDevil> i need to set it up on my linux
20:59 < WhiteDevil> i think i dont need drivers for this one
20:59 < WhiteDevil> i am on debian
20:59 < WhiteDevil> lookignt hrought the control panel for hints
21:01 < mango_99> hi, is there any keyboard shortcuts for grub? i want to be able to hold a key when i restart to boot into linux, and hold a key for windows for quick reboots
21:01 < mango_99> without having to scroll through grub screen
21:03 < lnnb> Equinsu ocha!
21:04 < jim> what's that thing?
21:04 < searedvandal> mango_99, best bet would probably to set grub to hide the menu unless you hold down shift, set windows to default and hold down shift whenever you want to boot into linux
21:04 < mango_99> searedvandal, neat trick thanks
21:05 < searedvandal> mango_99, https://wiki.archlinux.org/index.php/GRUB/Tips_and_tricks#Hide_GRUB_unless_the_Shift_key_is_held_down
21:05 < searedvandal> that explains how it's set up
21:09 < plexigras> does look allow using unsorted word lists?
21:10 < ayjay_t> anyone know a lightweight markdown viewer? like a plugin for feh or zathura?
21:16 < WhiteDevil> when iuse make to compile the drviers
21:16 < WhiteDevil> it says command not found
21:16 < WhiteDevil> i am using debian
21:16 < WhiteDevil> whats the substitute to make in debian ?
21:17 < adrian_1908> WhiteDevil: make should be found, what does `gcc --version` say?
21:18 < WhiteDevil> i just read i should reinstall the build-essenstials package
21:18 < adrian_1908> WhiteDevil: yes, that would have been my first suggestion, depending on the output. do that!
21:19 < WhiteDevil> gcc (Debian 6.3.0-18+deb9u1) 6.3.0 20170516
21:20 < rascul> ayjay_t python-markdown includes a cli tool (called markdown_py here) which will convert markdown to html, also there is pandoc which has a variety of input and output formats
21:20 < adrian_1908> try reinstalling the package and see if that fixes it. surprises me though, would not expect make to be unavailable in the first place.
21:20 < lnnb> ayjay_t: http://lmgtfy.com/?q=markdown+editor
21:20 < adrian_1908> @WhiteDevil:
21:20 < bratchley> make isn't part of gcc afaik
21:21 < bratchley> it's part of automake I think
21:21 < rascul> make is part of make
21:21 < adrian_1908> bratchley: but shouldn't make be part of just about any vanilla linux distro?
21:21 < bratchley> not necessarily
21:21 < WhiteDevil> ya it was that
21:21 < rascul> make isn't always installed by default but is generally made available in repos
21:21 < bratchley> why would it be?
21:21 < WhiteDevil> but for some reason this drive dint compile properly
21:21 < WhiteDevil> :(
21:22 < WhiteDevil> i just got a new wireless adpter and trying to make it run on my linux
21:22 < WhiteDevil> debian and redhat
21:22 < adrian_1908> WhiteDevil: btw, check out the `checkinstall` package, really nice if you want to keep your compiled binaries for later reuse, and for easy removal.
21:23 < WhiteDevil> ill look into it
21:24 < dfcnvt> quick question -- in terminal (bash) how do you make a quick math?  Say, 2^5
21:25 < pankaj> How to deal with 'ls' when I am using only tty and I have large number of files (like movies with different names that I want to choose) and I want to choose one. The list is very big and that too in one line.
21:25 < pankaj> Using 'less' also makes it big
21:27 < koala_man> pankaj: ls -C | less if you want them in columns in less
21:31 < ayjay_t> ugh but i hate latex
21:32 < ayjay_t> i do everything i can to avoid latex
21:32 < koala_man> polyurethane 4 lyfe
21:33 < dfcnvt> latex, that's the math's annotation?  Or just special editor with all included characters for any notation purpose (science, math, etc) ??
21:33 < azarus> dfcnvt: LaTeX is a typesetting system
21:34 < ayjay_t> it's a bajillion packages fora  typesetting system which doesn't give one heck about the semantics
21:34 < ayjay_t> i'll admit tho, the default font is realllll elegant
21:35 < azarus> ayjay_t: well do you prefer troff or something?
21:35 < ayjay_t> html
21:35 < ayjay_t> no joke.
21:35 < ayjay_t> i do everything i can in html and markdown.
21:36 < azarus> fair enough
21:36 < azarus> I'm a bit of a retro guy in that regard
21:36 < azarus> troff -me
21:36 < koala_man> can you typeset math in html yet?
21:38 < ayjay_t> okay so math type setting is the one thing i like latex for
21:40 < azarus> i just adore the typesetting of "The C Programming Language" by bwk and dmr
21:40 < azarus> so I really want to learn to use troff proficientlöy
21:40 < azarus> so I really want to learn to use troff proficiently*
21:41 < adrian_1908> dfcnvt: echo $((2 ^ 5)) but you probably want 2 ** 5 in there.
21:41 < rypervenche> dfcnvt: bc also works; bc <<<"2^5"
21:41 < adrian_1908> dfcnvt: or consider opening an interactive python session, if you have that installed.
21:42 < adrian_1908> bc looks like a better solution, wasn't aware of that.
21:52 < Celmor> I hate it when I accidantly unplug or the monitors connector when moving it, now linux doesn't pick up the monitor anymore and I have to reboot apparently...
21:52 < Celmor> unplugging and plugging back in both connectors of the cable doesn't help, reloading DE doesn't and xrandr thinks the monitor is still there
21:53 < twainwek> i'd probably blame it on your distro... i don't have that issue
21:54 < twainwek> if xrander detects them, you can try changing it's location and see if that works
21:54 < toothe> anyone know how to make a command execute every time you run a new command
21:55 < toothe> ?
21:55 < toothe> i am trying to create my own git prompt bash thingy
21:55 < toothe> the ones I found online are quite slow.
21:55 < plexigras> how can i do `grep "^$word"` and have $word not be interpreted as regex?
21:55 < koala_man> toothe: like PROMPT_COMMAND?
21:56 < Celmor> twainwek, I blame it on the monitor
21:56 < koala_man> plexigras: I don't think you can. you can use awk or bash instead
21:56 < plexigras> its not posible is it?
21:57 < tty01> `grep "\^\$word"` i think should work
21:57 < plexigras> can i use look with an unsorted list?
21:58 < koala_man> tty01: I assumed $word would be something like a.b where the . shouldn't be regex
21:58 < plexigras> tty01: i want to grep words that start with the content of $word
21:58 < tty01> ooh
21:59 < twainwek> can you not escape $word before you pass it into grep
21:59 < plexigras> how would i do that?
21:59 < xyz111> Hi Guys, I'm a regular linux user, and wanted some advice on which new laptop to buy. My current laptop is AMD with on-board graphics, and the graphics performance has not been too great. Can anyone suggest a laptop that they've had a bit of success with? Thanks in advance!
22:00 < twainwek> dunno... some function that escapes all regex alphabet(?)
22:02 < bls> xyz111: you're going to need to provide some more criteria other than "had some success with"
22:02 < plexigras> from what i have heard people generaly have good sucess with dell and thinkpads
22:03 < bls> my personal anecdotes are thinkpads are good, dells are awful
22:03 < twainwek> i'd stay away from anything with an optimus card
22:04 < twainwek> unless possibly when you can turn it off from the bios
22:04 < bls> but that includes and excludes so many different models it's not really worth much
22:11 < MrElendig> I hope the new ryzen a1000 serie doesn't suck too much
22:11 < MrElendig> and that some laptop makers will pick it up soon
22:11 < tty01> plexigras: i think this is what you're trying to do? https://pastebin.com/YwV6A2Ch or am i still off?
22:11 < MrElendig> v1000*
22:12 < MrElendig> V1605B at 12w could make for an interesting laptop
22:12 < widp> Is there a way I can make different "startup profiles"?
22:12 < MrElendig> widp: sure, systemd targets
22:13 < MrElendig> create a target wiht the stuff you want, boot with system.unit=yourthing.target
22:13 < widp> MrElendig: can I create profiles that I can choose on login?
22:13 < widp> ahh, right, the grub menu, correct?
22:14 < widp> thank you MrElendig
22:15 < MrElendig> widp: depending on what you  mean with "startup profile" ofcourse
22:15 < widp> right, so currently I have a bunch of programs running during startup.
22:16 < MrElendig> system or user session?
22:16 < nmschulte> what tools exist to merge stdio streams? specifically, I want to take would-be-tty output (things w/ ANSI escape sequences for color and cursor movement a la progress bars, spinners) from multiple processes and feed them "line-by-line" in a single display.
22:16 < widp> I want to create 2 profiles, one with these programs running at startup and another without them
22:16 < MrElendig> nmschulte: that is going to be a mess
22:16 < widp> MrElendig: I don't understand.
22:17 < nmschulte> I've got a sort of hacky solution working in JavaScript, but it suffers from "ANSI escape artifacting" at points.
22:17 < bls> nmschulte: you could use some variations of cat or paste, but it's going to be awful
22:17 < MrElendig> nmschulte: shouldn't be too hard to write using ncurses though
22:17 < bls> heh, javascript and what I assume are hard coded escape sequences? what could go wrong?
22:17 < nmschulte> MrElendig: I hope to make this cross platform, Linux, Mac OS X, Windows 10.
22:17 < nmschulte> bls: no escape sequences at all; the processes output those.
22:18 < nmschulte> there's enough new-lines that it "works alright"; it's just feedback to the user "things are working right", but would be nice if it could be trusted to be unartifacted/accurate
22:18 < lnnb> what do you mean feed them line by line
22:19 < MrElendig> nmschulte: that is really hard, since windows is horrible
22:19 < nmschulte> MrElendig: Indeed.  The JS solution works well enough for now.
22:19 < lnnb> tty doesn't automatically mean it's always line buffered output
22:19 < bls> and you have no control over what the individual applications are assuming about the current state of the espace/color modifiers are
22:19 < nmschulte> lnnb: I want to get better at what I mean here -- can you help?!
22:20 < lnnb> maybe, if you can explain why tmux/screen can not do what you are imagining
22:22 < toothe> Am I missing something here that everyone is seeing except me?
22:22 < toothe> if [ $retval -ne 0 ]; then return; fi
22:22 < toothe> -bash: [: -ne: unary operator expected
22:22 < bls> toothe: $retval is null
22:23 < bls> if [ "${retval:-1}" -ne 0 ];then return;fi
22:23 < toothe> blah!
22:23 < bls> or whatever is the failsafe default
22:23 < lnnb> if [ "$retval" != "0" ]  ?
22:23 < nmschulte> MrElendig, lnnb: this is the JS script that kicks off processes w/ ANSI sequences: https://paste.debian.net/1030865/
22:24 < toothe> how do I set retval?
22:24 < toothe> wait...i justwrote a 1000 line bash script, how do I not know this...
22:24 < toothe> 1300 line bash script*
22:24 < nmschulte> you can see on line 57 how I parse/relay the streams "line by line"
22:24 < tty01> might wanna add a -x to that script so you can debug all those lines
22:25 < lnnb> nmschulte: i don't have time to learn JS today
22:25 < nmschulte> lnnb: no worries.
22:25 < toothe> bls: thanks
22:25 < nmschulte> lnnb: I collect output from the process until new line characters, then flush that collection/fragment.
22:26 < nmschulte> it works pretty well, because the output is mostly line-based, but some steps in the processes do things like ASCII progress bars or "wait" indicators, and sometimes when there's color/style, it "bleeds".
22:27 < nmschulte> I didn't want to maintain a script, was hoping a tool existed already.
22:27 < Trel> Is there a good article write/up on the differences between zsh and bash, not from a which is better standpoint but more from what one might run into while using the one they're not familiar with (so in both directions)
22:28 < bls> if an application is using terminal control codes to move the cursor, that application was never intended to be used in a pipe
22:29 < bls> the usual approach to this issue is to strip all the control, escape, and movement sequences
22:29 < nmschulte> bls: or inform the process to do that for you (processes here are Docker commands, for reference).
22:30 < bls> so you're using custom JS to automate docker via the CLI? shouldn't this be using a CM system or kubernetes
22:30 < nmschulte> bls: I really just wanted the colored/bolded/underlined... output from the processes, not the fancy progress bars etc.  I guess tr and other progs can strip everything but the color?
22:31 < bls> no, the control sequences are all multiple characters, so you either need to tell the program to not output, trick it via a dumb TERM setting, or use something that's aware of how to parse the control sequences for stripping them
22:31 < lnnb> oh you want to eliminate color code bleed through from other programs?
22:32 < nmschulte> bls: sure! this is for development, not production; asking devs to run Kubernetes locally is silly.
22:32 < nmschulte> lnnb: that at least, yeah.  I'm sure there are other issues w/ cursor movement I'm not noticing, but that's the holy grail.  fixing color would be a great start.
22:33 < toothe> bls: in the example you gave earlier, what was -1 ?
22:33 < lnnb> just make sure to always terminate the colored block, sometimes programs don't bother like grep and you get a huge chunk of colored text
22:33 < nmschulte> I think simply adding a reset ([0m; ?) after every line of output would go far.
22:33 < toothe> "${retval:-1}
22:33 < toothe> I didn't understand that part.
22:34 < bls> toothe: :- is a sigil for substituting something if the variable is unset, I randomly picked 1 as a guess at what a safe default would be
22:35 < bls> that's not going to handle output like: =>\e[31m;1%\r==>\e[33m;\r===>\e[32m;
22:36 < nmschulte> I'd have to re-start the color block across process fragment/line outputs.  bleh!
22:37 < nmschulte> bai bai ANSI  :) -- ty all.
22:37 < lnnb> for cursor you have to save/restore position too
22:37 < bls> terminal color is pretty dumb/useless for most things anyway, you're better off stripping it
22:37 < nmschulte> so that was my original Q: is there a program that does this already?!
22:37 < lnnb> so you need program context variables for everything merging into the pipe
22:38 < lnnb> usually you just save output form one program to it's own file
22:38 < lnnb> what beenfit does merging them have?
22:40 < bls> I usually turn terminal color off via commands that support the switch, or run some things with no switch with TERM=dumb or something similar to try to get them to stop
23:04 < nmschulte> lnnb: these processes run persistently, they're "auto-builds" of various components via inotify and polling; this output/merging is just a status indicator -- the specific information is only important if things go awry.
23:05 < nmschulte> the script lets the dev run a single command (single terminal/console) and have a single log of the output/status.
23:06 < bls> you could also do something as simple as have the commands output to a file (if they're capable of producing suitable output), then tail all the output files
23:11 < za1b1tsu> What does "Run command as login shell" even means?
23:12 < bls> za1b1tsu: there are certain things you want to happen the very first time you log in to a system that you don't want to happen if you open new shells/programs
23:13 < za1b1tsu> *scratches head*
23:13 < bls> za1b1tsu: so shells could be told they were the login shell, in which they'd process certain dotfile, or that they were just another interactive shell, that'd process a different set of dot files
23:14 < raindev> za1b1tsu, INVOCATION section of bash man page is quite extensive if you're interested in details.
23:14 < justsomeguy> za1b1tsu:   ...for example, you might want to set an initial value for environment variables that tell programs what locale you're using, where to load certian programs from, and what you standard editor.
23:15 < bls> or you want to have a certain program running while your logged in, but only a single copy
23:16 < justsomeguy> za1b1tsu: If you put that in the configuration file for a login shell, it will only be run once. If it's in ~/.bashrc, or something for a interactive non-login shell session, it will be run every time you open a terminal window.
23:16 < raindev> showing package updates is a good example too (Ubuntu server does it): you only want it displayed once after logging in (e.g. via ssh), not for every new shell.
23:16 < bls> displaying the motd or the server issues or fortune were other more classic examples
23:17 < za1b1tsu> I see
23:17 < za1b1tsu> rvm requires it
23:17 < bls> they've likely misunderstood and/or are making some pretty shakey assumptions
23:18 < nothos> So, I'm about to go mad. Client wants PHP 5.3 w/ TLS 1.2 support and classic ASP running on the same site/server.
23:19 < nothos> Which means fudging together a hacky thing with docker on windows and custom built PHP in a container :(
23:19 < compdoc> thats madness!
23:19 < bls> why does it mean that?
23:20 < nothos> bls Because building PHP 5.3 with modern TLS support on windows is a damn nightmare
23:20 < za1b1tsu> / #macos
23:20 < za1b1tsu> sorry
23:20 < bls> why does it have to be done on windows?
23:20 < nothos> Because of the classic asp
23:21 < bls> let windows host the ASP, have something more sane host PHP
23:21 < nothos> bls Client wants one box and one box only
23:21 < bls> hehe, fire your client :P
23:21 < nothos> They're paying us far too much
23:21 < nothos> Boss wouldn't allow it :D
23:23 < Dagmar> Warn them that this will cost them more than usual because of their inadvisable requirements
23:24 < nothos> Dagmar They've already accepted that
23:24 < nothos> They're the sort of "Throw money at the problem" clients
23:24 < nothos> (You can see why my boss is keen on them)
23:25 < Dagmar> In that case they can have whatever they're willing to pay for
23:26 < MrElendig> sounds like a bad idea compared to just get some extra hardware
23:26 < nothos> It's...weird
23:26 < nothos> Just, yeah
23:26 < Pentode> seriously. bunch of masochists
23:27 < hailhydra> how could i see what port an email server is using?
23:27 < nothos> Pentode Sado-Masochists
23:27 < bls> hailhydra: ss
23:27 < MrElendig> nmap
23:27 < hailhydra> ss?
23:27 < nothos> Since it's my sanity that's being expended to fudge this ;)
23:27 < MrElendig> if you have access to it, just read the config
23:27 < bls> hailhydra: yes, the ss command
23:27 < nothos> hailhydra ss -tulpn
23:27 < hailhydra> ss someone@somedomain.com
23:27 < nothos> Oh wait, is this not local hailhydra ?
23:28 < nothos> Do you want to find what port a different server is running its email on?
23:28 < bls> if it's not local, then the mail server is on a standard port
23:28 < BerenErchamion> Are there any good driving simulators available for the Linux OS?
23:28 < bls> BerenErchamion: tuxracer?
23:28 < bls> heh
23:28 < hailhydra> yeah cause the client is a moron and doesn't know what port connect to
23:28 < hailhydra> nothos: yes
23:29 < nothos> hailhydra Just to confirm, not on AWS is it?
23:29 < nothos> They have some strict no-no rules on port sniffing before asking their permission :D
23:29 < hailhydra> no I think its a godaddy site
23:29 < hailhydra> oh no its Dreamhost
23:29 < BerenErchamion> I'm looking for a realistic driving simulator to train driving students.  I figure I have to use Windows, but just thought I would ask here, first
23:29 < MrElendig> man ss, it has examples
23:29 < lnnb> BerenErchamion: check steam they might have something
23:29 < MrElendig> BerenErchamion: there are commercial software for that
23:30 < nothos> BerenErchamion Your best bet would be trying a windows one I imagine
23:30 < MrElendig> specificly made for that task
23:30 < MrElendig> you can even map in your own town in them
23:30 < nothos> Not sure there's any hyper-realistic driving sims like that on linux
23:30 < bls> it's not port scanning if you connect to the two standard SMTP ports
23:30 < MrElendig> nothos: physics doesn't have to be that realistic, it is 99% about traffic
23:30 < nothos> bls I'd assume they're not is the issue :D
23:30 < BerenErchamion> Yeah: I figure Windows has the best driving simulators
23:31 < BerenErchamion> but they'll be expensive compared to Linux simulators
23:31 < MrElendig> idealy it would tell you when you screwed up
23:31 < lnnb> simulators aren't cheap
23:31 < bls> in which case nothing else that hasn't been custom configured is going to be able to communicate with them, and they should know if they've done that
23:32 < MrElendig> lnnb: the driving school ones are pretty reasonable
23:32 < lnnb> how much
23:32 < MrElendig> varies *a lot* depending on the features you want and the country
23:32 < MrElendig> and how big of a customer you are
23:32 < lnnb> how cheap do they get
23:32 < MrElendig> anything from a couple of 100 bucks a year to a few thousand per seat
23:33 < MrElendig> which is nothing for a driving school, unless you are in india
23:33 < bls> left hand driving please!
23:33 < MrElendig> in which case you don't need one anyway
23:33 < lnnb> you don't need one to begin with
23:33 < akk> Train 'em on tuxkart. :)
23:33 < bls> "here's the gas pedal, here's the horn. have fun!"
23:33 < MrElendig> https://www.youtube.com/watch?v=xGfLNqjh4j0
23:33 < MrElendig> bls: almost
23:34 < lnnb> that could be 1-12 months rent out the window just because you wanted a nice simulator
23:34 < MrElendig> bls: it is slighly harder now, you have to drive in a circle now
23:34 < toothe> bls: your solution from earlier worked, but I have actual working examples where I've done it differently...yet its not working now.
23:35 < toothe> https://gitlab.com/farhankhan/cis/blob/master/cis_linux_script.sh#L940
23:35 < justsomeguy> This is a little strange, but, is there such a thing as a linux tutor?
23:35 < bls> toothe: if the script is littered with other uninitialized variables, that wouldn't be a surprise
23:35 < toothe> bls: pardon?
23:36 < bls> toothe: the snippet I gave was a way to deal with uninitialized variables. it's a problem that tends to cascade down through a script
23:36 < bls> if not handled via best practices at every step
23:37 < toothe> bls: I see...so, I explicitly set the variable eralier.
23:37 < toothe> which I find odd.
23:37 < toothe> as in, I did : RETVAL=`some-command`
23:37 < MrElendig> a good example of a script that shouldn't be a shellscript
23:37 < bls> toothe: that's capturing the output of some-command, not its return value
23:37 < toothe> i see...
23:37 < toothe> that would be a problem.
23:37 < toothe> a big problem.
23:38 < bls> toothe: might want to try out shellcheck.net on it, fix what it tell you, then go from there
23:38 < bls> some_command_output=$(some-command); retval="$?"
23:39 < toothe> bls: that is...amazing...
23:39 < toothe> that website.
23:46 < akk> j
23:49 < toothe> bls: this code you rpovided is giving me a return code of 1
23:49 < toothe> even if/when I manually run the command myself and get 0
23:50 < phogg> toothe: What is the command? What parameters are you using?
23:50 < phogg> you get help faster if you can provide an exact sample
23:50 < toothe> true. a moment.
23:51 < toothe> phogg: https://dpaste.de/dTHh
23:53 < bls> toothe: you can't simultaneously capture a commands output and evaluate its return code
23:53 < toothe> Right. I was getting frustrated.
23:55 < bls> and even in languages that allow it, it's considered a bad practice/anti-pattern
23:56 < compdoc> is wordpress still a security nightmare?
23:57 < meyou> keep it patched up and it's not so bad
23:58 < compdoc> is there a small distro with wordpress that runs in a vm?
23:58 < meyou> just has a much higher ROI for someone hunting exploits since a good exploit in the current wp build gets you into a loooot of boxen
23:58 < xamithan> You can run it inside a docker container
23:58 < xamithan> or lxc
23:59 < xamithan> wordpress was never a security nightmare,  where did you hear that?  Maybe you mean outdated wordpress plugins
23:59 < dunnousername> Hey, I'm trying to cross compile a kernel, but whenever I set up my config the way I want it, I `make ARCH=arm CROSS_COMPILE=...`, and then it reconfigures everything. I can't seem to get it to actually build the kernel, just stuck configuring it
--- Log closed Wed Jun 27 00:00:15 2018