--- Log opened Sat Jun 30 00:00:08 2018
--- Day changed Sat Jun 30 2018
00:00 < JordiGH> Why is it that postscript seems to be generic default when I'm trying to add a new printer from the Gnome settings?
00:00 < Hasimir> JordiGH, because it's been the de facto standard for decades
00:00 < phogg> JordiGH: It's the traditional printing format, or rather the traditional one after ASCII, especially on Unix systems where things like the low-margin desktop-printing printer business weren't really factors.
00:00 < phogg> tl;dr historical reasons
00:00 < bls> because it's about as close to a standardized PCL (or PDL if you will) you'll get
00:01 < phogg> Free software also tends to prefer open standards, or at least some kind of standards, for all purposes.
00:02 < bls> yeah, everywhere I've gone to school or worked had HP laserwriter/laserwhatever printers for the unix systems that all understood PS
00:02 < runjutsu> omg.. first of all: I asked in the wrong channel: Secondly: why am I wasting your time on trivial googleable things.
00:02 < JordiGH> Wait, I'm confused.
00:02 < runjutsu> thirdly: I'm way past bedtime
00:02 < JordiGH> Can printers today understand postscript or not?
00:02 < Hasimir> phogg, yes, but the preferences of the FSF and friends didn't have much weight back when PostScript launched
00:02 < phogg> postscript was a key component of the start of the "desktop publishing" revolution in the mid 80s. Apple went to Adobe and gold was the result.
00:02 < Hasimir> JordiGH, yes; most, if not all
00:02 < phogg> Hasimir: but it matters a lot to GNOME
00:02 < Hasimir> true
00:03 < bls> it wasn't until the advent of cheap soho printers that dispensed with the memory and processing power needed to interpret postscript that we got into the printer compatibility issues
00:03 < phogg> yep
00:03 < lukey_> sensibel: Sorry I'm using debian so i can't help you there, try asking in #centos
00:03 < phogg> it's such a shame. You could probably have supported .eps natively without much more expense and then did the compiling on the computer
00:04 < JordiGH> Wait, so postscript was standard because printers were huge monsters, then printers got cheap and couldn't process postscript anymore, and now printers can afford to play videogames and send email again so therefore they can run Postscript again?
00:04 < phogg> runjutsu: that is by far not the most off-topic question we've gotten in here. And at the right time of day you might have even received a prompt answer.
00:05 < phogg> JordiGH: yes, yes, and yes
00:05  * Hasimir ponders the PDF ISO standards again
00:05 < JordiGH> phogg: Thanks, I see!
00:05 < Hasimir> nah, I can understand wanting to avoid that mess
00:05 < phogg> Hasimir: at least it doesn't embed an actual programming language
00:05 < phogg> unless they tried to standardize the JS parts
00:05 < Hasimir> phogg, no, it's worse
00:06 < Hasimir> have you looked under the hood of that evil beast?
00:06 < phogg> of PDF? I have been studiously avoiding it.
00:06 < bls> no, they basically tried to pass a standard that says, parts of this are defined by whatever adobe has implemented
00:06 < Hasimir> you're lucky
00:06 < boingolov> lol, printing
00:06 < JordiGH> lol
00:06 < boingolov> who does that anymore
00:06 < bls> namely the JS and XML form stuff
00:06 < phogg> postscript I kind of get, never wanted to know what makes pdf tick
00:07 < phogg> bls: well that's useless. Might as well just use OS Office Open XML.
00:07 < phogg> s/OS/MS/
00:07 < Hasimir> phogg, I had to; a job a while back and understanding it was the only way to complete the damned thing
00:07 < Hasimir> it did pay well, though
00:07 < bls> boingolov: I have to maybe 2-3 times a year, feel like such a dinosaur
00:07 < Hasimir> that spec is ... urgh
00:07 < curiosikey> I'm way over my head, but what would cause a box to start without networking? We had to restart the service to bring it back online after a reboot. We are in run level three right now and the default target is multi-user
00:07 < bls> phogg: I feel like that's where they learned the trick
00:07 < phogg> Hasimir: I've been there for other formats. Fortunately for PDF I managed to get away with "this is all the library I found supports"
00:07 < boingolov> bls: I was halfway kidding
00:07 < Hasimir> fortunately the NSA came to my rescue (I shit thee not)
00:07 < boingolov> that said, I print less and less
00:08 < phogg> Hasimir: "You can't use PDF because it can embed information we intended to redact."?
00:08 < twainwek> boingolov: my boss insists on emailing me a document, then printing it, and then physically bringing it to my office
00:08 < boingolov> once upon a time it was my job to configure print queues in *nix, that sucked
00:08 < boingolov> well, I mean, it was fine I guess, but it was a chore
00:08 < phogg> if managing print queues sucks you're using the wrong software
00:09 < boingolov> phogg: the queues themselves were fine, it was the things on either side of it that sucked ;)
00:09 < Hasimir> phogg, no, much better than that ... I found something they took offline shortly after I found it ... but I kept a copy and it's here: http://okbounty.adversary.org/pdf_risks.pdf
00:09 < boingolov> i.e., users and printers ;)
00:09 < phogg> twainwek: The part I love is when there's a fillable PDF form and someone prints it, fills it, scans it and then sends it to me as an email attachment... and I have to print and file the result.
00:09 < boingolov> some of the printers were serial
00:09 < bls> I loved the paid print queue software my uni used...only deducted your charges after the last page was ejected, so everybody would just add a blank page at the end, get a friend to stand by the printer, and lprm as the last page was spit out
00:09 < boingolov> that was an additional nightmare
00:10 < phogg> Hasimir: ah, so you used their paranoia about hidden embedded information to argue that PDF was a bad idea?
00:10 < Hasimir> the rest of that site/subdomain relates to the job I delved into it for
00:10 < Hasimir> no it was a digital forensics job
00:11 < Hasimir> it involved Bitcoin personalities, so it was a bit of a train wreck
00:11 < phogg> Hasimir: I remember seeing some news reports like "Redacted PDF vulnerable to "select all" attack"
00:11 < Hasimir> this was more stupid than that
00:11 < bls> heh, where they just positioned a black bar in a layer over the text
00:12 < phogg> bls: yep.
00:12 < phogg> classic case of not knowing how it works being a really bad thing
00:12 < phogg> warms my geek heart
00:12 < Hasimir> I'll grant that's pretty stupid too
00:13 < Hasimir> anyway, that nsa doc is to pdf what o'reilly's sql in a nutshell is to every sql server
00:14 < phogg> Hasimir: I'm getting that impression from skimming it. It's a nice reference.
00:15 < Hasimir> yeah, it's nice to keep for some future need
00:15 < phogg> Hasimir: I find it funny that a document talking about the vulnerabilities in PDF is published as a PDF.
00:15 < Hasimir> yeah, that too
00:16 < Hasimir> phogg, the real chuckle-fest is in the emails archived on that site too
00:17 < phogg> I spent a few horrible weeks a few years ago reverse engineering how MS Excel conditional formatting rule serialization works so I could generate complicated spreadsheets. I had read about their XML formats being bad, but it took that to teach me exactly how bad. It's literally no better than BIFF rendered as plain text.
00:17 < phogg> I've seen worse abuses of XML, but not by much.
00:18 < Hasimir> wow
00:18 < Hasimir> I've looked at docx, but not any of the rest
00:18 < bls> I recently had to deal with apple's plists, and wow is that infuriating
00:19 < goldstar> anyone know the netstat options to run if wanting to see ICMP stats ?
00:19 < goldstar> netstat -s gives them all
00:19 < Hasimir> bls, did you find the decode and re-encode command?
00:19 < phogg> Things like... go into file1.xml, find a tag with an attribute xref="id" where id is a number. That number is an index into the tags named <foo> under a tag named <bar> in another XML file. You just have to know the relation.
00:20 < Hasimir> that's retarded
00:20 < bls> Hasimir: found a python library to turn it into an actual hierarchical data structure
00:20 < Hasimir> oh my
00:20 < phogg> they do that constantly and with multiple different kinds of refs. They all have to line up "just so" and they can *sometimes* be omitted, and when you can omit you *must* or Excel chokes.
00:20 < Hasimir> jesus
00:20 < Hasimir> bls, which lib?
00:21 < phogg> with helpful errors like "corrupt document at offset 376821"
00:21 < bls> https://docs.python.org/3/library/plistlib.html
00:21 < Hasimir> phogg, just like when someone screws up a pdf mod ...
00:21 < Hasimir> oh, a std lib, cool
00:21 < phogg> Hasimir: I wouldn't know, but probably.
00:23 < phogg> docx is probably worse, but only because the choice of which tags go around which bits of text is purely based on the MS Word implementation and not anything logical
00:23 < bls> started out trying to use xml.etree and XPath, until I realized they'd flattened certain portions of the structure
00:23 < Hasimir> bls, urgh, no ...
00:23 < Hasimir> I only ever modified those things by hand
00:23 < phogg> fortunately you can mostly write HTML with the right sugar, name it .docx and word will open it. As long as your documents aren't too complex.
00:24 < bls> can docx be validated with a DTD?
00:24 < Hasimir> and not often
00:24 < phogg> bls: Technically? Yes. You can't know what most of it means just from the spec, though.
00:24 < Hasimir> er ... not sure
00:24 < twainwek> and yet for some reason people insist on emailing an image or plaintext as a pptx or docx file
00:24 < phogg> .docx isn't one format, it's several different XML formats in a zip file
00:25 < Hasimir> yeah
00:25 < phogg> depending on which features you use
00:25 < phogg> Excel is the same. The rules for styling change *completely* if you have ANY "table styles" in a sheet
00:25 < Hasimir> I normally ignore everything except the file with the content
00:25 < phogg> Hasimir: if you're just trying to extract data you can do that. Sadly my job seems to always involve "can you export that to Word?"
00:26 < twainwek> jesus
00:26 < bls> ah, sort of like some of the ebook formats where instead of figuring out how to encode graphics they just zip up html and images
00:26 < Hasimir> when doing that it's a little less ugly than ODT/ODF too
00:26 < phogg> bls: yes
00:26 < phogg> twainwek: yeah, "Send me a screenshot" to most people seems to mean "Send me a .pptx with an image pasted in to it."
00:27 < phogg> not sure why
00:27 < Hasimir> bls, we don't want encoded ebooks ... companies always interpret that as a good enough reason to push drm
00:27 < sensibel> Still the same error on opening bash bash: ulimit: open files: cannot modify limit: Operation not permitted although tried this
00:27 < Hasimir> case in point: Mobipocket
00:27 < sensibel> https://unix.stackexchange.com/questions/359189/fixing-ulimit-open-files-cannot-modify-limit-operation-not-permitted
00:27 < phogg> sensibel: what is your file limit set to?
00:28 < bls> thankfully my current shop is primarily a macos ecosystem for desktops, so little to no Office files getting passed around
00:28 < sensibel> [dellemc@localhost bilal]$ ulimit -n 100000
00:29 < bls> Hasimir: yep, and I've quit buying from amazon/kindle store now that there's no drm stripping
00:29 < phogg> sensibel: any limits set in /etc/security/limits.conf (or under limits.d)?
00:30 < sensibel> phogg: dellemc hard nofile 100000 dellemc soft nofile 100000
00:31 < Bunk> hello
00:31 < phogg> sensibel: are you familiar with the information as given here: https://www.tecmint.com/increase-set-open-file-limits-in-linux/
00:32 < sensibel> phogg: [dellemc@localhost bilal]$ cat /proc/sys/fs/file-max 500000
00:32 < phogg> sensibel: if your distro uses systemd note that it can also set limits in its config file which affect services it spawns and anything they run in turn
00:32 < phogg> sensibel: those limits all seem relatively high, so the chance of hitting them is not great
00:32 < Hasimir> bls, no drm stripping?  Surely it's still crap and easily removed?
00:32 < phogg> Bunk: hi
00:32 < sensibel> phogg: then how to set it to 100000?
00:33 < phogg> sensibel: set it for what kind of scope?
00:33 < Hasimir> mind you, I never buy kindle devices or ebooks
00:33 < bls> they've switched to a new scheme in the past couple years that hasn't been resovled yet
00:33 < sensibel> phogg: actually I am performing genome analysis that creates a lot of files
00:34 < Hasimir> so that's what they chose to do instead of embracing utf-8 ... fucking typical
00:34 < phogg> sensibel: I don't care. Are you wanting to set the global limit? Looks like the kernel's is already five times that value.
00:34 < bls> I buy DRM free when at all possible, but some of the things I need to read for work aren't available, so I'll pay and strip them for backup purposes
00:34 < Hasimir> bls, yeah, I do that
00:35 < phogg> wait, amazon doesn't use utf-8?
00:35 < Hasimir> and sometimes drm is added by a distributor, so I always give explicit permission to rip it off any of my publications
00:35 < sensibel> phogg: yes its five times high but then why it shows error upon opening terminal?
00:35 < Hasimir> phogg, not in mobipocket, it's windows-1252
00:35 < phogg> sensibel: What is the limit for that process at that time?
00:35 < phogg> Hasimir: FFS, really?
00:36 < phogg> This is 2018 not 1988.
00:36 < Hasimir> I'm afraid so
00:36 < phogg> Fun story, do you know what an Oracle database chooses as the default encoding for CHAR and VARCHAR columns when you install it?
00:37 < micrex22> phogg do you know what ORACLE stands for?
00:37 < phogg> If you said "CP-1252" you'd be a winner.
00:37 < Hasimir> phogg, bls, enjoy my Unicode cheat sheet ... and the anti-mobipocket rants contained therein: http://files.east1.us.adversary.org/files/UnicodeNotes.pdf
00:37 < phogg> The justify it in their docs as: "Most clients are Windows"
00:37 < Hasimir> also the anti-LaTeX rants
00:37 < sensibel> phogg: the process showed error once therefore I changed this limit but dont know the  limit for the process at the that time , but the same process is running currently is there a way to know the limit for a current process?.
00:37 < phogg> hey now, don't be down on my LaTeX!
00:38 < Hasimir> phogg, there's a 3 page take down in that cheat sheet
00:38 < phogg> sensibel: run ulimit -n from the shell that spawned the process to see its limit
00:38 < boingolov> phogg: oracle defaults are designed to ensure job security for oracle dbas
00:38 < bls> hehe, I'm still love-hate with LaTeX. if I don't need something to look nice, I've switched to groff with mom macros
00:38 < boingolov> nobody uses oracle defaults
00:38 < boingolov> if they have a clue
00:39 < mawk> I have a 5V voltage regulator that is outputting 12V
00:39 < mawk> good thing I tested it first
00:39 < jim> brokd
00:40 < Bunk> I noticed a problem. My hdd has 5 partitions. And when i copy a file to the "parking" partition, i read " access denied "
00:40 < sensibel> phogg: the process is currently running , how this command can be given in the same shell?
00:40 < phogg> Hasimir: I'll agree that the fact that latex wasn't improved to the point where it could dominate PDF and ODT is disappointing, but I still will not hear an ill word.
00:40 < iflema> parking?
00:40 < jim> Bunk, so maybe you don't have write permissing to the dir you're trying to copy to?
00:40 < Bunk> where i move files to
00:41 < phogg> boingolov: every naive person uses them... the first time. And then later pay a member of the priesthood to perform the ritual needed to unbreak it without losing any data.
00:41 < Hasimir> phogg, there's a challenge in a footnote in the LaTeX section, if you can meet that challenge I'll reconsider
00:41 < Hasimir> good luck
00:41 < iflema> is there a big ring with lots of keys...
00:41 < Bunk> jim: Seems so, but i never had the problem
00:41 < phogg> sensibel: if the process is currently running you cannot change its ulimit in any case
00:41 < bls> a document with some computer modern, equations, and tikz? oh boy, that's obviously something important and smart
00:41 < ryouma> Bunk: it is the filesystem and permissions (and maybe some weird permission like things) that affect that, not hte partition per se
00:42 < jim> Bunk, where is this partition mounted?
00:42 < sensibel> phogg: then how this error can be resolved?
00:42 < boingolov> phogg: I was a member of the priesthood, once upon a time.  went back to development.  less buggary
00:42 < Hasimir> heh
00:42 < Bunk> ryouma: jim  The partition is /media/win_d
00:42 < Hasimir> well, bbl, gotta go check out of a hotel
00:42 < ryouma> Bunk: lsblk -i on it
00:42 < phogg> Hasimir: I am in no way that cool. Just a fan.
00:43 < jim> ok, and where on that partition do you copy files?
00:43 < Hasimir> then enjoy the swipe at Jeff Bozos instead ...
00:43 < phogg> sensibel: If you can't set a ulimit up then you need to raise the maximum limit and then try again. A ulimit on a running process cannot, as far as I know, be changed.
00:44 < phogg> Hasimir: oh I will
00:44 < Hasimir> and if Emacs be your thing, the config for as wide support of unicode as it's possible to get ...
00:44 < jim> Bunk, (my next question is...) ok, and where on that partition do you copy files?
00:44 < sensibel> Is there a way to add extend root partition instead of attaching a live usb/cd of operating system?
00:44 < Dan39> phogg: maybe think it can be up to the hard limit using gdb and setrlimit() ?
00:44 < Hasimir> my fontset config from my init file is at the end
00:44 < phogg> boingolov: you shouldn't say things like that to me or I'm liable to spend an hour yelling at you for things that aren't really your fault.
00:45 < jim> sensibel, do you have any interest in trying out lvm?
00:45 < sensibel> phogg: I already has changed it but the problem is the error that comes on terminal when I open it?
00:45 < jim> (it would give you that ability)
00:45 < phogg> Dan39: Hmm, I hadn't thought of that. I'm not sure he's up to that kind of trickery.
00:45 < Dan39> only read your message, nothing else :P
00:45 < boingolov> phogg: it's okay, I'm used to it
00:45 < phogg> sensibel: that error says that the limit could not be changed
00:46 < phogg> sensibel: where exactly are you trying to set ulimit that causes that error?
00:46 < Bunk> ryouma:  |-sda7   8:7    0 146,5G  0 part /media/win_e
00:46 < Bunk> jim: I moved the file from /Downloads to the partition
00:47 < sensibel> phogg: I set it using su ulimit -n 100000 at that time it gave no error
00:47 < phogg> sensibel: that isn't going to do what you want
00:47 < sensibel> jim: yes
00:47 < jim> Bunk, and you're using linux to do the copy? (maybe doing it from a term?)
00:48 < jim> sensibel, you  -are- interested... ok, give me a few mins
00:48 < phogg> sensibel: In a shell type: ulimit -n. This gives your current setting. Go ahead and try to change it that way, then check ulimit -n again. It will not have changed.
00:49 < Bunk> jim: MC did also fail. I dont know what is wrong // Yes, i just clicked copy in dolphin
00:49 < jim> sensibel, while I'm occupied, maybe this would entertain you: https://www.youtube.com/watch?v=BysRGDgqtwY
00:50 < phogg> sensibel: run 'ulimit -Hn' to check the hard limit. This is the maximum you are allowed to set it to without changing something as root.
00:50 < sensibel> phogg: [dellemc@localhost bilal]$ ulimit -n 100000
00:50 < sensibel> [dellemc@localhost bilal]$ ulimit -Hn 100000
00:50 < phogg> sensibel: don't provide a value, too, or you won't be told what the current setting is
00:50 < sensibel> phogg: is this error for limits.conf or for kernel settings?
00:51 < sensibel> phogg: I have not provided any value I have shown you the output
00:51 < phogg> sensibel: for? It's not for either of them, it's tell you about your current limits. If you need to raise your limits you will need to do so in limits.conf (or possibly also in the kernel)
00:51 < phogg> sensibel: Okay. That means your limit is already 100,000 files. Isn't that what you wanted?
00:52 < jim> Bunk, could you run: (ls -ld /media/win_d ; echo =========== ; df /media/win_d) | nc termbin.com 9999
00:52 < sensibel> phogg: yesss
00:52 < phogg> sensibel: Then what is the problem?
00:52 < Bunk> jim:  As user ?
00:52 < jim> sure, that's fine
00:53 < jim> Bunk, it would give you a url
00:53 < sensibel> phogg: the problem is whenever I open terminal i see this line on top of it bash: ulimit: open files: cannot modify limit: Operation not permitted why this is so?
00:53 < phogg> sensibel: One of your shell initialization files (.bashrc or .bash_profile or the like) is trying to set the file limit to above your hard maximum
00:54 < Bunk> jim: Ah, uh. All in one command. Lemme see
00:54 < jim> yeah :) because: all one pastebin
00:54 < jim> don't forget the ( ... )
00:55 < phogg> sensibel: fgrep ulimit ~/.bashrc ~/.bash_profile # you can try and find the offending line
00:56 < Bunk> ========== not found it says @ jim
00:56 < phogg> Bunk: you copied it wrong, then
00:56 < phogg> probably missing the 'echo'
00:56 < jim> it just puts a pretty line in the pastebin
00:57 < sensibel> phogg: thanks it worked , I set it to 100000 and now there is no such line
00:57 < jim> can you copy/paste it?
00:57 < Bunk> jim:  phogg  i need to try. i copied in kwrite first
00:58 < Bunk> (ls -ld /media/win_d ; echo =========== ; df /media/win_d) | nc termbin.com 9999
00:58 < jim> right, put that in your shell
00:59 < Dan39> a note... i just tested something interesting the other day. if you use `ulimit -n 2000` when your soft limit was 1024 and hard limit 4096, your hard and soft will be both set to 2000 now and you can't set above 2000 anymore. so use `ulimit -Sn 2000` to change just soft limit
00:59 < Bunk> zsh: ========== not found
00:59 < boingolov> honestly the reason I'm no longer a dba: the best day for a dba is the most boring day ever.  nothing bad happened, nothign good happened, the data is fine, everything's fine.  the worst day for a dba is everyone is upset at them, every decision they've ever made up to that point is called into question, it's absolutely awful.  there is no in between
00:59 < Bunk> jim: zsh: ========== not found
00:59 < Bunk> shrg
01:00 < phogg> Dan39: interesting. Just confirmed this, too.
01:00 < jim> hmm... yeah, that's sh scripting... try it this way:
01:00 < Dan39> phogg: yea, i accidentally did that while testing and was like "wtf?"
01:00 < jim> first, type: bash
01:01 < phogg> boingolov: that's why I stick with writing code and wearing a DBA hat only when it is needed. As a full time job it would not be fun.
01:01 < jim> Bunk, is bash actually installed?
01:01 < boingolov> phogg: that's wise
01:01 < phogg> zsh should be able to cope with that
01:01 < phogg> might need more whitespace
01:01 < boingolov> I made that mistake once.  I was the person in our group that sucked the least at database stuff, so they sent me to some very expensive training
01:01 < phogg> jim: yep, works just fine in zsh for me
01:02 < phogg> or wait, no it doesn't...
01:02 < Bunk> phogg:  jim  fish ate it, but no reply
01:02 < boingolov> I got certified
01:02 < boingolov> and then the group grew, and positions became more specialized
01:02 < phogg> Bunk: use this: (ls -ld /media/win_d ; echo '===========' ; df /media/win_d) | nc termbin.com 9999
01:02 < phogg> zsh is picky about unquoted = where bash is not
01:03 < jim> phogg, oh ok
01:03 < jim> thanks :)
01:03 < boingolov> I don't work there anymore.  but I've made it a point to downplay my systems and database experience in my programming carrer because of that
01:03 < jim> I'll remember that
01:03 < phogg> as will I
01:04 < Bunk> jim: Very cool. It spat out an url http://termbin.com/rmtc
01:04 < phogg> boingolov: I limit mine to putting down "Knows SQL" and later "Oracle/Postgres/MySQL" all in one breath. Keeps the focus where it should be.
01:05 < phogg> boingolov: hey, did you hear? sqlplus in the latest release gained command history support (if you supply an additional flag)!
01:05 < jim> Bunk, take a look at that url in your browser if you want
01:05 < phogg> Finally Oracle is joining the rest of us in the 1980s.
01:05 < Bunk> Yes, i did
01:06 < boingolov> lol @ phogg
01:06 < boingolov> no, I didn't.  but I always used rlwarp wth sqlplus anyway, so I had that anyway
01:06 < boingolov> I haven't used oracle since 11g
01:06 < boingolov> err, 11i
01:06 < phogg> boingolov: I use that when I can, but I can't always get people to agree to install it.
01:06 < boingolov> whatever it was called
01:06 < jim> ok, so you chmod 777ed it, and it's owned by root... which user are you likely to write in there as?
01:06 < jim> bunk: ^^
01:07 < phogg> boingolov: they also upped the varchar2 max to 16k, again if you supply a non-default mode. I think that was in 11g, though. So modern!
01:07 < phogg> identifiers still limited to 30 chars
01:07 < phogg> because COBOL says so
01:07 < boingolov> hehe
01:07 < jim> bunk, oh, you have a german locale active :)
01:08 < boingolov> well, I ran into windows limiting paths to 260 chars today
01:08 < boingolov> trying to compile go on windows
01:08 < Bunk> jim:  I thought the 100% was the prob. Ahm. Normally i write as user
01:08 < boingolov> I hate windows
01:08 < phogg> boingolov: only on some of the file APIs, I believe they have a low level one which is 260 per path component instead
01:09 < Bunk> i need to switch that locale off then
01:09 < boingolov> yes, that's all well and good, but go get / git fetch / etc don't know about them
01:09 < boingolov> git clone fails because the directories are too deeply nested
01:10 < boingolov> the ironic part is the offending package was the Azure SDK
01:10 < phogg> boingolov: probably not
01:11 < phogg> another classic rule of MS is that the lowest common denominator always wins and prevents progress
01:11 < phogg> which is why their directory delimiters still lean the wrong way even though NT has literally always supported both directions
01:11 < Bunk> jim: phogg  I tried mc to redistribute rights on those partitions, but failed
01:13 < boingolov> I think server 2016 solves the issue anyway
01:14 < boingolov> so that's fine
01:14 < phogg> boingolov: according to https://docs.microsoft.com/en-us/windows/desktop/FileIO/naming-a-file you can opt in to removing MAX_PATH restrictions by changing a key in HKLM (win10 only)
01:14 < boingolov> right, for win10
01:14 < boingolov> or server 2016
01:14 < boingolov> but not server 2012r2
01:14 < boingolov> which we are running
01:15 < nekoseam> wew
01:15 < phogg> a dirty Windows secret is that it really does have a single VFS, just hidden and with a very weird naming convention
01:17 < revel> I think someone yelled at me on this channel for claiming something like that :<
01:18 < phogg> \\ is the real root, you could literally change all local file path refs to use \\ if you know the weird syntax
01:18 < phogg> with the exception that not all of their APIs will accept such paths
01:19 < phogg> so technically there is a single VFS, but without universal support it may as well be three or four parallel ones which reach the same things
01:22 < jim> Bunk, had something come up... any progress?
01:23 < Bunk> jim:  No, im not experienced on setting rights on partitions
01:24 < Bunk> I took some attempts previously, but did not fly
01:29 < Bunk> jim: wouldn't it be chown command ?
01:29 < jim> Bunk, yeah... is that sda7 partition still empty right now?
01:30 < Bunk> no no, it is almost full.
01:33 < jim> Bunk, almost full. ok, are all the files in it owned by that user?
01:35 < jim> (you can use ls -l to see who owns the files)
01:36 < kerframil> Bunk: please run the following and convey the results in a pastebin post: id; findmnt -n /media/win_d; ls -ld /media/win_d
01:36 < Bunk> on the partitions questionable, the files are mostly owned by root
01:36 < kerframil> partitions are not filesystems. the filesystem type does matter - hence the inclusion of findmnt.
01:37 < Bunk> /media/win_d is all root
01:39 < Bunk> Here is an output of the command @ jim  https://hastebin.com/ogatoyeliz.makefile
01:39 < Bunk> And thanks for the support, as well in case i can not solve
01:40 < kerframil> Bunk: mount -o remount,uid=Bunk,gid=Bunk /media/win_d
01:40 < kerframil> Bunk: then ls -ld /media/win_d again and let use know what you see
01:40 < kerframil> us*
01:40 < Bunk> hm k
01:43 < Bunk> No change, it seems https://hastebin.com/ayocetazav.makefile
01:45 < kerframil> Bunk: that's curious. you might just need to unmount and mount it again. here's a template that you can use that is reasonably safe (grants ownership to you, doesn't allow every user to write): mount -t vfat /dev/sda6 -o uid=Bunk,gid=Bunk,umask=002,dmask=002 /media/win_d
01:46 < kerframil> Bunk: the reason that these options are needed is because vfat doesn't natively implement unix permissions. therefore, these options are used to 'fake' the desired permissions.
01:47 < kerframil> that said, the permissions already looked wide-open due to the mask being 0. but the combination of root being the owner and allow_utime being 002 could have caused some problems.
01:48 < Bunk> Jebus, where did you get that command from ? kerframil
01:50 < revel> Bunk: `man mount`
01:50 < revel> (hehe)
01:51 < Bunk> ya
01:51 < kerframil> Bunk: the options are explained in the mount manpage. what it means is to set the effective owner and group on all dir/file entries to appear as Bunk, and to use a file and directory mask of 002.
01:51 < Bunk> seems i need to be better
01:51 < kerframil> Bunk: masks are harder to explain but that basically means that it will mask out the write bit for users other than Bunk. which is safer.
01:51 < Bunk> Maybe it was a mistake to make the partitions FAT when i set up
01:51 < revel> 002 mask == 775 permissions.
01:52 < kerframil> Bunk: after mounting, you should see that the perms for a directory look like rwxrwxr-x for example (not the absence of the w bit in the last three characters. those apply to other users.
01:52 < kerframil> oops, note
01:54 < kerframil> Bunk: for the record, FAT has no permissions model at all (neither for Linux nor Windows). that's why chown/chmod aren't the solution here.
01:54 < Bunk> Ah, that is why mc did not work here
01:54 < kerframil> probably, yeah
01:55 < Bunk> i wonder why i was able to mv files to there at all, but it is like it is
01:57 < kerframil> Bunk: save this and read it when you have some time: http://termbin.com/078h
01:57 < kerframil> Bunk: that should help you to better understand
01:58 < Bunk> Very cool. I will pack that in my Linux dictionary
02:02 < Bunk> jim: phogg  kerframil  I will apply the command and see. Maybe, i need to check back on the problem
02:03 < jim> Bunk, umm which command?
02:03 < Bunk> the mount -t vfat /dev/sda6 -o uid=Bunk,gid=Bunk,umask=002,dmask=002 /media/win_d of kerframil s
02:04 < jim> oh, to mount it with unixlike ownerships
02:04 < kerframil> Bunk: just to be clear, you should umount /media/win_d before attempting to mount again - seeing as the -o remount trick wasn't good enough.
02:04 < Bunk> all clear
02:05 < jim> I'm confused though... what's /dev/sda7?
02:06 < jim> I thought it was 7, turned out to be 6... so I'm wondering what's 7
02:06 < jim> Bunk, ^^^
02:07 < Bunk> I was wondering about that, too.
02:07 < kerframil> judging by the mount naming scheme, most likely another vfat (or even ntfs) filesystem. findmnt can be used to confirm, either way.
02:30 < nothos> I think I've created an abomination
02:31 < nothos> PHP-FPM 5.3 running in ubuntu 16 in Windows' Linux Subsystem being proxyed to by IIS
02:31 < nothos> So dirty
02:32 < nothos> Would not recommend trying to do that
02:33 < revel> You're correct. You have.
02:33 < jason85> Any fun f2p games for linux?
02:34 < nothos> jason85 If you don't mind wine, League of Legends works well
02:35 < MtotheM> Is there any good documentation on what the difference between Linux and BSD are?
02:36 < jim> sensibel, hi... still around?
02:38 < nothos> MtotheM https://www.over-yonder.net/~fullermd/rants/bsd4linux/01
02:38 < tonymke> ^
02:38 < tonymke> an excellent writeup
02:38 < nothos> Gives a good summary of linux vs bsd from a real world perspective of using one vs the other
02:38 < MtotheM> ty
02:39 < jim> hi, please expand ty as thank you
02:41 < MtotheM> why be anal about it? it means the same thing. jim
02:41 < MtotheM> But yes, i really appreciate this. going to read it now.
02:42 < jim> MtotheM, we (the moderators of ##linux) feel it would help with understanding, especially where new english speakers are concerned
02:45 < triceratux> MtotheM: https://news.ycombinator.com/item?id=15711418
02:53 < Loshki> nothos: aquamarine on dark gray was such a bad choice
02:53 < nothos> Loshki Pardon?
02:53 < nothos> Oh, the site :D
02:54 < nothos> I just swap the theme right away
03:02 < sarcastico> Access IRC SERVER ---> /server irc.heckmann.top
03:02 < revel> No thanks.
03:04 < nothos> Ah, heck mann
03:09 < kubast2> Hey how can I check compression ratios on btrfs ?
03:09 < dannylee> fedora 27 might really be G0D
03:09 < kubast2> like even an overall ratio would be allright
03:09 < kubast2> ./compsize compsize.c
03:09 < kubast2> compsize.c: SEARCH_V2: Operation not permitted
03:09 < kubast2> ah
03:09 < kubast2> I'm dumb
03:09 < kubast2> nvm
03:09 < kubast2> nvm got it
03:09 < dannylee> its a great weekend for linux
03:10 < dannylee> some of you guys are Genius..but i'm learning from you>>> 8-)
03:13 < dannylee> ok i dont have gvim working..but i have bluefish..its really great editor..
03:14 < dannylee> i do have vi working in the terminal..just for kicks
03:15 < infinisil> Not often do I see Arrows in the wild in ##linux (>>>)
03:15 < dannylee> it come with linux
03:15 < dannylee> gvim is a bit of a weeeeeny
03:21 < dannylee> ok i dont have gvim working..but i have bluefish..its really great editor..
03:22 < infinisil> That's not how you spell emacs
03:23 < pnbeast> My osteopath spells emacs "b-o-a-t-p-a-y-m-e-n-t".
03:28 < ryouma> what doesthe arrows mean
03:31 < infinisil> ryouma: It's from Haskell
03:32 < infinisil> (>>>) :: Arrow a => a b c -> a c d -> a b d
03:32 < ryouma> no idea what that means.  give it to me in lisp syntax.
03:32 < Psi-Jack> Alright.. Beware.. I've got my hands in the NetworkManager source code, and I'm making changes!
03:32 < infinisil> ryouma: ((((>>>) :: Arrow a => a b c -> a c d -> a b d)))
03:33 < ryouma> you probably believe in anti anti anti missile missiles too
03:33 < triceratux> Psi-Jack: ha! you think you can improve on 1.12 ? https://www.phoronix.com/scan.php?page=news_item&px=NetworkManager-1.12-Released
03:34 < Psi-Jack> triceratux: Yes, actually. I'm fixing a little issue with the way it uses dnsmasq for a WiFi Hotspot.
03:34 < Psi-Jack> Yes, I know 1.12 just released today. :D
03:34 < phinxy> Are you going to send a pull request for that?
03:34 < Psi-Jack> Yes I am.
03:35 < Psi-Jack> Hence I said, BEWARE! Muahaha
03:35 < infinisil> ryouma: Sorry too hard to explain to a lisper in a sentence
03:36 < infinisil> ryouma: https://en.wikibooks.org/wiki/Haskell/Understanding_arrows
03:36 < Psi-Jack> Thanks to learning C++ for home automation, I know enough C itself to get by as well. :)
03:42 < Psi-Jack> Yeah, part of the problem I saw with NM and WiFi AP was how it loads in /etc/dnsmasq.conf and tramples over things it shouldn't, and at the same time wanted to use the default NM path for the dhcp-leasefile, when it should've used its own /var/lib/NetworkManager/dnsmasq-$unique.leases file tracking it specifically for that device.
03:43 < Bunk> phogg: jim ryouma Thanks again
03:43 < Bunk> for support
03:43 < Bunk> I need to go out !
03:43 < Psi-Jack> And now... I am running NetworkManager 1.12.0 with my applied fix, and it works great.
03:59 < Psi-Jack> Annnd, there we go. My code submission passes the vigilant NM CI tests.
04:28 < Dagmar> Psi-Jack: ^5
04:28 < Psi-Jack> heh
04:35 < hatp> Nothing will render unicode characters for me anymore after my latest update (at least my terminal and firefox). What could have been fucked? I really know nothing about font rendering
04:37 < Psi-Jack> hatp: Kindly mind the langauge here, please.
04:38 < MtotheM> Double check your locale / LANG maybe hatp
04:39 < MtotheM> Should work if you have UTF-8 set.
04:39 < hatp> cat /etc/locale.conf gives LANG=en_US.UTF-8
04:41 < Guest27487> Can I force lzo compression on new and existing files for /boot on btrfs ?
04:41 < Guest27487> and use zstd:4 for the rest of my partition ?
04:41 < rpifan> hello im back with some debug codes. do_page_fault(): sending SIGSEGV to pineapd for invalid read access from 0000005c
04:41 < rpifan> [  242.896155] epc = 77cd28a1 in libpcap.so.1[77cca000+2b000]
04:41 < rpifan> [  242.901557] ra  = 004053f8 in pineapd[400000+e000]
04:53 < AfonsoHenriques> howdy! I need to execute dozens of command-lines -- one per terminal, simultaneously. Is there a way to automate this?
04:56 < AfonsoHenriques> isnt fair to ctrl+shift+n (or t) 1000 times and ctrl+shift+v another thousand times
04:56 < pnbeast> AfonsoHenriques, cssh can be used to do it.  Tools like GNU parallel might be useful.
04:58 < koala_man> AfonsoHenriques: lots!
04:59 < koala_man> AfonsoHenriques: do you actually want them in separate terminals, or is that more of a manual workflow thing you did?
05:00 < AfonsoHenriques> pnbeast, see, I already made the command lines, I just need something to automate the boring process to copy the line from gedit, paste it to the terminal and run
05:01 < AfonsoHenriques> koala_man, separate terminals
05:01 < pnbeast> AfonsoHenriques, what?  If you describe your real problem/goal, you might get better advice.
05:01 < koala_man> AfonsoHenriques: are you familiar with screen or tmux?
05:01 < AfonsoHenriques> pnbeast, i'm trying, lets go on
05:02 < AfonsoHenriques> koala_man, no :( I tried tmux, but no results
05:02 < pnbeast> You don't seem to be.  You woke up this morning and decided "Hmm, today's a great day to run some program in many terminals!"?  I suspect you have some other goal.
05:03 < koala_man> AfonsoHenriques: which terminal are you using?
05:03 < AfonsoHenriques> yes, let me explain (not a native speaker + sleepy/tired [took already 60mg of ritalin but still tired]), wait
05:06 < AfonsoHenriques> I have tons of validation tests to do. A python script does those tests.
05:07 < AfonsoHenriques> so I need to put, e.g.,
05:09 < AfonsoHenriques> python script.py --validation-test-parameters [stuff-to-test]
05:09 < AfonsoHenriques> every single time.
05:10 < koala_man> how do you check if they fail?
05:11 < AfonsoHenriques> the terminal outputs the result
05:11 < pnbeast> AfonsoHenriques, why do you want multiple terminals to do this?
05:11 < koala_man> and you want to go through and check each of the terminals, rather than just get a simple "one of the tests failed, sorry"?
05:12 < Psi-Jack> This is making less and less sense.
05:12 < koala_man> sounds like a manual workflow half automated instead of fully automated
05:12 < pnbeast> Yes, I'm thinking that if he can write Python scripts, the rest should follow fairly naturally.
05:13 < AfonsoHenriques> because I want to do it simultaneously <==> want to see the output (the result) as fast as possible
05:14 < pnbeast> AfonsoHenriques, your test code is stochastic/non-deterministic?
05:14 < koala_man> AfonsoHenriques: so "no, I don't care about separate terminals, I just want the results immediately?"
05:14 < pnbeast> Regardless, this is not the fastest way.  It should all be automated.
05:16 < AfonsoHenriques> pnbeast, uh, no. I'm a noob person, just saw the meaning and it doesnt fit to "stochastic process"
05:17 < AfonsoHenriques> yes, it should, but in my context and considering my knowledge limits, it would be better to me as I said
05:17 < koala_man> AfonsoHenriques: if parallel -u --halt now,fail=1 < yourfile; then echo "All tests passed"; else echo "There was a failing test"; fi
05:18 < pnbeast> AfonsoHenriques, are you running the same test on different terminals?  If so, why?  Do you mean you're running different tests on each terminal?
05:19 < AfonsoHenriques> no
05:19 < koala_man> AfonsoHenriques: this runs one job per cpu. add -j 0 to run all jobs immediately
05:19 < AfonsoHenriques> the only thing that repeats along the commands is the python script and its parameters. So, no, we're talking about different tests
05:20 < ryouma> must ... master ... parallel
05:20 < ryouma> i keep mastering i8t and then having to look it up
05:21 < AfonsoHenriques> " Do you mean you're running different tests on each terminal?" YES!
05:21 < justsomeguy> parallel is bae
05:21 < koala_man> should you wish to run them in multiple terminals instead,    parallel gnome-terminal --tab -- bash -c 'eval "$1"; read' _ < yourfile
05:22 < pnbeast> AfonsoHenriques, if you don't want to write the parallelism into your testing script, then probably GNU parallel is a good choice.  Your tests should return some kind of pass/fail indicator, and write other information to a log or multiple logs, if required.
05:22 < pnbeast> AfonsoHenriques, when your testing fails, read the log file(s).
05:24 < AfonsoHenriques> god knows how fucking tired I'm now, but only dead (or with a solution for it) I shall find a solution
05:24 < AfonsoHenriques> pnbeast koala_man wait, please
05:24 < pnbeast> It's always best to do difficult coding when you're tired.  Good job!
05:26 < AfonsoHenriques> the clock just give 0 fucks about me. Need to do it stat. I'm glad you guys are helping me with this
05:27 < AfonsoHenriques> I'll follow koala_man suggestion first. brb
05:34 < AfonsoHenriques> koala_man, I almost cried from joy when I saw the terminals popping out, but no command-line were executed -- and:
05:34 < AfonsoHenriques> "Failed to parse arguments: Unknown option -c
05:34 < AfonsoHenriques> "
05:39 < AfonsoHenriques> now I see how awful my English looks. Please abstract it (the language I spoke as "native" have an Asperger level of logics and rules at syntax perspective [Poortuguese]).
05:39 < noregret> anybody used arpalert before? I don't quite understand the white/black lists
05:41 < AfonsoHenriques> anyway I'm still looking for a way to robotize this.
05:41 < pnbeast> noregret, can you give some detail?  From the website, it looks like it might be pretty simple.  If a machine is white listed, don't respond?
05:45 < AfonsoHenriques> pnbeast yes, it looks like parallel it the thing, but - as said and as it looks - probably I'm kinda dumb, I don't know how to handle parallel. In fact, never used until koala_man's suggestion.
05:46 < pnbeast> koala_man, I recommended it first!  Thief!!
05:46 < AfonsoHenriques> If you, well... Know what I need to write to parallel the commands from the list,
05:47 < koala_man> pnbeast: we recommended it in parallel but my output was buffered
05:48  * pnbeast curses at himself for forgetting the value of buffering.
05:50 < Tech_8> hi
05:52 < AfonsoHenriques> oy... the 'human nature' at least appeared after one suggestion; looks like I'll need to figure the fuck out how parallel and gnome-terminal works in a way to do what is needed here.
05:52 < AfonsoHenriques> thanks though
06:03 < AfonsoHenriques> oh
06:05 < jim> AfonsoHenriques, please watch your language while you're here
06:05 < AfonsoHenriques> the validation test I mentioned were actually a exploitation tool. The "validations" actually is servers and websites with cute vulnerabilities. Thousands of them.
06:06 < avis> he don't look in kid's or adult showers, or not afraid to check.  (he flee's)
06:06 < AfonsoHenriques> and my goal actually is root them all
06:06 < AfonsoHenriques> thanks again
06:06 < noregret> pnbeast: well, i have both deny and allow lists empty, which should mean that any device will trigger the script, i tried it but the script wasn't triggered. ALso, I don't know how to clear/reset the already detected devices, i'm confused
06:09 < avis> you could be punk
06:09 < Tech_8> ?
06:10 < avis> some people jack people.  they act like priest too, or at least -- "dumass" had the remote.
06:10 < avis> i can't stand photographers
06:11 < avis> i got swiss clock
06:12 < avis> you play punk / pong ?
06:14 < pnbeast> Here's the Friday night I know and love!
06:17 < CoolerZ> hey anyone here who uses flex and bison?
06:17 < CoolerZ> where is the complete regex specification given for flex?
06:18 < CoolerZ> i found this http://people.cs.aau.dk/~marius/sw/flex/Flex-Regular-Expressions.html
06:18 < CoolerZ> but that doesn't seem to have stuff like \n
06:19 < CoolerZ> not \n i mean what it actually provides as a match
06:42 < Loshki> CoolerZ: anything here? https://en.wikipedia.org/wiki/GNU_bison
06:45 < CoolerZ> Loshki, i am little bit confused on how you interoperate flex and bison
06:46 < CoolerZ> if i match something with a regex in flex, and then you want to access that value in bison
06:46 < CoolerZ> in a bison action, for example you match a number and you want to parse the number and maybe add it with another number
06:47 < CoolerZ> yytext seems to have something to do with it
06:47 < CarlenWhite> Is there any particular problem when you run mkfs on a disk and not a partition of that disk?
06:48 < CoolerZ> also what about capturing groups and such, if your regex matches a long string and you want to capture parts of it
06:48 < CoolerZ> in javascript regex you have that concept of capture groups
06:50 < Loshki> You are not obliged to use them together, but flex provides a lexical analyser in exactly the right format to interface to bison, which is a syntax parser. These are the first two stages of compiling/interpreting most languages.
06:59 < willd_> anyone have any interest in helping me troubleshoot a rclone issue I am having with google drive?
07:00 < Pentode> 333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333
07:03 < kubast2> Hey how do I properlly add :<number> to fstab ?
07:03 < BlindWiz> hey, I just did an apt-get dist-upgrade on my ubuntu 16.004 server, and it brought up a prompt telling me that it couldn't figure where to install grub to.  and its giving me a list of partitions of where to install it.  Does grub gon on my /boot partition?
07:03 < kubast2> option=value:value2
07:03 < kubast2> is what I want to do
07:04 < kubast2> I can't mount -o remount btrfs rootfs for whathever reason so I can't put it as soon as fstab is done
07:06 < strive> CarlenWhite: You may wipe data.
07:13 < CarlenWhite> Mmhm. Was mostly seeing if it'd cause some weird program crashes.
07:13 < CarlenWhite> I'm having a fun time with MSSQL and I kinda want to die.
07:20 < BlindWiz> hi all, just did an apt-get dist-upgrade and it wants to upgrade grub.  but it as asking me where to upgrade it for some reason.  I am not sure where it is installed, this is a virtual instance, and was spun up via a cloud automation...  How can I determine where grub is currently installed?
07:25 < curiousx> BlindWiz: which are the options ?
07:27 < curiousx> Although, tobe honest i would ask the support if you are paying for a VPS :p
07:28 < kerframil> BlindWiz: to begin with, try `dd if=/dev/sda count=1 | hexdump -C` (or /dev/vda)
07:28 < TheNH813> What's the best way to resolve hostnames?
07:28 < TheNH813> At least for ssh connection.
07:29 < TheNH813> Because I'm tired of my desktop occasionally ignoring static ip settings and having to use nmap to find it.
07:31 < curiousx> TheNH813: i don't understand what are you trying to do
07:31 < BlindWiz> curiousx /dev/sda, /dev/sda1, /dev/sda2 /dev/sdb and /dev/dm-0
07:32 < TheNH813> curiousx: Basically, I want to be able to type "ssh administrator@hostname" and have it work.
07:33 < TheNH813> curiousx: I don't want to run a local DNS server, so should I set up nmbd?
07:34 < kerframil> TheNH813: a functioning DNS infrastructure. if your router uses dnsmasq, this could be an easier option than it seems. and no, nmbd is not appropriate.
07:34 < BlindWiz> curiousx /dev/sda1 says /boot, and /dev/dm-0 says ubuntu root, the rest say virtual_disk next to it.
07:34 < curiousx> Well, a dns like google's (8.8.8.8 - 8.8.4.4) or opendns should do that TheNH813
07:34 < kerframil> not for private hosts, it won't
07:34 < BlindWiz> can'
07:35 < curiousx> TheNH813: you could set remote dns servers on /etc/resolv.conf although that file get overwrited it :p
07:35 < TheNH813> Hmmm...
07:35 < BlindWiz> t you tell your router to assign a  static ip based on your card's hardware id
07:35 < RustyShackleford> I need to access this web app remotely as if i'm on localhost:32400
07:35 < TheNH813> Funny enough, it dosen't support that. And it dosen't support OpenWRT. Those were my first thoughs.
07:35 < kerframil> BlindWiz: yes. and if it runs dnsmasq, chances are it's possible to pick an arbitrary DNS suffix and it all just works.
07:35 < curiousx> BlindWiz: then /dev/sda1 is the one
07:36 < RustyShackleford> need to open the app locally and allow remote access. But thats only allowed if you're localhost
07:36 < curiousx> BlindWiz: tobe sure, you could mount it and see what it has inside
07:36 < BlindWiz> couriousx thanks.  I thought so, but I wanted to make sure.  thanks a million
07:36 < BlindWiz> curiousx I meant. :)
07:36 < curiousx> np
07:36 < RustyShackleford> wanna help me find the correct ssh options?
07:36 < kerframil> curiousx: why do you think that? it's far more common for grub to be installed to the MBR. that's why I recommended an inspection of the first sector.
07:36 < TheNH813> I used to have a local dns cache server. Maybe I'l go turn it on and get it running again. A bit troublesome though.
07:37 < curiousx> kerframil: cuz if it has the boot flag, then grub is there
07:37 < kerframil> curiousx: no. you're conflating the block device given to grub-install with the device that back's grub's root filesystem. usually, they are different things.
07:37 < kerframil> I recommend checking to be sure
07:38 < curiousx> kerframil: ofc mount sda1 to see what's inside
07:39 < kerframil> no. if it's an MBR setup, the proof would come from dumping the first sector and finding GRUB's initial payload. it's easy to identify because it contains the word "GRUB".
07:39 < curiousx> kerframil: but he's on a VPS, things are diferents
07:39 < kerframil> curiousx: nobody knows that unless it's tested for; that's all I'm saying
07:39 < CarlenWhite> Is there an fstab option that asks that a dependent mount be mounted first?
07:40 < CarlenWhite> e.g. a block device inside a partition.
07:40 < curiousx> it's not a personal PC, so yeah, BlindWiz mount it a make a paste of it, to check first
07:40 < kubast2> Which kernel version enables this feature: https://btrfs.wiki.kernel.org/index.php/Compression#Can_I_set_the_compression_level.3F
07:41 < kubast2> I presume any over 4.15.0-20
07:41 < kubast2> it's my 7th reinstall cause I cucked my fstab
07:41 < searedvandal> "ZSTD level support is planned."
07:41 < kubast2> ah
07:42 < kubast2> so it's not fully implemented yet searedvandal
07:42 < searedvandal> you can't add compression level yet.
07:42 < kubast2> ah
07:42 < kubast2> that explains a ton
07:44 < kubast2> I have no idea how I made remount,compress-force=zstd:4 work at some point
07:44 < kubast2> then :shrug:
07:44 < kubast2> unless it works only for the default level or smthn
07:44 < kubast2> yeh
07:44 < kubast2> nvm
07:47 < TheNH813> Okay, neq question. Why does zsh give errors when typing in ip addresses for nmap?
07:47 < curiousx> xD
07:48 < curiousx> idk, never i always used bash :p
07:48 < TheNH813> I tried to type: sudo nmap -sn 192.168.0.*, but it returns "zsh: no matches found: 192.168.0.*"
07:48 < curiousx> idk, never *use zsh*, i always used bash :p
07:48 < TheNH813> I guess it's trying to interpret them as filenames?
07:48 < searedvandal> probably
07:48 < TheNH813> I just decided to try zsh for a day.
07:49 < TheNH813> Guess unless I can figure out how to exclue those, it's back to bash.
07:49 < curiousx> use capital S
07:49 < TheNH813> Hmmmm.... no difference
07:50 < curiousx> then idk, i think the last time i use nmap was like 6 years ago :p
07:50 < pnbeast> TheNH813, what does 192.168.0.* represent?
07:50 < go|dfish> TheNH813: try escape (quote) the * e.g. \*
07:50 < TheNH813> Ip addresses,, of course.
07:50 < curiousx> Oh! i know use sudo nmap -sn 192.168.0./24
07:50 < pnbeast> TheNH813, wrong.  Look up shell wildcards.
07:51 < curiousx> cdir notation FTW \m/-_-\m/
07:52 < CarlenWhite> Today we try to fit a 20GB file in a 22GB partition. And fail. Somehow.
07:53 < TheNH813> curiousx: That's actually not a bad idea at all. Except I use a /22 prefix, just because. :p
07:54 < curiousx> Well... 24 will look from 0 to 255 idk how many /22 will look after on you netwrok, you gotta grab an ip calculator to know that
07:58 < CarlenWhite> 20GB file in a 22GB partition. 32% used.
07:58 < CarlenWhite> Hrrrm.
07:59 < CarlenWhite> Okay I guess.
08:01 < revel> That darn compression, am I right?
08:03 < CarlenWhite> It's a ext4 blockfile that was moved and it's not completely used so...Eeeeeh?
08:03 < CarlenWhite> I'm doing some weird loops to get MSSQL to play nice inside a zvol.
08:03 < CarlenWhite> Like some weird condom.
08:12 < Elladan> Aren't sparse files fun.
08:16 < pnbeast> curiousx, you need an "ip calculator" to do 2 ^ 10 (or so)?
08:24 < iflema> muh hat
08:24 < toothe> anyone know how in grub to list the kernels?
08:24 < iflema> toothe: mkconfig  with out writing to file?
08:25 < iflema> grub-mkconfig that is for those that update-grub
08:26 < CarlenWhite> Well the ext4 block file inside the ext4 partition from a zvol works.
08:26 < CarlenWhite> Time to fuckit-shipit
08:27 < iflema> CarlenWhite: flip it.... shipit... fuckit
08:29 < CarlenWhite> It's Good Enough™
08:30 < CarlenWhite> It works ontop ZFS so I get what I want from that.
08:30 < curiousx> pnbeast: i do need a ip calculator to divide a subnet, how do you do that ?
08:30 < CarlenWhite> It doesn't suffer from a chick-egg issue
08:30 < CarlenWhite> Everything's good now.
08:31 < CarlenWhite> fstab should run sequentially, right?
08:32 < CarlenWhite> Er. Some pages are mentioning it might not, but it should recognize that a mount exists inside another.
08:33 < pnbeast> curiousx, /23 is 2 times /24 (2 ^ 8 * 2), and /22 is 2 times /23.  It's pretty straightforward.
08:35 < CarlenWhite> And I don't have a testicular fortitude and energy to test that by restarting the server when I'm 40 minutes away from it at 3AM.
08:35 < dnanib> curiousx: You should be able to do that in your head, otherwise on many Linux variants you have the ipcalc command
08:37 < curiousx> yeah, but i'm not too much of an Einstein :D
08:37 < pnbeast> Okay.  Why the excitement over CIDR notation?
08:37 < curiousx> ik ik, it's not that complicated but i'm just lazy, i do you bc for simple * and / as well -.-
08:38 < curiousx> The excitment over CIDR to me is to really ban troll players on a game server \m/-_-\m/
08:39 < dnanib> You needn't be Einstein. It is like knowing basic arithmetic.
08:39 < curiousx> ik
08:40 < curiousx> bb
08:49 < mini0n> yo
08:49 < mini0n> r8188eu  <-- having disconnect issues with this driver .. any clues ?
08:54 < CapTn_MoRgun> r8188eu  <-- having disconnect issues with this driver .. any clues ?
08:54 < CapTn_MoRgun> wifi
09:29 < Pentode> CapTn_MoRgun, try turning off power management for the adapter.
09:29 < Pentode> e.g, iwconfig wlan0 power off
09:35 < paulcarroty> folks, evolution can check an emailbox for new mail in background? Seems like I have enabled this feature, but actually I never got any notifications.
09:46 < jim> paulcarroty, you should check to see if you're receiving mail
09:48 < za1b1tsu> Hello, how come moving files to /usr/local/bin does not require sudo?
09:50 < SuperSeriousCat> Welcome to the matrix
09:52 < Dynamicfail> what is the best way to change default gateway for current session, but would be lost on restart?
09:52 < cluelessperson> I have an issue,  when I do,  "umount /dev/loop0"  it says,  "is not mouned"
09:52 < cluelessperson> when I do,  mount /dev/loop0    "already mounted"
09:52 < cluelessperson> wtf
09:54 < SuperSeriousCat> Dynamicfail, sudo route add default gw <temp-gateway> should work
10:10 < NetTerminalGene> local priviledge escalation requires physical access to pc?
10:13 < hexnewbie> NetTerminalGene: Not physical, just local user and/or code executions. It depends on the attack vector, some may require physical access, but that would be an exception
10:17 < hexnewbie> Er, that was misleading. Physical vulnerabilities are way more common, as physical access makes any access pretty easy, but those aren't often counted as vulnerabilities, and when they are counted (e.g. xscreensaver), they aren't considered under the umbrella of local privilege escalation.
10:19 < MrElendig> za1b1tsu: because you did bad things
10:55 < Dagmar> Too right
11:12 < toothe> Is docker built on lxcs or cgroups
11:12 < toothe> that's a question :) ?
11:12 < MrElendig> those are not mutally exclusive
11:12 < toothe> oh? perhaps I don't understand the relationship of the two.
11:13 < MrElendig> lxc uses cgroups too
11:14 < toothe> so, the primitive is cgroups?
11:14 < MrElendig> "Docker is primarily developed for Linux, where it uses the resource isolation features of the Linux kernel such as cgroups and kernel namespaces, and a union-capable file system such as OverlayFS and others"
11:15 < MrElendig> "Since version 0.9, Docker includes the libcontainer library as its own way to directly use virtualization facilities provided by the Linux kernel, in addition to using abstracted virtualization interfaces via libvirt, LXC and systemd-nspawn."
11:15 < toothe> I see...I think I need to fully understand both and their relationship
11:16 < MrElendig> docker can use multiple interfaces, including lxc
11:16 < toothe> i see.
11:16 < MrElendig> see the wikipedia page and the docker documentation
11:16 < toothe> if I understand correctly, there are multiple ways to do the same functionality?
11:22 < Pusteblume> how can i convince apt-cdrom to read the metadata from /media/user/pendrive?
11:33 < socomm> Pusteblume: probably, ask #debian
11:33 < toothe> err...how do I set a static IP in ubuntu?
11:33 < toothe> my configuration seems to not be working.
11:44 < user1_fn> I connected to a server with "ssh -R 612345:localhost:612345 user@host.tld". "netstat -tulpen" on the server side shows me that there is LISTENd to this port. On the client side I execute "while ( true ) ; do nc -l 612345 ; done", and then on the server side "date | nc localhost 612345". The last command fails with "connect_to localhost port 612345: failed." Any suggestions?
11:44 < user1_fn> (Background: I would like to get some text from the stdout of a program on the server side to the stdin of a program on the client side)
11:45 < barometz> probably want to use a port below 65536
11:46 < zenix_2k2> one question, i think there is a command which monitor log-files changes on linux but i kinda forgot, so what is it again ?
11:46 < user1_fn> barometz, 612345 is I typo =:) Of course under 65536.
11:47 < longxia> zenix_2k2: inotify ?
11:47 < zenix_2k2> no like when i add changes to a file, that change is gonna be printed out
11:48 < longxia> tail -f
11:48 < zenix_2k2> oh right, tail
11:48 < zenix_2k2> thk
11:48 < zenix_2k2> thank*
11:53 < zenix_2k2> so er, one more question that is related to tail -f, so i want to add something to the beginning of every time it prints out like this --> https://pastebin.com/jCbGFUwh, is it possible ?
11:54 < kerframil> zenix_2k2: tail -f file.txt | while read -r; do echo "=> $REPLY"; done
11:58 < zenix_2k2> wow, that is easy
11:58 < sy> Hello. So I bought a gaming mouse and the drivers are only for windows
11:59 < sy> I'm wondering. Is there anything I can do with this https://0x0.st/spA7.txt to tweak the settings
11:59 < sy> I only need to turn off the annoying LEDs
11:59 < sy> and id like to do that without getting out my soldering iron and screwdriver
12:02 < hexnewbie> sy: A no-name mouse with proprietary Windows-only drivers? I don't think it's likely there would be a tool. You didn't say which directory the tree was for, but there doesn't appear to be any LED interface for the mouse from what I can tell.
12:02 < hexnewbie> sy: Open the mouse and unsolder the LEDs?
12:06 < Pusteblume> oh come on. why do they put ufw on dvd-2. one of the first things you do after an install is to setup ufw, no?
12:07 < hexnewbie> Pusteblume: Certainly not something I've done, or known someone do, no.
12:07 < sy> ^
12:07 < Pusteblume> you guys go without firewall?
12:07 < sy> no
12:07 < sy> we use iptables
12:07 < Pusteblume> :)
12:07 < Pusteblume> nftables you mean
12:07 < CapTn_MoRgun> iptables rock
12:07 < sy> but maybe not right after install
12:08 < sy> i actually hate iptables and use pf at every opportunity
12:08 < kbob> packet filtering?
12:09 < hexnewbie> Figuring what you want firewalled and how is certainly more difficult than writing the iptables rules.
12:12 < socomm> Pusteblume: depends, if you've network firewall you might be able to get away without a firewall on your end-host.
12:15 < searedvandal> no firewall and open all the ports, that's the best way to internet
12:15 < socomm> searedvandal: certainly more fun.
12:15 < searedvandal> a bit of excitement is important
12:16 < pingfloyd> average joe internet
12:16 < socomm> pingfloyd: average joe is behind a NAT.
12:17 < pingfloyd> firewall isn't just nat
12:17 < hexnewbie> Why are the ports open even if you have no firewall? I mean, firewall is good, but it's not substitute for real security. Does a default GNU/Linux install open (m)any ports by default?
12:17 < searedvandal> Hi, I'm average Joe. I hide behind NAT and my ISPs brilliant firewall setting on the router. Am I safe now?
12:18 < socomm> hexnewbie: I don't understand your question.
12:18 < pingfloyd> hexnewbie: probably because they run a bunch of daemons they don't even use
12:18 < pingfloyd> a port isn't really open unless there is something listening
12:19 < hexnewbie> socomm: If you have no open ports, you don't really need firewall. And I don't thing nowadays you get inetd, finger, ident, etc. installed and open by default like in the good old days
12:19 < socomm> searedvandal: Sure. Normally its not the lack of firewall that gets most users, its the fact tehy install junk software on their computers.
12:19 < pingfloyd> I'd still use a firewall
12:19 < searedvandal> socomm, good :)
12:19 < pingfloyd> especially today
12:19 < socomm> hexnewbie: you'll inevitably have some form of software that listens on a network port.
12:20 < hexnewbie> Yeah, sure, it's good. I like to explicitly disallow any incoming connections until I explicitly allow them, in case I accidentally install some daemon, etc.
12:20 < searedvandal> I'll be honest, I hate messing around with iptables, so on my desktops and laptops I just use ufw. just deny everything and it's all good
12:20 < pingfloyd> the firewall gives you one point to manage it all
12:21 < bazhang> gufw is the front end iirc
12:21 < hexnewbie> socomm: Then your jabber file transfers and DCC won't work. ;)
12:21 < pingfloyd> most of the time, users install software without knowing they're also installing a daemon that listens on a non-local interface
12:22 < pingfloyd> I wonder how many open ports the average dists comes with
12:22 < pingfloyd> a lot of dists have many daemons/services installed by default
12:22 < bazhang> including minero miners?
12:22 < Dagmar> If you install as a desktop role, very few, if any
12:22 < searedvandal> speaking of firewalls, pfSense vs ClearOS, is there one that's better than the other?
12:22 < pingfloyd> they don't follow the philosophy of secure by default at all
12:22 < Dagmar> Usually just ssh
12:23 < Dagmar> We have absolutely yelled at distro maintainers about that enough times
12:23 < hexnewbie> I usually have to install the ssh *server* manually, so I'd guess zero.
12:23 < Dagmar> I actually called some RH reps incompetent for thinking it was a good idea to emulate Windows in that respect
12:23 < Dagmar> To their faces.
12:24 < pingfloyd> hexnewbie: having to install it is a more sane policy (debian is like that)
12:24 < BluesKaj> Hiyas all
12:24 < pingfloyd> well, unless you choose the server tasksel in the installer
12:24 < pingfloyd> I usually uncheck all tasksels when I do an installation.  I wonder if some users check them all
12:25 < socomm> Dagmar: emulate windows in what respect?
12:25 < hexnewbie> You would have CUPS on a desktop, though. I don't know if that listens on localhost only, or all interfaces, though.
12:25 < pingfloyd> I can see that happening.  The reason being it's all greek to the user so they think, "I better just install everything to be safe" (ironic)>
12:25 < Dagmar> socomm: Install every damn thing active and listening on the network by default
12:26 < Dagmar> Now your main risk is basically mDNS buggery
12:26 < pingfloyd> redhat more or less pushed systemd on the world...
12:26 < pingfloyd> so I would say it's not a stretch at all that they're trying to be like windows
12:26 < oiaohm> pingfloyd: it was not that everything was that great before systemd.
12:27 < Dagmar> They're just trying a different way
12:27 < pingfloyd> (pun intended)
12:27 < Dagmar> I know I wasn't the only one calling RH people out at conferences over that policy
12:27 < pingfloyd> good
12:28 < pingfloyd> I hate when dist communities become cults
12:28 < pingfloyd> that mentality that whatever their dist devs do is right.
12:28 < oiaohm> Dagmar: really I have done a lot of custom distribution build over the years.   assmebling sysvinit system from scratch was down right painful.
12:29 < pingfloyd> oiaohm: you mean writing all the startup scripts?
12:29 < Dagmar> oiaohm: Being that I converted some things which weren't sysV _to_ sysV by writing the scripts myself, I didn't think it was that hard
12:29 < oiaohm> Dagmar: pingfloyd its not writing the scripts.
12:29 < oiaohm> It was having like udev and the like that were all independant projects cat fight.
12:30 < Dagmar> That doesn't really have anything to do with sysV
12:30 < oiaohm> Dagmar: its not like sysv without hotplug and console control and other things was highly useful.
12:31 < Dagmar> This is a perfect example of why this systemd stuff pisses me off
12:31 < Dagmar> Hotplug is _not_ part of init, man
12:31 < Dagmar> It's not init's job
12:31 < pingfloyd> yeah, look how it's changed people's thinking
12:31 < oiaohm> Dagmar: the script did not help that I had a BSD background rcorder was normal.
12:31 < pingfloyd> now people think init is more than it is because of it
12:31 < Dagmar> All this hardware enumeration and management stuff are system facilities
12:32 < pingfloyd> exactly
12:32 < Dagmar> They've nothing whatsoever to do with init
12:32 < pingfloyd> but now everyone thinks its init's job along with everyone else
12:32 < oiaohm> Nothing to do with init.
12:32 < oiaohm> Hmm
12:32 < oiaohm> where do you get that idea.
12:32 < hexnewbie> My main issue with systemd is its tendency to fight what *I* am doing by introducing unanticipated changes of system behaviour across the board. For example, if I'm converting a partition table  I don't like the filesystems on it activated while I'm converting it.
12:32 < oiaohm> old udev would be calling sysvinit services.
12:32 < pingfloyd> for their firewall, system logging, networking, and so on
12:32 < Dagmar> A few f**King decades of hands-on experience with a ridiculous number of Unix machines.
12:33 < lupine> it's certainly different
12:33 < lupine> but that's a neutral statement
12:33 < pingfloyd> sysv init was around long before udev even existed
12:33 < pingfloyd> if udev is tightly coupled with sysv init, that's the dists' fault
12:33 < hexnewbie> It's a bit late to blame (just) systemd for it, though. Since everyone has adopted it, even the kernel and other subsystems have switched to activating everything by default.
12:33 < Dagmar> Init's only job is launching the binaries that the system needs to perform in its current role
12:33 < oiaohm> pingfloyd: that is part of the problem.   hotplug support was very hacky on top it.
12:34 < pingfloyd> Dagmar: something that LP and friends are in denial about
12:34 < Dagmar> It's why they're struggling
12:35 < oiaohm> Dagmar: zombie clean up being given to PID1 was a sysv idea.
12:35 < oiaohm> Dagmar: so sysv init has the start of service management clean up.
12:35 < oiaohm> Dagmar: not very good service management clean up.
12:36 < Dagmar> No, zombie cleanup was handed to init, which did not necessarily need to be PID1
12:36 < drellok> hi guys, my machine is giving me segmentation fault on almost every program I am trying to run. any ideas what can be wrong?
12:36 < Dagmar> ...and it was handed to init because it's basically the only way to get rid of zombies without _refactoring POSIX_
12:36 < oiaohm> Dagmar: zombies being linked to PID1 for cleanup comes straight from sysv
12:36 < pingfloyd> systemd's design is a house of cards
12:37 < Dagmar> I used enough bloody systems that did _not_ have a reaping init
12:37 < Dagmar> It was deeply and profoundly annoying
12:37 < hexnewbie> drellok: When did it start, and did you do something to possibly trigger it?
12:37 < pingfloyd> if you want to know the future of systemd, look at Windows Server
12:37 < Dagmar> init would _automatically_ be the parent of an orphaned child _anyway_
12:37 < oiaohm> pingfloyd: reallly sysvinit design was also house of cards.
12:38 < Dagmar> So, it's more a matter of it being the only process able to clean up that mess
12:38 < pingfloyd> oiaohm: not really
12:38 < drellok> hexnewbie: it starts after several hours of running a pc, if I reboot it is gone and then it returns after a few more hours
12:38 < pingfloyd> oiaohm: it doesn't matter with sysvinit if other daemons crash
12:38 < pingfloyd> as long as init is still running stable, you're fine
12:38 < drellok> hexnewbie: I must say this is a remote dedicated server accessible via SSH only
12:38 < hexnewbie> drellok: You can't run dmesg either, right? Anything bad in /var/log/kern.log after that happens?
12:39 < pingfloyd> with systemd, somethingd crashes and it can systemd init with it
12:39 < oiaohm> pingfloyd: So init running all remote services down and you are locked out the box is stable hmm that does not seam to be that useful.
12:39 < Dagmar> You really haven't spent a lot of time on architecture
12:39 < pingfloyd> oiaohm: that's a rigged comparison
12:39 < pingfloyd> losing remote access has nothing to do with init
12:39 < Dagmar> It's an insane comparison
12:39 < drellok> hexnewbie: I can run dmesg
12:40 < oiaohm> pingfloyd: no it not the real world what I would run into.
12:40 < pingfloyd> totally conflating things
12:40 < drellok> hexnewbie: I see a lot of those - rejecting I/O to offline device
12:40 < oiaohm> pingfloyd: with watchdog and other things failing.
12:40 < hexnewbie> pingfloyd: systemd crashing is, from my limited experience and observations, way rarer than kernel crashes, so I'm not sure if that's a genuine problem.
12:40 < pingfloyd> oiaohm: what does that have to do with sysv init?
12:40 < drellok> hexnewbie: looks like disk problems to me
12:40 < hexnewbie> drellok: Your disk is dying
12:40 < oiaohm> pingfloyd: sysvinit design lacks proper service tracking.   So a service crashes it may not restart properly.    So having  watchdog to restart a service is not much use.
12:41 < Dagmar> Bullshit.
12:41 < oiaohm> Dagmar: no bullshit most of my problems traced to that with sysvinit.
12:41 < Dagmar> Ensuring that a service starts properly is the responsibility of the service-launching script.
12:41 < pingfloyd> see what I mean by how systemd has brainwashed many users
12:41 < Dagmar> Your problems stem from that you have a wildly overblown idea of what responsbilities fall to init
12:41 < pingfloyd> they're not being logical
12:42 < drellok> hexnewbie: I can do ls and list through the files though
12:42 < hexnewbie> drellok: I seem to recall segmentation fault when running programs off a crashed filesystem and/or forcibly detached/timed out root volume. I assume you're not using RAID there? You need to migrate the system to another disk so it's stops crashing (and do this before you need to restore it from your backups)
12:42 < Dagmar> It's not init's job to _keep_ services running, or to ensure that they're sane, or to even care
12:42 < pingfloyd> thinking in systemd's terms of how things work
12:42 < Fulgen> Dagmar: systemd is a daemon manager and an init process
12:42 < Fulgen> not only an init proces
12:42 < oiaohm> sysvinit it self did not stop at just defining init process.
12:42 < hexnewbie> drellok: Yeah, in that scenario the cached files tend to be still readable/executable.
12:42 < Dagmar> It's job is to launch the scripts that launch those processes, the way the administrator intended, as the system enters the appropriate runlevels
12:42 < Dagmar> Fulgen: So what
12:43 < pingfloyd> see what I mean
12:43 < oiaohm> Dagmar: and those sysvinit script designs like using PID that recyled and so on are paths to failure.
12:43 < Fulgen> "<Dagmar> Hotplug is _not_ part of init, man <- so it isn't only an init process, so it is its job
12:43 < Dagmar> oiaohm: We already covered this.  They didn't "using PID that recycled"
12:43 < Dagmar> Fulgen: Are you seriously going to claim that hotplugging hardware is init's job?
12:44 < oiaohm> Dagmar: Linux kernel is design to recycle PID numbers it part of it.
12:44 < drellok> hexnewbie: I actually thought about that and after reboot I checked smartctl status and it seemed ok.. is it possible that smartctl does not see problems on dying disks?
12:44 < Fulgen> Dagmar: no. it isn't _init_'s job. but systemd is _more_ than an init system
12:44 < pingfloyd> Fulgen: nss
12:44 < Dagmar> Fulgen: That's nice.  We're not talking about system.
12:44 < Dagmar> er systemd
12:45 < oiaohm> Dagmar: the big problem with the sysvinit scripts is that it was design for a operating system that is not LInux.   So they are basically dud off the start line.
12:45 < pingfloyd> we've got a Captain obvious
12:45 < Dagmar> oiaohm: You're free to apply a patch that will randomize PIDs for no good reason, but again... this has nothing to do with the bloody number 1
12:45 < hexnewbie> It's not a question of what's init's job, though. The real question is, is it useful to have service restart handled by your init system. And I tend to prefer that to have all my services basically ran by supervise.
12:45 < oiaohm> Dagmar: randomising does not fix recycling clash.
12:46 < Dagmar> I prefer to be a mature adult and write service scripts that correctly start things in all cases, and service monitors to let me know when something actually falls over
12:46 < pingfloyd> Fulgen: how is any of that germane?
12:46 < Dagmar> oiaohm: THere is no "recycling clash"
12:46 < Dagmar> There are, however, people dumb enough to assume that a PID is a good thing to track a job with
12:46 < oiaohm> Dagmar: I have had service not restart because another one has taken the PID it had and it was claiming it was running.
12:46 < Dagmar> You had crap service scripts
12:46 < oiaohm> Dagmar: and everything was to sysvinit designs.
12:47 < oiaohm> Dagmar: no crap service scripts designed as per sysvinit documentation scripts.
12:47 < oiaohm> Dagmar: basically sysvinit in this department was total crap.
12:47 < Dagmar> Well, I'm completely sorry you didn't understand those scripts were not designed to be perfect and it was up to local administrators to write ones which served any further purposes
12:48 < Dagmar> I never found it that bloody hard to check the process names
12:48 < oiaohm> Dagmar: I had two apache servers so matching process names.
12:48 < Dagmar> I did however see very many lazy tools writing service scripts that made assumptions from kill 0
12:48 < Dagmar> oiaohm: You failed then
12:49 < Dagmar> Sysv can be summed up with the one sentence I recited earlier
12:49 < pingfloyd> could have used a sane dist that came with decent scripts
12:49 < pingfloyd> not many of them left at this point though as most have jumped on the systemd bandwagon
12:49 < Dagmar> It is not responsible for reminding the sysadmin to get up and have a wee every once in awhile either
12:49 < oiaohm> Dagmar: it works under systemd or BSD init or smf
12:49 < Dagmar> Your insanity is not my problem.
12:49 < oiaohm> Dagmar: It was only a broken problem under sysvinit.
12:49 < drellok> hexnewbie: it turns out I have RAID controller on that server... I am thinking it might be the problem since smartctl says everything is good
12:50 < hexnewbie> Sane init scripts is something I've rarely seen, even if I exclude the horrific nightmares I've written.
12:50 < pingfloyd> oiaohm: that's a problem of piss poor scripts, not with sysvinit
12:51 < pingfloyd> I'll concede that sysvinit gives you more rope to hang yourself with though
12:51 < pingfloyd> but that's unix in general
12:51 < Dagmar> hexnewbie: Someday, you'll get really bored, and write some bulletproof ones just because
12:51 < oiaohm> pingfloyd: really its the sysvinit main package documentation at fault.
12:51 < bazhang> this is about the longest systemd rant to date
12:52 < Dagmar> THe one I was using for restarting mysqld was a bit insane, but then it was on a box that was spontaneously rebooting because of a driver issue
12:52 < oiaohm> pingfloyd: you follow that documentation make broken scripts every single time.    bsd init point to using tty and other things to track services that works better.
12:52 < Dagmar> So, it had to go that "extra mile" and vet the table sanity, repair them it not, or restore the most recent backups if that failed
12:52 < pingfloyd> oiaohm: poor documentation is a bigger problem though
12:53 < Dagmar> ...and absolutely none of this was ever init's responsibility
12:53 < pingfloyd> oiaohm: a very common one unfortunately
12:53 < MrElendig> oiaohm: one of the problem with sysvinit is that every distro ever have slightly different init scripts
12:53 < pingfloyd> with many programs
12:53 < hexnewbie> drellok: Having a RAID controller doesn't mean you have RAID. If you're using a hardware RAID, access to the real disk's SMART is more difficult than just calling smartctl.  You need to use the -d option for your controller model and disk ID, or if smartctl doesn't support - a tool provided by the vendor of the RAID controller.
12:53 < oiaohm> pingfloyd: this case was not poor documentation.   But basically incorrect documentaiton leading to method that cannot work dependable.
12:53 < bazhang> does upstart qualify as a variant
12:53 < drellok> hexnewbie: yea, I am using it like this smartctl -a -d megaraid,9 /dev/sda
12:53 < Dagmar> upstart was a step in the right direction
12:54 < drellok> hexnewbie: disk parameters look good to me
12:54 < pingfloyd> upstart wasn't handled very well
12:54 < MrElendig> good luck having sysv scripts sanely kill process trees btw
12:54 < oiaohm> MrElendig: yes each distribution attempted to design their own fixes to sysvinit without updating main project documentation and advice to have everyone doing the same thing.
12:55 < pingfloyd> but that's the norm for canonical
12:55 < MrElendig> pingfloyd: worst bit was that cannonical basically refused any external contribution
12:55 < pingfloyd> Canonical handles many things very poorly
12:55 < Dagmar> The common existance of hardware environments that might change after boot (like a wireless NIC going up and down), and the possibility that service relevance might hinge on that required something like upstart to happen
12:55 < pingfloyd> MrElendig: yeah, typical of them
12:55 < MrElendig> they just replied with "fork and maintain it downstream"
12:55 < chocolate> I have a server running linux text mode only, and it has a collection of books... I want to have it online to be read in the kindle by the webpage or smartphone connecting by opds for example (everything synced)... anyone know if is there any released solution to manage books like this?
12:55 < hexnewbie> drellok: The dmesg error that just means your (logical or physical) disk went offline, which is the cause of the subsequent segmentation faults. That can be a lot of things - from overheating RAID controller to disk trouble not visible in SMART.
12:55 < MrElendig> (also the CLA)
12:56 < Dagmar> MrElendig: "sysv scripts sanely kill process trees" <-- How is this supposed to be hard?
12:56 < Dagmar> Let's be clear
12:57 < MrElendig> Dagmar: you would have to implement something using cgroups by hand in every init scripts
12:57 < drellok> hexnewbie: ok thanks for your help. I am already opened a ticket with the hosting provider but no response yet, so I thought it is a good idea to ask here meanwhile :)
12:57 < MrElendig> since none of the distroes bothered to make helpers that did it
12:57 < Dagmar> If someone can't reliably manage processes via shell scripts, has no buisness attempting to write init scripts.
12:57 < hexnewbie> drellok: I've seen *visible* damage from overheating on a RAID controller that was half-operating like that before taking it out. Also, I've seen disks with clean SMART grind down to a halt due operations (don't know if the kernel would offline a disk for that, but your RAID controller might)
12:57 < hexnewbie> Also the disk firmware might as well, too.
12:57 < Dagmar> MrElendig: Why?  WOuld I or any other programmer have recently been struck a stout blow to the head?
12:58 < MrElendig> Dagmar: eg if you have something that forks so it is no longer under the parent
12:58 < Dagmar> Who the heck has an init system that _doesn't_ have a pile of library functions that can pull in?
12:58 < MrElendig> you have no idea what the pid is, so you can't terminate it
12:58 < Dagmar> MrElendig: Pop quiz... Who's the daddy?
12:58 < Dagmar> If it forked and exec'd it's on it's own now.  I don't have to care about the pid.
12:58 < Penguin> Dagmar: "its"
12:59 < Dagmar> From the standpoint of the service script, I merely have to be able to correctly identify that daemon process to tell it to go away later
13:00 < Dagmar> ...and let me assure you, pretty much bloody everything forks, because otherwise it's not a daemon
13:01 < Dagmar> Why do you think all these programs are still writing pid files/
13:02 < Dagmar> It's not for collation into powerpoint presentations to middle management later
13:04 < Dagmar> I'm half thinking the nonsense about zombie processes was just a troll
13:05 < oiaohm> Dagmar: really when you come form smf and other placess and you have logs of what services are creating lots of zombies.   Not having that information with times like php under apache can be a pain.
13:06 < oiaohm> Dagmar: same with missing information to locate what services have only half restarted and leaving stuff hanging.
13:08 < oiaohm> Dagmar: basically we need better than what sysvinit was.    Now that does not mean systemd is ideal either.
13:09 < bazhang> oiaohm, I hope to return to a discussion of pulseaudio when I get back
13:10 < pingfloyd> bazhang: hold your breath
13:10 < bazhang> :x
13:10 < jim>  bazhang, you're welcome to start/continue such a discussion :)
13:10 < bazhang> hey jim!
13:11 < oiaohm> bazhang: really with pulseaudio I hope one day we get rid of pulseaudio and jackaudio and finally have single audio server that we can focus on getting right.
13:11 < oiaohm> bazhang: pulseaudio was a improve over the 8 sound servers it go rid of.
13:12 < jim> bazhang, heya :)
13:13 < oiaohm> artsd vs esound vs NAS vs a few other and having applications dependant on them was not particularly fun back before the existance of pulseaudio.
13:14 < hexnewbie> oiaohm: And a regression to those of us who didn't use a sound server. :p (But yeah, I cheered the disappearance of artsd. Ugh.)
13:15 < oiaohm> hexnewbie: put it this way pulseaudio may not be the best.   But its a hell of a lot better than all the older sound servers that died out because of pulseaudio.   Jackaudio was about the only sound server project that stayed alive after pulseaudio.
13:16 < pingfloyd> better to not use a sound server at all
13:16 < pingfloyd> edit asoundrc instead
13:16 < pingfloyd> pulse is buggy af
13:16 < oiaohm> pingfloyd: really alsa still need to fix hotplug problem.
13:16 < hexnewbie> My main concern with PulseAudio these days is that the anointed mixer sucks, and that the audio is still choppy so many years later. Well, not under normal circumstances, but opening pavucontrol causes it to interrupt temporarily.
13:17 < MrElendig> oiaohm: pulse and jack have the opposite focus, trying to do both in one server is probably not going to be easy...
13:17  * MrElendig have never had it chop when opening a mixer
13:18 < pingfloyd> hexnewbie: not to mention the levels have a mind of their own
13:18 < pingfloyd> especially if you log in with more than one user
13:18 < hexnewbie> pingfloyd: Oh, i fixed that by switching to flat volumes, but yeah, that was insane
13:18 < oiaohm> MrElendig: https://pipewire.org/   the attempt is way more evil lets get rid of pulseaudio/jackaudio and fix audio to video sync problem at same time.
13:18 < hexnewbie> Though both PA and ALSA suck from a multi-user perspective.
13:18 < dgurney> I've had nothing but good experiences with pulseaudio, but I know it wasn't always so trouble-free
13:19 < pingfloyd> dgurney: you probably just haven't pushed it
13:19 < MrElendig> most of the issues I've had have been caused by alsa itself
13:19 < rcf> No sound server is great on my BSD systems with modernized OSS, but ALSA is just terrible.
13:19 < MrElendig> (and stupid hardware makers9
13:19 < hexnewbie> PA could at least be used by several users (though odd things happen when it assumes one user is using it), I don't think dmix can.
13:19 < oiaohm> rcf: there is a sound server inside oss
13:19 < pingfloyd> oss is even worse
13:20 < MrElendig> pingfloyd: which of the 8 different implementations?
13:20 < MrElendig> :p
13:20 < MrElendig> can't really speak of oss as one thing anymore
13:20 < pingfloyd> I think of oss as two.  Before OSS sold out, and after.
13:21 < oiaohm> rcf: https://en.wikipedia.org/wiki/Sndio  most bsd systems run this sound server.
13:21 < rcf> FreeBSD’s variant has never bothered me over a decade. ALSA drove me to PA despite its faults.
13:21 < MrElendig> v3, v4, v5, and all the forks of v3 and v4 that are used on the various BDSs that have diverged so much from their base that they are basically seperate pieces of software now
13:21 < pingfloyd> the biggest problem with sound in linux is the architecture (lack of thought out one), which has created an environment of conflict.
13:22 < pingfloyd> like nobody every thought in terms of protocols
13:22 < pingfloyd> and scope
13:23 < oiaohm> pingfloyd: sorry to say there is an architecture to the Linux alsa system its written before we had hotplug sound devices.
13:24 < MrElendig> eh, there were hotpluggable sound devices when alsa was written
13:24 < MrElendig> wasn't as common though
13:24 < oiaohm> pingfloyd: its more that we need to rewrite and redo the complete architecture to be something modern.   Instead of being 90 percent before the existance of hot plug bluetooth and usb sound devices.
13:24 < pingfloyd> we need to define roles and their scope
13:25 < MrElendig> there were usb and firewire sound devices out at the time, that could be hotplugged
13:25 < oiaohm> MrElendig: none of those writing ALSA documentation early had a hot plug sound device.   Yes not common.  And not common enough so alsa was designed without the support.
13:25 < pingfloyd> then sound system related programs won't stomp all over each other.
13:26 < hexnewbie> One would think that a server architecture like X would work for sound, and still, PA works way worse than X despite X being a complete nightmare of legacy features on some levels.
13:27 < oiaohm> hexnewbie: there was extention to X11 for sound.
13:27 < rcf> hexnewbie: how easy is it to run fbdev programs in X? Because that’s what PA would be trying to do.
13:28 < MrElendig> why should it?
13:29 < oiaohm> hexnewbie: X11 protocol has had many strange things like print servers, sound ....   Sound was found very early on to be non workable with X11 protocol and never got out of draft.
13:29 < BluesKaj> some fedora devs were working on a comprehensive media suite for audio and video called pipewire  dunno if it's gonna be successful though. Work on i seems to have stalled.
13:30 < BluesKaj> https://pipewire.org/
13:30 < oiaohm> BluesKaj: pipewire resources at the moment seam to be sucked up by attempt to get desktop capture to work.
13:30 < hexnewbie> Like, a sound server gives you dedicated software mixing in the right layer, a plugin (and feature) architecture that's more user-friendly than ALSA with plugins that can be enabled/disabled in realtime, ssh forwarding (and thus potential network transparency), etc. And it should be possible to do it right, yet PA still manages to do something wrong.
13:31 < pingfloyd> hexnewbie: one sound server to rule them all
13:32 < pingfloyd> hexnewbie: one userland to rule them all
13:32 < pingfloyd> welcome to the redhat coup
13:32 < BluesKaj> oiaohm, that seems too wide a scope for ordinary use
13:33 < oiaohm> BluesKaj: that is why I think pipewire might be doomed.
13:33 < oiaohm> Getting something like sndio that is fast enough that jackaudio can be on top of it and can provide system wide sound mixing would be good.
13:34 < oiaohm> I do understand not wanting to do all sound mixing in kernel space.
13:35 < BluesKaj> oiaohm, let's hope not, but we dos need some thing that's more integrated than alsa nad pulse
13:35 < BluesKaj> nad=and
13:36 < hexnewbie> That sounds bold - get all of jackd and PA into one (so far so good - layering PA on top of jackd kinda gives you that, and some PA features can be implemented in Jack without issue), but adding video on top of that sounds bold.
13:36 < pingfloyd> all you need is a sane standard
13:38 < pingfloyd> don't need another monolithic program
13:39 < hexnewbie> Er, on a second thought, from the little I've used jackd, I don't think there's a PA feature (e.g. hardware mixing and per-app volume controls) that it doesn't do. All it needs is a nice frontend? What am I missing? I guess it can't do network audio over a high-latency network, but there are like two people in the world who use PA like that.
13:41 < BluesKaj> no kidding pingfloyd, gonna be difficult to wipe the slate clean and build a decent media suite from scratch and I doubt that will ever happen
13:41 < oiaohm> hexnewbie: jackaudio has netjack so it can go over network.    Jackaudio interface is quite evil for normal usages.
13:42 < pingfloyd> BluesKaj: imagine if all this work that went into reinventing the wheels with sound servers, went into making a good standard instead.
13:42 < oiaohm> hexnewbie: also jackaudio is not that power wise in design.
13:42 < BluesKaj> jack is waay too complex for  ordinary users
13:42 < hexnewbie> oiaohm: Yeah, but when I tried netjack over a high-latency network, it was unusable, which is when I switched to PA
13:42 < pingfloyd> BluesKaj: then we all could be writing to the standard and leave the sound debacle in the rear view mirror.
13:43 < oiaohm> BluesKaj: really windows and OS X sound stack is fairly much designed to handle the roles jackaudio and pulseaudio handle combinded.
13:43 < oiaohm> BluesKaj: so it do able.
13:44 < pingfloyd> the thing is windows and osx are closed, so they don't have to the same dilemmas
13:45 < BluesKaj> pulse is a necessary evil , that's how i see it , i ha d to drink the pulse kool-ade after my new audio card would do a thing without it altho it's chip was intel
13:45 < BluesKaj> would not
13:46 < oiaohm> pingfloyd: and they do have the management to lock everyone in a room and write a sane standard.
13:46 < pingfloyd> I wouldn't say necessary evil, but it's about the best we got right now.  Which is very sad.
13:46 < MrElendig> if it works with pulse then it works with alsa too
13:46 < BluesKaj> intel used to be able to stand alone even with web audio without pulse
13:47 < oiaohm> MrElendig: there are a few cases of no.   Pulseaudio has a few work arounds at times to get particular audio cards to behave self.
13:47 < pingfloyd> I'd like to think of pulse audio as a stop gap solution
13:47 < BluesKaj> any way I'm repeating myself ...been here done that a hundred times at least
13:47 < MrElendig> which you can backport to alsa itself too
13:47 < oiaohm> The volume of broken hardware out there is insane.
13:48 < pingfloyd> but it's more like the defacto standard because practically everything else available is even worse.
13:48 < BluesKaj> pingfloyd, sad but true
13:49 < oiaohm> esound, artsd, nas for sure were total horrible.   And rememember the first 2 kde and gnome use to mandate that you would use them.
13:49 < oiaohm> or their application would not work.
13:49 < oiaohm> Yes the time of having applications starting with k for kde and applications starting with g for gnome and x for hopefully netural.
13:50 < pingfloyd> back in the days of artsd and esound one of the first things I do on a new install, was get rid of those and configure dmix.
13:50 < oiaohm> Things have got better but they could get a lot better.
13:50 < oiaohm> pingfloyd: I had gnome applications and kde applications that would not use alsa directly.
13:50 < oiaohm> pingfloyd: so no esound or artsd no sound.
13:50 < oiaohm> pingfloyd: and program crashing at odd times.
13:50 < pingfloyd> to me, those would have been good candidates for replacement
13:51 < pingfloyd> that would have been a pretty big deal breaker in choosing alternatives
13:51 < oiaohm> pingfloyd: I wish my work at the time did not require me to use them.
13:51 < pingfloyd> that's always a problem
13:51 < pingfloyd> (work always forcing terrible software standards)
13:51 < oiaohm> This is why I don't  see pulseaudio as that bad.
13:52 < oiaohm> If I am forced to use it the thing is not absolutely nightmare.
13:52 < pingfloyd> I got to a point with IT where I was like "fuck it, let them eat cake"
13:52 < oiaohm> My last remaining cat fight is between applications needing pulseaudio and application needing jackaudio they do at least kind of get along.
13:53 < pingfloyd> it works out good from a dependency standpoint though.  Their terrible choices keep you in business.
13:55 < oiaohm> My most recent audio issue with pulseaudio was TiMidity service deciding to grab audio output while it was playing nothing.
13:55 < oiaohm> Pulseaudio gui does need to get better messages when something else has nicked off with the audio device instead of redirecting into void of nothing.
13:56 < oiaohm> and having you scratching head why do I have no audio.
13:56 < pingfloyd> redhat doesn't want you to understand the messages
13:56 < pingfloyd> they're in the business of selling support
13:56 < pingfloyd> s/selling/extorting/
14:31 < NewbProgrammer10> Hello.
14:37 < sadasaulna> hello
14:40 < NewbProgrammer10> Is it Linux, or GNU/Linux?
14:40 < kvesel> yes
14:40 < sadasaulna> thesedays its more like SystemD/Linux :(
14:40 < NewbProgrammer10> I've heard about this controversy just yesterday.
14:40 < MrElendig> depends on what you are refering to
14:40 < NewbProgrammer10> Linux
14:40 < MrElendig> linux is the kernel, gnu/linux is the operating system
14:40 < MrElendig> problem solved, next!
14:41 < hexnewbie> I simply code for Python/Qt and avoid the controversy.
14:41 < NewbProgrammer10> MrElendig: well, Linux is more than a kernel. It's the opposite of the UNIX philosophy.e
14:41 < sadasaulna> NewbProgrammer10: how so?
14:42 < MrElendig> NewbProgrammer10: no, linux™ is the kernel
14:43 < lukey> Linux is a Registered Trademark of Linus Torvalds. All trademarks are property of their respective owners.
14:44 < NewbProgrammer10> It has character devices like tty and all that bloated crap that a kernel doesn't need (yes, you can disable them in the config, but whatever). Yes, Linux can work without the GNU ecosystem with the exception of the GCC and GPL.
14:44 < hexnewbie> NewbProgrammer10: I'd say ‘UNIX philosophy’ is not a real thing, but an imaginary object to which we like to ascribe attributes that support their arguments.
14:44 < NewbProgrammer10> Over and out about the naming controversy.
14:44 < hexnewbie> s/their/our/
14:44 < NewbProgrammer10> hexnewbie: precisely.
14:44  * MrElendig thinks NewbProgrammer10 should revisit the trolling 101 course
14:45 < sadasaulna> character devices are bloated crap, now i've heard it all
14:45 < sadasaulna> yep his trolling sucks
14:45 < NewbProgrammer10> O_o
14:45 < NewbProgrammer10> That was trolling?
14:45 < NewbProgrammer10> How so?
14:46 < NewbProgrammer10> I'm following the freenode channel guidelines.
14:46 < NewbProgrammer10> If not, then how?
14:46 < sadasaulna> You seem to not understand what a kernel is... and then call character devices "bloated crap".  You're either a troll or clueless
14:46 < hexnewbie> I can do one better. Character devices are not only the span of evil, but should also be renamed to octet devices, because they don't support Unicode, and thus not real characters.
14:46 < kvesel> what are character devices?
14:47 < sadasaulna> hexnewbie: noice! :)
14:51 < lukey> Also Syscalls are the worst !1!!
14:52 < sadasaulna> lol
14:52 < sadasaulna> Windows10 is clearly the best kernel of all, it has the Internet I hear.
14:54 < hexnewbie> All non-file related system calls shall be declared anti-UNIX and replaced with electroshock devices.
14:55 < lukey> exit()
14:55 < lukey> ZAP
14:55 < lukey> Ouch
14:56 < sadasaulna> exit() is a tool of the patriarchy
14:56 < FreeFull> nanosleep? ZAP
14:59 < hexnewbie> Who needs exit()? One should be able to simply do write(open("/proc/self/suicide", O_WRONLY), "\x00", 1)
15:00 < lukey> hexnewbie: open("/dev/mem", O_WRONLY) is the way to go
15:05 < hexnewbie> Also mmap() is very anti-UNIX, probably slipped in by Multics double agents. It should be removed, along malloc() and pointers. All memory access should be handled by opening a new file /proc/self/allocations, and reading and writing to it. Allocation simply happen by writing after the end of the file.
15:17  * Pentode checks hexnewbie's drink to make sure it's safe
15:22 < limbo_> Any suggestions for an ftp server that doesn't require user accounts on the system itself to be added? I don't need any write access over ftp.
15:22 < limbo_> I've tried proftp and pure-ftp, and some others. couldn't grok any of them/didn't work.
15:23 < Pentode> you _really_ just want to use ftp?
15:23 < Pentode> vsftp will allow you to log in using the systems current users. and it's at least kind-of secure
15:24 < twainwek> where are X authoirization listed under `xauth list` stored? i was under the impression that they were under ~/.Xauthority but looks like that file hasn't been updated for a while
15:24 < limbo_> yes. I want to download files from my server to an underpowered device without ssh support.
15:24 < Pentode> +d
15:24 < limbo_> Pentode: I'd prefer to not use system users.
15:24 < limbo_> since it's meant ot be read-only.
15:24 < Pentode> oh i thought thats what you asked, i must've misunderstood
15:25 < Pentode> not sure of anything like that for linux but that doesnt mean it doesnt exist ;)
15:25 < twainwek> and contents of ~/.XAuthority differs than what `xauth list` outputs
15:25 < Pentode> i mean ftp is inherently a user based protocol
15:26 < Pentode> you just want something to transfer data over ethernet using sockets i guess?
15:26 < MrElendig> limbo_: httpss
15:26 < MrElendig> minus one s
15:26 < Pentode> oh yeah, that'll work
15:26 < MrElendig> faster, less painful,  more secure
15:26 < MrElendig> easier to implement
15:32 < sadasaulna> or minus two s if he doesn't care about security and certs
15:32 < MrElendig> no reason not to either self sign and pin or use le
15:32 < sadasaulna> heh, he said underpowered right
15:33 < MrElendig> https is so cheap you can do it on an avr 8 bit micro
15:33 < sadasaulna> and KISS principle... if he were transferring credit card details I doubt he'd be using FTP amirtie?
15:33 < MrElendig> so not going to cause an issue with performance
15:34 < sadasaulna> MrElendig: i was thinking more of hassle
15:34 < sadasaulna> MrElendig: and space
15:34 < MrElendig> doesn't take more space
15:34 < sadasaulna> MrElendig: libraries etc for https surely?
15:34 < MrElendig> going to have those anyway
15:35 < sadasaulna> might not.. don't know what his device is.
15:35 < sadasaulna> i've worked on devices didn't have hardly any libraries at all
15:35 < sadasaulna> like I had to write my own code to do an md5 or sum shit
15:35 < sadasaulna> pun not intended
15:43 < stormland> hey
15:48 < UKCoder> Hi all, I have a 5 raspberry pi machines all running raspbian and I was wondering, is there a way to centrally manage passwords across all machines. Meaning, if I have the same username on each box and want to update the password to that account, that the password is updated across all the nodes in the network. Is this possible?
15:49 < jamesd__> UKCoder: puppet, ldap    to name a couple ways
15:49 < oiaohm> UKCoder: there are many different ways to pull that off.
15:49 < sadasaulna> NIS?
15:50 < jamesd__> Active Directory?  ;-p
15:50 < oiaohm> UKCoder: first question that needs answering is this a configuration update like puppet or central server like ldap/activedirectory solution you looking for.
15:50 < sadasaulna> AD is really just LDAP anyway
15:51 < triceratux> http://www.datacenterknowledge.com/open-source/cloud-east-meets-cloud-west-google-and-tencent-go-linux-foundation-platinum
15:51 < jamesd__> UKCoder: not exactly what you want, but ssh shared keys makes the passwords just go away
15:52 < UKCoder> oiaohm:  I'm agnostic to the implementation, and more looking for what's going to take the least amount of maintenance/setup.
15:52 < sadasaulna> ssh keys then
15:52 < jamesd__> ssh shared keys
15:54 < UKCoder> jamesd__ : is there a way then to distribute/update those keys centrally, rather than create/update keys across all nodes?
15:54 < MrElendig> ipa is also an alternative
15:55 < oiaohm> UKCoder: ssh keys for pushing on the confguration and https://docs.ansible.com/ansible/latest/modules/user_module.html  for setting the configurations.
15:55 < jamesd__> UKCoder: you create a key pair, and distribute to all hosts, and it just basicly works.   one time setup, and it just works.
15:55 < pingfloyd> ssh keys make the most sense if it's just about managing logins
15:56 < pingfloyd> a directory service would be overkill and extra maintenance if that's all you need
15:56 < pingfloyd> and screw AD, this is linux
15:57 < oiaohm> pingfloyd:  multi replication between samba ADs I have seen used by It is in class over kill in most cases.
15:57 < UKCoder> oiaohm : thanks, I'll take a look.  jamesd__ : would I need to create key pairs for each combination of hosts I'd want to log into? e.g. key pair for RP1->RP2, RP2->RP1,RP1->RP3,etc.?
15:57 < pingfloyd> okay, but why taint your network with AD if you don't have to
15:58 < jamesd__> UKCoder: you can use one key set for all your hosts...  it will make things easier.. but you can setup one pair each cluster of hosts or one to one if you are paranoid
15:59 < UKCoder> jamesd__ : thanks :)
16:00 < oiaohm> UKCoder: notice you can push out a ssh login key to users with ansible.      So system you run update from you could have a pair with every machine with a root level login.   Yet have a general user with a shared key pushed out by ansible.
16:00 < jamesd__> shared keys is what use at work,  a cluster of 3000 hosts, all with the same key.. Connect to bastion host with two-factor authenication and then go to any host no passwords needed.
16:00 < oiaohm> UKCoder: think about configuration edit route is that ansible is only running when you are pushing out a change.
16:01 < pingfloyd> for paranoia factor I would just use your account with the keys very judiciously.  For example use a different account for surfing.
16:01 < UKCoder> oiaohm : does ansible have to be installed/configured on each node, or just the one node that you're authoring/distributing the key pair from?
16:02 < oiaohm> UKCoder: ansible connect to machine by ssh then sends commands over ssh to get everything set.
16:02 < oiaohm> UKCoder: basically instead of having to configure each machine indivually you get to use a template.
16:02 < revel> oiaohm: Why ssh and not telnet?
16:02 < oiaohm> revel: really???
16:03 < revel> wot
16:03 < oiaohm> revel: how many distributions have telnet server any more.
16:03 < jamesd__> simple deploy is   ssh-keygen ; for i in host1 host2 host3 host4......  ; do   ssh-copy-id $i  ; done   ... put your passwd in  your copy-paste buffer and  paste once per host done.
16:03 < revel> Dunno. Most of them, probably?
16:03 < oiaohm> revel: almost none.
16:04 < revel> Got any examples?
16:04 < revel> Gentoo and Debian do.
16:04 < jamesd__> most have it.. but not installed by default, unless they have totally removed it from the repos, seriously haven't checked
16:04 < oiaohm> revel: packet sniffering software taking telnet password over network because it plain text has kind of seen telnet as option on Linux disappear.
16:04 < revel> I know why telnet is bad, but as far as I know, it's still in most repos.
16:05 < revel> (telnet daemons, even, not just the client)
16:05 < MrElendig> just because it exists doesn't mean that you should use it
16:05 < sadasaulna> its not even bad if you don't care about the security
16:05 < kvesel> telnet irc.freenode.net 6667
16:05 < UKCoder> jamesd__ : woah, that looks super straight forward :)
16:05 < MrElendig> kvesel: s/irc/chat/
16:05 < pingfloyd> depends how you're using it
16:05 < pingfloyd> I would only use it for anonymous access
16:06 < jamesd__> UKCoder: if you want one host to all others that is complete... if you want all hosts to any host you need to copy your private host to each of the others.
16:06 < sadasaulna> pingfloyd: not that its worth bothering, but if its just hosts in your own wired home network, security on the wire hardly matters
16:06 < sadasaulna> unless your wife likes to use a packet sniffer lol
16:07 < MrElendig> your iot devices do
16:07 < pingfloyd> as long as you keep rogues off your home network
16:07 < pingfloyd> problem is there is sometimes surprises
16:07 < sadasaulna> I wouldn't have any IOT devices on the same VLAN as my home network, fuck that
16:08 < pingfloyd> like you didn't know about the idiot that decided to attach an AP to it, that didn't know what they were doing.
16:08 < sadasaulna> pingfloyd: if you have rogues on your home network, you got bigger problems!
16:08 < pingfloyd> of course.  You should be preventing that to start with, but shit inevitably happens.
16:09 < jamesd__> i worked for a company that had 350 active users all telneting to a single unix server... running a legacy cobol app, sad to say that was in this decade.
16:09 < sadasaulna> frankly, if they've gotten the access they can watch my packets on my LAN, then i'm already pwned AFAIC
16:09 < pingfloyd> jamesd__: that still goes on
16:10 < pingfloyd> sadasaulna: being on the lan and being able to access hosts aren't one in the same
16:10 < oiaohm> revel: https://github.com/marado/netkit-telnet/tree/master/netkit-telnet-0.17  <<  As such, their use for remote logins is discouraged.  <So the ones with telnetd the documentation tells you not to use telnet in most cases.  Also most telnetd servers are no longer in active mantaince.   So even worse security risk.
16:10 < pingfloyd> being on the lan is getting a foot in the door
16:10 < pingfloyd> doesn't mean the vault has been cracked yet
16:10 < kvesel> :verne.freenode.net 404 telnetusr ##linux :Cannot send to channel
16:10 < kvesel> :(
16:11 < pingfloyd> now if you just leave the money laying out, they don't have to bother
16:11 < revel> oiaohm: I know. This doesn't really back up your claim that no distros have it any more though.
16:11 < jamesd__> pingfloyd: not where I am...  the auditors stat there bloomers,  protip, if someone says FEDRAMP and your salary,  run, hide, jump ship, get the hell out, and find a new job.. thankfully i'm hourly with OT eligilble.
16:11 < sadasaulna> pingfloyd: i understand that, but i'm comfortable I don't have any such people on my LAN and therefore not paranoid about plaintext on the LAN
16:11 < CapTn_MoRgun> sup yo
16:12 < sadasaulna> I use SSH anyway, but I would hardly be scared of letting a password go in plaintext over my own LAN because i'm confident enough that it isn't compromised
16:12 < CapTn_MoRgun> sheshechh
16:12 < pingfloyd> sadasaulna: I wouldn't use telnet unless you have to.  And if you have to, you have to think carefully about things.
16:12 < CapTn_MoRgun> dude telnet is like so 50's
16:12 < pingfloyd> sending authentication in clear text is always a bad idea regardless of environment
16:12 < CapTn_MoRgun> post war protocol
16:12 < oiaohm> revel: sorry I wrote it bady.   Most distributions have ssh server in the default installers.    these use to have telnetd in the installers.   So the ablity to find it already installed sshd is way more likely.
16:12 < CapTn_MoRgun> used on ze bombe
16:13 < sadasaulna> pingfloyd: i agree, I was just pointing out that its not ALWAYS bad.
16:13 < revel> Yeah, telnetd by default is unheard of.
16:13 < pingfloyd> sadasaulna: for MUDs you usually have to use it, or some mud client which is just as bad.
16:13 < oiaohm> revel: except for windows server
16:13 < revel> Yeah...
16:13 < sadasaulna> pingfloyd: and you take the sensible precaution of using passwords you care less about for stuff like that
16:13 < pingfloyd> but your mud account getting compromised isn't the end of the world.  It's really the host of the mud that should be worried.
16:14 < pingfloyd> sadasaulna: you should never reuse passwords
16:15 < sadasaulna> pingfloyd: I reuse them all the time, but in grades, from "couldn't give a shit" to "only use this for one special account"
16:15 < sadasaulna> so for instance my password for identifying this nick is in the grade "couldn't give a shit"
16:15 < oiaohm> revel: to be correct common versions of windows server no telnet server.   But cluster versions of windows server its really old but yes you find telnet server on by default.  hopefully soon come sane ssh.
16:15 < kvesel> passwords should be no less than 20 character sin length, contains mixed case letters, numbers, special characters, unicode characters, and never be used more than once and memorised
16:16 < sadasaulna> lol kvesel
16:16 < aksis> i have a computer which has suddenly started to regularly (once per day) become completely unresponsive. the monitor also stops working. i set up kdumps and sysrq-trigger works, but my crash kernel never gets loaded when the computer becomes unresponsive. ideas?
16:16 < oiaohm> kvesel: come on nothing is more fun than giving a person 2048 char password.
16:16 < oiaohm> kvesel: and tell them they can change it latter.
16:16 < kvesel> Я
16:17 < jamesd__> password management needs to be intergrated with OS and APPs at a more fine tuned level, the browser and OS should handle 90% of it by default, and should offer a way to get free two factor using a phone or tablet so passwords just becomes something that happens in the background.
16:17 < kvesel> от7fw%V^E%^vvvbЦС;%
16:17 < sadasaulna> kvesel: hope you memorised that
16:17 < kvesel> i did
16:17 < kvesel> but i have to forget it because you know it now
16:17 < sadasaulna> lmao
16:18 < kvesel> :)
16:18 < sadasaulna> no, all I see is *************** here
16:18 < sadasaulna> irc has a special feature that protects passwords
16:18 < kvesel> oh okay im safe then
16:18 < sadasaulna> try it by pasting yours
16:18 < sadasaulna> :)
16:18 < jamesd__> eibcccdhnlblhikjdukebhnbucgjcrghciubfcdlgrij
16:18 < sadasaulna> yep just stars here... its a great feature of IRC ;)
16:19 < kvesel> F4ncy.BeaЯ$%76454v545
16:19 < jamesd__> oh crap there goes mine.
16:19 < kvesel> oh shit can you see that?
16:19 < sadasaulna> kvesel: nah, just all stars lol
16:19 < kvesel> oh okay im good then
16:19 < oiaohm> kvesel: windows stalls at 127 chars for a password.    Linux/many unix systems unlimited lenght allows you to be evil.
16:19 < jamesd__> 5 points that identify the method i used to generate my password above.
16:20 < kvesel> well after a certain amount you will just get collisions on unix
16:20 < kvesel> i think like aes/sha256 you get a decent length?
16:20 < jamesd__> passwords good,  pass phrase better.
16:20 < kvesel> salted
16:20 < oiaohm> kvesel: that is how you generate a huge long MD5 password that is in fact really short.
16:20 < kvesel> /dev/random has the best passwords
16:21 < jamesd__> echo password | md5sum
16:21 < sadasaulna> jamesd__: yeah i like phrases, and you can mix in you own junk according to some memorised rule
16:21 < infinisil> # echo hi
16:21 < infinisil> Aw, no shell bot
16:21 < sadasaulna> # pkill infinisil
16:26  * jamesd__ wonders if a linux kernel 1.0.9 based distro would still boot on a modern CPU.. things that you wonder. 
16:26 < dw1> how can i add a single configure parameter after already being configured
16:26 < dw1> without changing what's already configured
16:27 < sadasaulna> jamesd__: shower-thoughts :)
16:27 < oiaohm> To be correct you can set a 256 password on windows machine.   Just one catch you will not be able to use that with any windows password dialog.   So a linux system running samba or rdesktop on Linux can login those running windows totally stuffed over.
16:27 < infinisil> dw1: Gotta be a bit more specific, I have no idea what you're talking about
16:28 < dw1> basically i downloaded the ubuntu source for curl and want to add --with-brotli to the configure
16:28 < dw1> without affecting anything else
16:29 < oiaohm> jamesd__: thinking you can get freedos to still work on modern cpu badly it might work.   But drivers would be a pure nightmare.
16:29 < jamesd__> dwl you can install your version on top of the one that comes with the OS, but its better to install at  /usr/local/ so you can go back to the OS provided version if needed.
16:30 < dw1> problem is i dont know what all paramaters they used and i want to keep it the same
16:31 < sadasaulna> there might be a way of extracting the build options from the package manager I forget now though
16:31 < jamesd__> dwl track it down... perhaps grab the source package from the distrobution, see if they documented there  configure parameters.
16:32 < dw1> yeah, i have their source, but i'm a configure/make noob :/
16:32 < RayTracer> dw1: I think it should be in debian/rules or a file besides it
16:32 < jamesd__> most distro's just use the defaults, since its just one of thousands of packages they need to maintain.
16:33 < dw1> CONFIGURE_ARGS is seen in debian/rules - nice
16:34 < dannylee> 8-)
16:34 < UKCoder> jamesd__ : thanks for the help before, keys now distributed across all nodes, super simple :)
16:35 < jamesd__> UKCoder: your welcome.
16:37 < pingfloyd> sadasaulna: use a password manager
16:37 < pingfloyd> sadasaulna: then you just have to remember one good one
16:38 < pingfloyd> sadasaulna: everything else you can use random generated ones
16:40 < sadasaulna> sadasaulna: I do use a password manager.  I don't use random passwords on everything though.
16:40 < pingfloyd> some sites make it hard to use randomly generated ones
16:40 < pingfloyd> because they don't allow many characters
16:40 < pingfloyd> webdevs are so dumb
16:40 < Li> one of the reasons that should me think to become rich is to pay for gnome, kde UX people. maybe they will produce something useful eventually .. who knows
16:41 < sadasaulna> I prefer memorable but strong passwords.
16:41 < pingfloyd> Li: would be great if that happens
16:41 < Li> anyway I will not be rich so you can forget it alerady
16:41 < Li> lol
16:41 < pingfloyd> there's a shortage of good UI development
16:42 < Li> yeah because it is free
16:42 < jamesd__> i used to like KDE back in the 3.0 days, but then i got lazy and gave up on installing it, and just use the basics of what ever I have at hand. and 99% of my life is command line anyway.
16:42 < pingfloyd> well if you do, there's a good cause to support
16:42 < FreeFull> My first DE was KDE 3.5
16:42 < FreeFull> Then I ended up using XFCE 4 for a really long time
16:43 < iodev> there is none like KDE 3.5
16:43 < iodev> Trinity desktop is still beta
16:43 < FreeFull> Nowadays I use i3wm
16:43 < pingfloyd> it's stupid how DEs keep reinventing themselves
16:43 < Li> FreeFull: my first DE was mistery and it will always be. simply because I didn't what a de is at that time
16:44 < pingfloyd> like gnome has progressively gotten worse in functionality and performance
16:44 < pingfloyd> you'd think if the performance got worse, it would get better
16:44 < jamesd__> I just installed lubuntu on my work laptop, got tired of compiz eating up so many cpus cycles and needing to give the vm  4GB of ram... my windows laptop just starts to crawl, even though I have 16GB of ram on a i5 with HT.
16:44 < FreeFull> Li: You don't remember whate it looked like?
16:44 < pingfloyd> like gnome1 was highly configurable
16:44 < iodev> FreeFull: I had i3wm
16:44 < pingfloyd> and very fast
16:44 < iodev> but it's not okay for modern apps
16:44 < Li> pingfloyd: don't remind about gnome .. it makes me feel like wanna through myself from a tower
16:45 < iodev> I tried using stacking WM and DEs
16:45 < FreeFull> iodev: I haven't been having trouble with i3wm and "modern apps"
16:45 < pingfloyd> kde is more of less the same way
16:45 < Li> FreeFull: I don't remember when I first used linux
16:45 < sadasaulna> pingfloyd: its funny when you run very old DEs and you're like "damn these are better than the modern ones"
16:45 < iodev> FreeFull: in the end I use Win 10 and WSL
16:45 < sadasaulna> even Win98 i'm like "this wasn't half bad"
16:45 < Li> probably was mandrake back in the day
16:45 < jamesd__> I got better results with Solaris 8 on a 32MB of ram and a 70mhz cpu, than i do with gnome with 2x cores and 4GB of ram...
16:45 < pingfloyd> kde 3.5 is better than 4
16:45 < pingfloyd> much better to use
16:45 < NoriusNotorius> Has anyone had a frustrating time going back to a DE after using i3 for a while? I was given a Macbook Pro for work and it has been a frustrating experience when it comes to window management
16:46 < pingfloyd> with 4 it's extra eye candy and effects
16:46 < Li> that is the only single worth invetion came out from France
16:46 < FreeFull> iodev: Did you know there is a tiling WM for Windows?
16:46 < Li> aside from metric measurments
16:46 < pingfloyd> I'd rather have functionality than stupid animations
16:46 < iodev> FreeFull: no thanks, I don't want tilling anymore
16:46 < pingfloyd> FreeFull: remember tabworks?
16:47 < pingfloyd> iodev: same here
16:47 < FreeFull> pingfloyd: No
16:47 < pingfloyd> iodev: it's so overrated
16:47 < sadasaulna> i've been playing with i3 but using twm right this minute.
16:47 < jamesd__> tilinf doesn't  sound bad if you have 2x large monitors  27" and a 29" monitor here.. fighting to keep the right windows in place and orginized gets to be a pain.
16:47 < pingfloyd> forces the user into its work flow, instead of the other way around
16:48 < Li> what other programs to open doc files without losing organization other than openoffice ad libreoffice
16:48 < pingfloyd> sadasaulna: i3 makes your keyboard combos longer
16:48 < pingfloyd> sadasaulna: kind of like tmux and screen do
16:48 < NoriusNotorius> jamesd__: in i3, your workspaces are like having multiple monitors
16:48 < Li> librewrite is a big big mess
16:48 < pingfloyd> extra keystrokes to do the same things that normally take less
16:49 < pingfloyd> i3 was one of those things that's good in theory
16:49 < FreeFull> pingfloyd: What do you mean?
16:49 < NoriusNotorius> no need for multiple monitors really
16:49  * jamesd__ wonders why did they make a window manager with the same name as a current intel processor, i guess they like confusing people
16:49 < d3nk> Hello all, I want to upgrade my ram. I'm having difficulty with understanding the clock speeds. this is my onboard memory at the moment: https://ptpb.pw/u_3U
16:50 < d3nk> I would appreciate help on deciding the clock speed I should go for
16:51 < pingfloyd> jamesd__: it's a play off words.  i3 is a continuation of wmii
16:51 < jamesd__> d3nk: something 2400mhz or faster.. but honestly it usually doesn't matter as long as your system supports, and your not a gamer trying to scrape out every cycle from everything.
16:51 < pingfloyd> jamesd__: i3 is basically what wmiii would be
16:51 < kvesel> .
16:51 < sadasaulna> so any i3 fans here.. ?
16:52 < d3nk> jamesd__: nope not  a gamer at all. s i should go for 2400 or above?
16:52 < jamesd__> i was a windowmaker fan... but remember i got lazy.
16:52 < Pentode> i like my windows stacked.
16:52 < FreeFull> To be fair, I only use i3 in tabbed mode, mostly
16:52 < jamesd__> d3nk: yeap just make sure your motherboard or system supports it
16:52 < Li> has anyone tried Calligra Suite before?
16:53 < FreeFull> And use workspaces to switch between programs
16:53 < d3nk> jamesd__: how do i do that?
16:53 < pingfloyd> FreeFull: it's floating mode is garbage
16:53 < NoriusNotorius> FreeFull: same here. Never got used to stacked mode. Feel like it's less efficient regarding real estate
16:54 < Pentode> d3nk, just buy 2400mhz ddr4 ram. no worries.
16:54 < jamesd__> d3nk: google motherboard  or system model number and the memory model number, you usually get results talking  about it being supported, or  search your system model number and memory in amazon.
16:54 < pingfloyd> most gui programmed are designed with the assumption of there being a stacking wm
16:54 < pingfloyd> so you also get to fight with that
16:54 < d3nk> ah thanks :)
16:54 < pingfloyd> s/programmed/programs/
16:55 < FreeFull> pingfloyd: It's annoying that floating windows are always on top, but otherwise it's fine
16:55 < pingfloyd> FreeFull: not just annoying.  Severely flawed.
16:55 < pingfloyd> (big UI flaw)
16:55 < pingfloyd> a real basic and no brainer one at that
16:56 < pingfloyd> that's the problem with projects like that, they become myopic
16:56 < jamesd__> yeah with two large monitors it gets annoying to see a dialog box pop up on the other monitor, and you sit wondering why did the application just lock up until you see the dialog box 5 minutes later ;-p
16:57 < strive> jamesd__: I've experienced that as well, lol
16:57 < pingfloyd> i.e., with i3, they're obsessed with the tree paradigm at the detriment to it.
16:57 < Li> pingfloyd: I can make huge UX differece if I just know an easy way to suggest changes without a big huss&fuss
16:58 < pingfloyd> Li: suggesting is fine, but most things never happen until someone rolls up their sleeves and implements them
16:58 < pingfloyd> Li: especially in FOSS
16:59 < pingfloyd> many things are something that many wouldn't mind, just nobody ended up implementing it (coding the needed changes).
16:59 < jamesd__> Li it might cool to work with tiling with the requirement that all dialog boxes pop up in a space defined by the user, so the user can just automaticly look at the location defined. like name one monitor primary and set a box that all dialog boxes pop up in.
16:59 < Li> wich brings us back to good reaso why you people should donate for me to become rich
16:59 < Li> I promise fixig it
16:59 < pingfloyd> the nice thing though, is if anyone wants something changed, nobody can stop them thanks to foss licenses.
17:00 < kvesel> whats wrong with terminals?
17:00 < pingfloyd> that's the other way you can get change--paying others to implement
17:01 < sadasaulna> Li, there is a prince in Nigeria I can introduce you too, he has a lot of cash
17:01 < jamesd__> kvesel: are you talking about a physical vt102 serial attached terminal, or terminal software  like  gterm, kterm, or my current favorite  terminator
17:01 < sadasaulna> vt52 for the win, at 300baud
17:02 < NoriusNotorius> pingfloyd: you probably won't get the same level of work though. In comparison to someone who will make the change because they want to
17:02 < kvesel> yes
17:02 < Li> sadasaulna: I'm not faggot if that is what you're reffering to
17:02 < pingfloyd> I can read faster than 300 baud
17:02 < sadasaulna> Li: no that was not what I was referring to
17:02 < jamesd__> i sold 1200 baud modems to a few highschool teachers back in the day $235 each.
17:03 < Li> but you are free to make money of your prince boyfried then donate for FOSS
17:03 < sadasaulna> nah he's just a rich prince likes giving his money away
17:03 < Li> sadasaulna: it sounded very much like what you're refercing to
17:03 < sadasaulna> *whoosh*
17:04 < pingfloyd> yeah, it did sound like prostitution
17:04 < pingfloyd> Li: c'mon, take one for the team!
17:04 < sadasaulna> yeah, its for the good of FOSS
17:04 < Li> sadasaulna: ok go get it and make donation, just make sure he put condom on while he is like giving it to you
17:05 < sadasaulna> He only likes Chinese... not interested in me sadly
17:05 < Li> pingfloyd: I don't team with faggots :))
17:06 < sadasaulna> he likes his teaming and bonding purely at the network level
17:08 < Li> Hmmm how did you know that?
17:09 < jack_rip_vim> Aha! prince~
17:13 < ShrikeX> hi
17:13 < jack_rip_vim> hi ShrikeX
17:13 < ShrikeX> Hi Folks I'm trying to run elastic search as a service in redhat 7.1 but I'm getting some java errors
17:13 < ShrikeX> Im not super familiar with alternatives
17:14 < jack_rip_vim> ShrikeX: what is the error?
17:14 < ShrikeX> Im getting the following error
17:14 < ShrikeX>  elasticsearch[2011]: which: no java in (/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin)
17:14 < ShrikeX> but my $JAVA_HOME is definitely set
17:15 < jack_rip_vim> ShrikeX: check /etc/profiles, see if you put the path behind it
17:15 < ShrikeX> ok
17:15 < ShrikeX> as root I'm assuming?
17:15 < jack_rip_vim> ShrikeX: yes
17:15 < Li> ShrikeX: what is the value of JAVA_HOME?
17:16 < ShrikeX> /usr/lib/jvm/jre1.8.0_161
17:16 < Li> are you on debian based linux?
17:17 < jack_rip_vim> Li redhat
17:17 < ShrikeX> Redhat 7.1
17:17 < ShrikeX> at the bottom of my /etc/profile
17:17 < ShrikeX> unset -f pathmunge JAVA_HOME=/usr/lib/jvm/jre1.8.0_161 PATH=$PATH:$HOME/bin:$JAVA_HOME/bin JRE_HOME=/usr/local/java/jre1.8.0_161 PATH=$PATH:$HOME/bin:$JRE_HOME/bin
17:17 < Li> what do you get when types javac -version?
17:18 < ShrikeX> -bash: javac: command not found
17:18 < Li> you need to add /usr/lib/jvm/jre1.8.0_161/bin to path as well
17:18 < ShrikeX> ok Ill give that a shot
17:19 < ShrikeX> in my /etc/profile ?
17:19 < Li> nope ... just put PATH=$PATH:/usr/lib/jvm/jre1.8.0_161/bin
17:21 < Li> anyone knows the reason for office2003-schemas on debian?
17:21 < ShrikeX> same error
17:21 < ShrikeX>  elasticsearch[2128]: which: no java in (/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin)
17:21 < Li> what do you get whe you echo $PATH
17:22 < jack_rip_vim> ShrikeX: put the path behind the PATH value, source it
17:22 < ShrikeX> /usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/root/bin:/usr/lib/jvm/jre1.8.0_161/bin:/root/bin:/usr/local/java/jre1.8.0_161/bin:/root/bin:/usr/lib/jvm/jre1.8.0_161/bin
17:22 < ShrikeX> god what a mess
17:22 < ShrikeX> I didnt configure this
17:22 < ShrikeX> just installing software
17:22 < jack_rip_vim> ShrikeX: lol
17:23 < ShrikeX> i just want this stupid service to start on server start
17:24 < jamesd__> ShrikeX: once you start down the java, and 3rd party apps, it gets worse, you haven't experienced hell untill you have done ps -ef and see a single pid using 5 lines on a full screen terminal session.
17:24 < ShrikeX> works fine if I run it manually
17:24 < jack_rip_vim> ShrikeX: delete the export JAVA_HOME, put the path behind the PATH value
17:24 < ShrikeX> I OH I KNOW
17:24 < ShrikeX> jack_rip_vim: from where?
17:24 < ShrikeX> from /etc/profile ?
17:24 < jack_rip_vim> ShrikeX: yes
17:24 < ShrikeX> ok sec
17:24 < Li> ShrikeX: I don't know why do you have duplications within your path values
17:25 < ShrikeX> yeah either do I
17:25 < Li> anyway try to reboot
17:25 < ShrikeX> this is some vmware image
17:25 < Li> javac --version should work
17:25 < ShrikeX> ok
17:25 < ShrikeX> so
17:25 < ShrikeX> delete JAVA_HOME=/usr/lib/jvm/jre1.8.0_161 from /etc/profile
17:26 < jack_rip_vim> ShrikeX: you may want to try yum install java or yum install openjdk
17:26 < jack_rip_vim> ShrikeX: yes
17:26 < Li> try which javac
17:26 < ShrikeX> no outside internet access
17:26 < ShrikeX> ok deleting
17:26 < ShrikeX> now
17:26 < ShrikeX> what was the bit about path
17:26 < ShrikeX> this is whats left
17:26 < ShrikeX> PATH=$PATH:$HOME/bin:$JAVA_HOME/bin PATH=$PATH:/usr/lib/jvm/jre1.8.0_161/bin JRE_HOME=/usr/local/java/jre1.8.0_161 PATH=$PATH:$HOME/bin:$JRE_HOME/bin
17:27 < jack_rip_vim> ShrikeX: delete it
17:27 < jack_rip_vim> delete those things
17:27 < ShrikeX> delete the ENTIRE THING?
17:27 < ShrikeX> all three lines
17:27 < jack_rip_vim> PATH=$PATH:/usr/lib/jvm/jre1.8.0_161/bin
17:27 < jack_rip_vim>                  JRE_HOME=/usr/local/java/jre1.8.0_161
17:27 < jack_rip_vim>                  PATH=$PATH:$HOME/bin:$JRE_HOME/bin
17:27 < jack_rip_vim> 11:27 < jack_rip_vim> ShrikeX: delete it
17:27 < ShrikeX> ok
17:27 < jack_rip_vim> yes
17:28 < ShrikeX> ok
17:28 < ShrikeX> the only thing left
17:28 < ShrikeX> is
17:28 < ShrikeX> export PATH
17:28 < ShrikeX> at the very end of profile
17:28 < jack_rip_vim> ShrikeX: after you delete it, put /usr/lib/jvm/jre1.8.0_161/bin behind PATH value
17:29 < jack_rip_vim> ShrikeX: do you already put this path behind it?
17:29 < jamesd__> being a geek that pays the eletric bill, means whether or not you buy another computer has little to do with the cost of system, and more about the impact to your electic bill.
17:29 < ShrikeX> no
17:29 < ShrikeX> im only sort of following
17:29 < jack_rip_vim> ShrikeX: put on it
17:29 < Li> do you redhat people have somethig similar to update-alteratives on debian?
17:29 < ShrikeX> do I want do
17:29 < Li> ShrikeX: if so I suggest you try from there
17:30 < ShrikeX> PATH=/usr/lib/jvm/jre1.8.0_161/bin:/usr/lib/jvm/jre1.8.0_161/bin ?
17:30 < ShrikeX> is that what I want?
17:30 < ShrikeX> followed by export PATH
17:30 < ShrikeX> ur shit
17:31 < jack_rip_vim> PATH=/usr/lib/jvm/jre1.8.0_161/bin:/usr/bin:/bin
17:31 < ShrikeX> PATH=/usr/lib/jvm/jre1.8.0_161/bin:$PATH:$HOME/bin
17:31 < jack_rip_vim> ShrikeX:
17:31 < ShrikeX> ok
17:32 < Li> ShrikeX: what is $HOME/bin ?
17:32 < ShrikeX> who knows
17:32 < ShrikeX> it was just hanging out there
17:32 < jack_rip_vim> ShrikeX: after finished, source /etc/profiles
17:32 < jack_rip_vim> ShrikeX: then java -version, see if work
17:33 < ShrikeX> source /etc/profile
17:33 < jack_rip_vim> source /etc/profiles
17:33 < ShrikeX> java -version java version "1.8.0_161" Java(TM) SE Runtime Environment (build 1.8.0_161-b12) Java HotSpot(TM) 64-Bit Server VM (build 25.161-b12, mixed mode)
17:33 < ShrikeX> i dont have a profiles
17:33 < ShrikeX> just profile
17:33 < jack_rip_vim> ShrikeX: OK
17:33 < jack_rip_vim> Work
17:34 < ShrikeX> should I try to reboot?
17:34 < jack_rip_vim> ShrikeX: javac -version,
17:34 < ShrikeX> ok
17:34 < jack_rip_vim> ShrikeX: no need
17:34 < ShrikeX> javac -version -bash: javac: command not found
17:35 < Li> the god damn new wireless keyboard is fucked already, fuck china production
17:35 < Li> good by hp
17:35 < jack_rip_vim> ShrikeX: javac should be in another dir
17:36 < ShrikeX> i need that in my path as well?
17:36 < pingfloyd> what are you doing to your keyboards?
17:36 < ShrikeX> or I need to run this from a different directory?
17:36 < jack_rip_vim> ShrikeX: you need to put that path on PATH too
17:36 < Li> pingfloyd: banging it
17:36 < pingfloyd> you're not to supposed to bang them
17:36 < ShrikeX> what directory is javac usually in ? :)
17:37 < pingfloyd> could only get away with that in the 80s
17:37 < Li> who's supposed to bang them then? you?
17:37 < jack_rip_vim> PATH=/usr/lib/jvm/jre1.8.0_161/bin:/usr/lib/jvm/bin:/usr/bin:/bin
17:38 < jack_rip_vim> ShrikeX: see if this can work
17:38 < sadasaulna> ive been typing wrong all these years?
17:38 < ShrikeX> ok
17:38 < ShrikeX> im wondering if I only have the JRE and not the JDK?
17:38 < jack_rip_vim> ShrikeX: jvm dir is the java dir?
17:38 < pingfloyd> ShrikeX: don't you want $HOME/bin before your other paths?
17:39 < ShrikeX> i took it out for now pingfloyd
17:39 < jack_rip_vim> pingfloyd: I use /usr/bin:/bin replace it
17:39 < ShrikeX> so for now
17:39 < ShrikeX>  which java /usr/lib/jvm/jre1.8.0_161/bin/java
17:39 < jack_rip_vim> to*
17:39 < jack_rip_vim> ShrikeX: ?
17:40 < ShrikeX> what did you want me to try?
17:40 < ghosalmartin> is it possible for the linux kernel to panic too fast for console-ramoops to catch it?
17:40 < jack_rip_vim> ShrikeX: tell me, where you unpack the java package?
17:40 < ShrikeX> good question
17:41 < Li> ShrikeX: try which javac
17:41 < ShrikeX> i dont know
17:41 < jack_rip_vim> lol
17:41 < ShrikeX> which javac /usr/bin/which: no javac in (/usr/lib/jvm/jre1.8.0_161/bin:/usr/bin:/bin)
17:41 < pingfloyd> did it just mysteriously appear?
17:41 < ShrikeX> some blasted vmware wizard creating this image
17:41 < ShrikeX> created
17:42 < pingfloyd> wizard?!
17:42 < ShrikeX> not a real wizard
17:42 < pingfloyd> sounds like Windows
17:42 < ShrikeX> no
17:42 < ShrikeX> its vmware :(
17:42 < jack_rip_vim> ShrikeX: cd /usr/lib/jvm, ls it, then show me the print
17:42 < ShrikeX> I am an AIX yoinker
17:42 < Li> ShrikeX: be wiser than the wizard and go ahead install jdk
17:43 < ShrikeX> jvm]# ls java-1.7.0-openjdk-1.7.0.161-2.6.12.0.el7_4.x86_64  jre  jre-1.7.0  jre-1.7.0-openjdk  jre-1.7.0-openjdk-1.7.0.161-2.6.12.0.el7_4.x86_64  jre1.8.0_161  jre-openjdk
17:43 < ShrikeX> looks like the 1.8 jdk is not there
17:43 < pingfloyd> wizard computing is dogshit
17:43 < bipul> Java virtual machine is a VM?
17:43 < ShrikeX> no my redhat machine is a vm
17:44 < ShrikeX> its literally vmware
17:44 < jack_rip_vim> ShrikeX: cd /usr/lib/jvm/jre1.8.0_161, ls it, then show me the print
17:44 < Li> Setting the default JDK with the /usr/sbin/alternatives Utility https://access.redhat.com/documentation/en-US/JBoss_Enterprise_Web_Platform/5/html/Installation_Guide/sect-use_alternatives_to_set_default_JDK.html
17:44 < bipul> Which version of redhat are you using ShrikeX ?
17:44 < ShrikeX> jre1.8.0_161]# ls bin  COPYRIGHT  lib  LICENSE  man  plugin  README  release  THIRDPARTYLICENSEREADME-JAVAFX.txt  THIRDPARTYLICENSEREADME.txt  Welcome.html
17:44 < ShrikeX> 7.1
17:44 < pingfloyd> the wizard installs jdk, but it's not ready to use?
17:44 < jack_rip_vim> ShrikeX: +_+, this is just a jre version....
17:44 < bipul> Aha, the latest one..
17:45 < Li> ShrikeX: the above link was meant for you
17:45 < ShrikeX> so I need to install the jdk?
17:45 < jack_rip_vim> ShrikeX: yes
17:45 < ShrikeX> thank you li, i will continue to read up on alternatives
17:45 < bipul> So what are you trying to do ShrikeX ?
17:45 < ShrikeX> start a service
17:45 < ShrikeX> but it can't find java
17:45 < ShrikeX> but its because I only have the jre
17:45 < ShrikeX> and not the jdk
17:45 < bipul> What services?
17:45 < ShrikeX> apparently
17:46 < pingfloyd> a Sunday service?
17:46 < ShrikeX> elastic search
17:46 < jack_rip_vim> redhat
17:46 < bipul> You want to run elasticsearch at the backend during booting?
17:46 < pingfloyd> red from all of the systems they've murdered
17:46 < ShrikeX> i just want it to start automagically
17:46 < pingfloyd> never trust a red-hatted wizard
17:47 < ShrikeX> I can start it manualy
17:47 < TheWild> hello
17:47 < TheWild> could ftp command f' up line endings despite claiming binary file transfer?
17:47 < pingfloyd> then you need to find the service file or make one
17:47 < ShrikeX> i have one
17:47 < ShrikeX> but
17:47 < ShrikeX> it fails
17:47 < ShrikeX> because it can't find java
17:47 < ShrikeX> because Im missing the jdk
17:48 < ShrikeX> :)
17:48 < ShrikeX> I will go get the jdk
17:48 < ShrikeX> thank you for your help
17:48 < ShrikeX> everyone
17:48 < pingfloyd> so systemctl enable foo.service
17:48 < jack_rip_vim> ShrikeX: download the jdk, unpack the dir to some paths, then put the [the path you put on]/bin behind the PATH in /etc/profile, source it again, you will get java
17:48 < bipul> ShrikeX, Just inform that to redhat fellows here #redhat
17:48 < jack_rip_vim> command
17:48 < Li> update-alternatives --all --verbose
17:49 < pingfloyd> he said he can start it manually
17:49 < bipul> wb prussian
17:49 < pingfloyd> this is looking to be an xy-problem
17:49 < birdbolt1> hi, im trying to figure out how to copy the contents of a file thats in a docker container
17:50 < d3nk> Hello, I still have a bit of confusion. The onboard ram on my laptop is " 4GiB SODIMM DDR4 Synchronous 2400 MHz" so when I decide to add a new stick I need to get a SODIMM ddr4 right? And can I go at a lower clock speed like 2133mhz? Thanks again :)
17:50 < jack_rip_vim> Aha! find an old tor version tor79~
17:51 < birdbolt1> `docker exec -it my-nginx-service nano /etc/nginx/nginx.conf | echo newfile
17:51 < birdbolt1> it created a newfile but didnt copy the contents of the file out of the vm
17:51 < birdbolt1> thats echo > newfile
17:52 < searedvandal> d3nk, probably a question for #hardware , but yes, you would need another stick of ddr4. and you can put in a slower speed stick, but the other stick will then clock down and run at that speed also.
17:52 < seveneleven> this saves minimum 30 minutes work https://github.com/Nyr/openvpn-install \o/
17:53 < d3nk> seveneleven: so my current stick will be clocked at 2133mhz too? Oh #hardware Oops
17:53 < jamesd__> birdbolt1: you got it backwards  | sends stdot to the next command  you possibly   want nano  file-to-edit  < contents  ... or you can just cp it from a shared volume or pull in via  wget or curl from a webserver
17:53 < searedvandal> d3nk, yes
17:55 < birdbolt1> jamesd__, damn i was hoping to avoid setting up a shared volume. Thanks though, ill go with that
17:55 < jamesd__> birdbolt1: shared volume or a webserver,  memcached.. with microservices in play there are so many ways to do things like that.
17:57 < bipul> Where does docker being used?
17:57 < jamesd__> bipul: huh?
18:00 < bipul> jamesd__, I'm wondering which companies uses docker as there infrastructure?
18:01 < searedvandal> bipul, closest answer you'll get is probably https://www.docker.com/#/world_leading_companies
18:01 < jamesd__> bipul: a lot of start up do, google  who uses  kubernetes   for a good start,  kubernetes is one of the leadingorgistration tools for docker
18:03 < searedvandal> unless companies actually say what they use, or docker phones home statistics, you're never gonna find a complete list.
18:03 < searedvandal> it's like asking which companies use linux as their infrastructure
18:04 < sadasaulna> hipsters use it.  So if the company is full of beards they probably use it
18:04 < searedvandal> that's as a good of a data point as anything else
18:04 < ShrikeX> hey li ! and jack_rip_vim
18:04 < ShrikeX> javac -version javac 1.8.0_172
18:04 < ShrikeX> yay
18:04 < jamesd__> ans much like personal computers in the 80's... people use docker without ever telling there company they are using it...  great for dev environments, even if you don't get the full benefit of docker everywhere
18:04 < bipul> searedvandal, Ah i see... Hope i will get a job where i have to work on dockers
18:04 < jamesd__> er s/ans/and/
18:05 < sadasaulna> bipul: you may come to regret that statement
18:05 < bipul> sadasaulna, why?
18:05 < searedvandal> bipul, chances are that you're gonna encounter some type of containerization in the wild when job hunting.
18:06 < sadasaulna> lets just say the world of devops is "fast moving"
18:06 < jamesd__> if you want to live docker, get ready high levels of indirection, and trying out how things work, and in somce cases its just blind trust that things will work right.
18:06 < bipul> Yes, that's true..
18:06 < sadasaulna> jamesd__: +1
18:07 < bipul> indirection?
18:07 < sadasaulna> you get a lot of devs who don't know shit about sysadmin copy-pasta everything into a nightmare
18:07 < bipul> docker = chroot + name space
18:07 < jamesd__> i think i spent like 5 hours researching how one simple thing in docker works,  persisent storage with ranch longbow  volumes... finally found what i needed in a bug report.
18:07 < searedvandal> those using docker eventually end up dockering everything. such a waste
18:09 < sadasaulna> they should rename docker "koolaid", dont drink too much of it
18:09 < jamesd__> searedvandal: but its so powerful, especially if you want to play with a new tool/service docker has pre made containers that just work for the simple stuff... some even work for complex stuff... want to deploy elasticsearch on 6 nodes... you can do it with docker and ranch in an afternoon.
18:10 < sadasaulna> gotta go, been nice chatting with all
18:10 < jamesd__> have a great day sadasaulna
18:10 < sadasaulna> cheers
18:10 < searedvandal> sure, it has a lot of good use cases, but people tend to overdo it. use it when not needed, just because they can.
18:13 < bipul> have great day sadasaulna
18:14 < kvesel> garden
18:26 < sanroot> Is it possible to install two operating system without partition on same disk
18:26 < hypercore> anyone use riot here?
18:26 < prussian> sanroot: yes.
18:27 < sanroot> How prussian can you explain
18:27 < sanroot> prussian i only know about KVM
18:27 < prussian> boot loader entry for whatever 2 -> configure pid 1 to chroot into whatever root fs on the same partition
18:28 < prussian> having two different "OSes" sharing the same root fs might be a bit much though
18:28 < sanroot> Hmmm
18:28 < sanroot> Prussian we can cinf bootloader to use chroot ?
18:29 < sanroot> Its conf not cinf
18:30 < prussian> i don't know. maybe you can pass a cmdline that makes whatever init/pid 1 do it for you
18:31 < sacules> Anybody using polybar with dwm? I have this issue when I start it: https://imgur.com/a/sxePS6l
18:38 < seveneleven> hi
18:49 < sanroot> Is here anyone using windows subsystem linux ,what are its limitations ?
18:51 < Mistell> sanroot: I use it for a number of things. It's solid, sometimes x-based apps work but I wouldn't count on it
18:52 < Mistell> Docker/LXC still don't work too well though :(
18:52 < Mistell> WSL is great for most other dev environments though
18:53 < sanroot> Cool..microsoft done good job ,currenty i updated to win 10 and testing its feature
18:53 < Mistell> I agree, I like it a good bit.
18:54 < pnbeast> I haven't touched it.  Two of the researchers at work who've tried it say that it does simple things, but it doesn't work for anything beyond that - I heard "clunky" and one or two obvious bugs, like something crashed, maybe.  And it's also not really on topic in this channel.
18:59 < jamesd__> if you want to see what happens when you take docker to its extreme on the desktop  https://github.com/jessfraz/dockerfiles   X apps with multimedia in docker.
19:00 < micrex22> pnbeast plus windows 10 will want to start in the middle of things or explorer.exe will stop working interrupting what you're doing with WSL and WIN32 :p
19:00 < micrex22> restart*
19:01 < pnbeast> Don't get me started on the randomly timed reboots.  I've have more users crying about that for the past few months than anything, and we use a fairly locked down "enterprise" version of 10.  What a POS...
19:06 < locrian9> I'm running a two linux guests of 'Arch Linux' over my Windows 10 host.  On the second Arch guest theirs some weird mouse connectivity issues, where sometimes I can boot the Arch guest and get good mouse click and highlight, and later on the clicking and highlighting goes away.  On the first Arch guest, I don't have any of these problems.  Is there a way to troubleshoot this?
19:07 < locrian9> Virtual box VM
19:13 < Dagmar> Yes, you could try a more stable distribution and see if the problem goes away
19:15 < ayecee> that's your answer to every problem
19:15 < ayecee> involving arch
19:15 < sauvin> That could have something to do with it being... well... arch!
19:16 < pnbeast> ARCH FIGHT!  ARCH FIGHT!
19:17 < mnrmnaugh> arch is fine. blame upstream
19:18 < pnbeast> LINUX FIGHT!  LINUX FIGHT!
19:18 < kvesel> slackware best ware
19:18 < mnrmnaugh> lol there was goodun the other day. /usr/bin/pinentry.pacnew O YE LETS JUST PUT BINARIES IN THE THINGY=()
19:18 < mnrmnaugh> slackware is bestware indeed
19:20 < azarus> slackware hah best last paradigm ware
19:24 < searedvandal> arch is the stablest
19:24 < azarus> "arch is the stablest" and other funny jokes you can tell yourself
19:25  * azarus stops
19:29 < micrex22> azarus I'd be more inclined to blame Windows 10 than a Linux distirbution....
19:29 < paulcarroty> arch is the stablest - joke of the year
19:30 < azarus> micrex22: blaming Windows for Linux instability -- genius
19:30 < Dagmar> micrex22: They did this inside VirtualBox
19:30 < micrex22> virtualbox is *terrible*
19:30 < Dagmar> micrex22: ...and apparently they managed to mess up the second install.
19:30 < azarus> what alternative hypervisor is there on windows anyway
19:30 < azarus> hyper-v?
19:30 < strive> vmware
19:30  * jamesd__ has seen unix systems with higher uptimes, than some people in here ;-p
19:30 < micrex22> Hyerp-V is baked in to windows 10 enterprise
19:30 < azarus> virtualbox is opensource at least
19:31 < micrex22> virtualbox has very poor acceleration
19:31 < azarus> it's OK
19:31 < micrex22> not for client-based stuff
19:31 < micrex22> it'd be okay for a headless OS
19:31 < strive> jamesd__: So far my personal server is at 15 days.
19:31 < azarus> micrex22: used it for a while, found no real issues
19:31 < azarus> i do use hyper-v too, but I don't like it that much
19:32 < MrElendig> long uptime just means more unpatches bugs/holes
19:32 < MrElendig> unpatched
19:32 < azarus> if I could, I'd use QEMU/HAX on windows, but it's so buggy
19:33 < Dagmar> ...and yet it works fine under Linux
19:33 < azarus> Dagmar: well on Linux it's QEMU/KVM, and that works very well
19:33 < azarus> and I use that often
19:33 < jamesd__> home fileserver, personal box with the higest uptime n40l:~$ uptime 12:32pm  up 512 day(s), 13:35,  1 user,  load average: 0.04, 0.20, 0.27
19:34 < micrex22> azarus well it's the graphical acceleration on virtuablox--it's extremely slow for any graphical work. Plus it does not support copyng files between guest and host
19:34 < pnbeast> jamesd__, can you give us the IP addr and which ports are open on that box?
19:34 < azarus> micrex22: it does do shared folders
19:34 < strive> jamesd__: Good and stable...or is it?
19:34 < micrex22> it can only copy text between guest and host
19:34 < micrex22> out of the box
19:34 < jamesd__> micrex22: virtualbox can share  mount a folder from the host os in the vm
19:34 < micrex22> well that's not the same as copying
19:34 < jamesd__> strive: yes, totally, but doesn't run linux.
19:34 < micrex22> :)
19:35 < azarus> micrex22: i copied tar.gz archives in virtualbox, worked fine
19:35 < azarus> from guest to host and back
19:35 < strive> I rarely get updates on CentOS.
19:35 < micrex22> virtuabox does not support copying files from the host and guest--you'd have to spin up a network share or something
19:35 < jamesd__> my D: drive on my windows 7 laptop, is 100% accessible in my ubtunu vm
19:35 < azarus> micrex22: no, you don't, it just works, you just need the virtualbox guest modules
19:37 < jamesd__> strive: are you running centos 5 or older?
19:37 < strive> Older.
19:37 < azarus> long, long time EOL
19:37 < jamesd__> thats becuse its EOL, no new patches unless its really dire, update to 6 or 7 and you will have security patches monthly
19:38 < strive> Woops. Misread that.
19:38 < strive> It's CentOS 7
19:41 < azarus> i'll try QEMU/HAX soon to see if it's less buggy now
19:41 < sensibel> Which is best for memory logging in linux, is this a nice option https://github.com/f18m/CPU-MEM-monitor?
19:42 < WhiteDevil> iam using QEMU
19:42 < WhiteDevil> it works great
19:42 < WhiteDevil> no problem so far
19:42 < azarus> WhiteDevil: which qemu
19:43 < azarus> on which OS, with which guest?
19:43 < WhiteDevil> debian
19:43 < WhiteDevil> linxu and windows
19:43 < azarus> on linux it's fine
19:43 < azarus> i'm talking about QEMU on windows, with a linux guest
19:43 < WhiteDevil> ahh
19:43 < WhiteDevil> it might not be well integrated ?
19:43 < azarus> that's what QEMU/HAX is
19:43 < azarus> kinda, yup
19:43 < WhiteDevil> ahh i see
19:45 < WhiteDevil> i thought vmware was a good option if virtualbox dosnt work out for windows
19:45 < azarus> eh, I prefer opensource software
19:45 < azarus> and vmware is really not a good company
19:46 < WhiteDevil> when i was on windows
19:46 < WhiteDevil> vmware was the shizz
19:46 < ironus> I have never really worked with it.. Is it good? Or you might as well use KVM?
19:46 < azarus> ironus: on windows, you have no KVM
19:47 < WhiteDevil> on windows vmware gets the job done
19:47 < ironus> I am aware.
19:47 < azarus> ironus: in my experience, some bootloaders freak out under QEMU/HAX
19:47 < azarus> syslinux and isolinux for example
19:47 < strive> azarus: Not true. You couldrun KVM through cygwin.
19:47 < ironus> I don't know if there is any truth to this but I have noticed a pattern with vmware vm's  not holding up under heavy IO
19:47 < azarus> strive: nah. KVM is kernel-side
19:48 < azarus> KVM cannot be run in cygwin
19:48 < strive> Ok.
19:48 < strive> My mistake, I meant virt-manager.
19:48 < azarus> *sigh* sure
19:48 < strive> That was a  misread.
19:48 < strive> I'll go take a nap now.
19:48 < strive> :/
19:49 < azarus> goodnight!
19:49 < searedvandal> sensibel, perf mem
19:51 < jamiejackson> how do i search a file for the first instance of this regex and return only the subexpression (the group in parens)? salt="([^"])*"
19:53 < sensibel> I want to increase the space for root using lvm, is this possible?
19:53 < Alabaster> hello all I have a question if I may I am doing a simple apt-get upgrade but everytime I hit a button by accident it seems to stop the proecess
19:56 < WhiteDevil> everytime you hit a button by accident
19:56 < WhiteDevil> how many times did you create the blunder ?
19:56 < Psi-Jack> sensibel: If you have available space in the VG, yes.
19:56 < Alabaster> me?
19:57 < Psi-Jack> Yes, you...
19:57 < Alabaster> oof now its showing a lot of carrots with differet letters
19:57 < WhiteDevil> we can safely say alabaster is retarded
19:57 < Alabaster> a good couple of times by looking at this. All I did was start typing letters on the wrong keyboard
19:57 < Psi-Jack> That's not very nice.
19:57 < Alabaster> YES YES you can
19:58 < Alabaster> I use ubuntu. New to Kali
19:58 < Psi-Jack> Kali is not a desktop distro.
19:58 < Alabaster> None the less I will agree I am retarded
19:58 < WhiteDevil> you want to be a hacker
19:58 < lnnb> i disagree!
19:58 < WhiteDevil> using kali
19:58 < WhiteDevil> break all the locks ..open the world for free information ?
19:59 < Alabaster> no following CCNA and CCNP and security videos and tuts
19:59 < WhiteDevil> Anonymous
19:59 < Alabaster> I wish
19:59 < Alabaster> no not anon
19:59 < WhiteDevil> we need free informatin
19:59 < Psi-Jack> CCNA and CCNP has literally nothing to do with Kali at all.
19:59 < Alabaster> i'm to old for that noise
19:59 < lnnb> aren't those fools just a govt front?
19:59 < ironus> jamiejackson: What are you trying to match? That reads to me as matching anything that's not a quotation mark ?
19:59 < Alabaster> It does if you want to mess around with your own network and yes
19:59 < sensibel> Psi-Jack: how?
20:00 < sensibel> Psi-Jack: VG means?
20:00 < Alabaster> in fact I think Kali is waaaaay to known to do anything stupid on
20:00 < kvesel> OSCP is for Kali
20:00 < Alabaster> either way. I plan on getting security certs later but I am noob so I am starting here
20:01 < Psi-Jack> LVM has 3 things about it. PV (Physical Volume), VG (Volume Group), LV (Logical Volume)
20:01 < Alabaster> either way it started unpacking and setting up things again. I guess it was just taking FOREVER
20:01 < Psi-Jack> Alabaster: Here, regarding Kali, you will mostly be shunned.
20:01 < jamiejackson> ironus, i got the asterisk wrong, but this means match what's in the quotes: salt="([^"]*)"  so if the file had salt="foo", it would return foo
20:01 < Alabaster> The first time I made this mistake I was so dumb I rebooted
20:02 < Alabaster> Really, I mean I guess it has a bad rap. But I am literally studying for certs
20:02 < Alabaster> but I get its stigma.
20:03 < searedvandal> no stigma, just not a desktop distro. in your case, if you need it for certs, fully valid use case. but those who install it and use it as main OS, I don't get.
20:03 < ironus> jamiejackson: what kind of characters are you potentially matching against? just letters and numbers?
20:03 < Alabaster> but yes the use of wireshark and other reading of legal traffic in Books and Vids and tuts ask you to follow along with them. Although they never tell you how THEY got there
20:04 < jamiejackson> ironus, yes. it's not the regex i need help with, though, as i've used this kind of thing lots. i need to know what command/syntax to use to extract that value out of a file
20:04 < sensibel> searedvandal:  so we give the desired command with perf and it records the memory usage like time command is used ?
20:05 < Alabaster> it just seems to automatically work and they say "lets look at this and the layers and remember all the things inside !
20:05 < ironus> jamiejackson: using grep?
20:06 < jamiejackson> whatever will do it. not sure whether to look at sed, grep, perl, or whatever
20:06 < jamiejackson> ^ ironus
20:09 < bls> it looks like you're parsing some kind of markup, so you probably want a tool that understands/can parse that markup instead of writing your own parser using regexps
20:11 < markasoftware> I'm writing a makefile for a front-end web project. Sometimes I want to build it in "development" mode, and sometimes "production" mode. In production mode, it does additional, slow, work, like minifying my JS files. In both modes, the output JS and CSS files should be written to the same spot
20:11 < markasoftware> but, if i do `make dev` and `make prod` one after another, it does not think anything has changed, and does not rebuild
20:12 < lnnb> the target files need to be cleaned up
20:12 < markasoftware> ah, can i somehow set the `clean` target as a prerequisite?
20:12 < searedvandal> sensibel, for that I think psrecord is a bit nicer and more human readable
20:12 < mawk> makefile for frontend markasoftware ?
20:12 < mawk> what is this witchcraft again
20:13 < sensibel> searedvandal: so it will show the memory usage throughout the command is running ?
20:13 < kriztmark> double o linux
20:13 < markasoftware> all the frontend tools have a CLI, i don't need to use gulp
20:13 < searedvandal> sensibel, https://pypi.org/project/psrecord/
20:13 < Alabaster> so since I was a moron and rebooted while apt-get was running I did a dp whatever the command it is to check the upgrade and force it to resume. How do I check to make sure it upgrade didn't miss anything. I can redo it but at this rate for some reason its been taking a whole day to upgrade this Linux
20:14 < mawk> running in which part Alabaster ?
20:14 < mawk> downloading or installing
20:14 < mawk> if it was downloading, just redo the command and it will resume
20:14 < Alabaster> it was unpacking
20:14 < mawk> ah
20:14 < kriztmark> apt-get delete package name
20:14 < mawk> just redo it
20:15 < kriztmark> then apt-get install packagename
20:15 < bls> if you want to clean before building a target, you either need to automate it via sentinels or explicitly call it yourself
20:15 < Alabaster> oh man
20:15 < Alabaster> so I have to wait a whole day again
20:15 < mawk> it won't download again
20:15 < mawk> don't worry
20:15 < Alabaster> Its not a package its an apt-upgrade btw
20:16 < mawk> yeah same thing
20:16 < azarus> Alabaster: are you like in a country with only dialup?
20:16 < Alabaster> no it's probably unpacking so slow because I am using a USB 2.0 stick running live
20:16 < azarus> well install it then
20:17 < bls> why are you applying updates to a live distro?
20:17 < Alabaster> I didn't know more than the usual apt-get update/upgrade basic stuff
20:17 < Alabaster> I have it set to persistence
20:17 < Alabaster> and my wifi card isn't working
20:18 < Alabaster> so I am making sure the first step of updating and upgrading is done
20:18 < Alabaster> the drivers wont take on the wifi adapter because failed or missing kernel headers for it
20:18 < azarus> why is it that there are so many people trotting into here with newbie kali questions
20:19 < azarus> Alabaster: also no, the reason why wifi isn't working is never missing kernel headers
20:19 < Pentode> azarus, use your imagination. ;)
20:19 < Alabaster> because if you live in a major city security jobs that start at 50k is a plenty
20:19 < Pentode> more likely missing / wrong firmware than anything else
20:19 < ExoUNX> afternoon
20:19 < bls> what made you pick this mode of operation? most people either make a normal install onto USB, or completely update the live distro ISO instead of trying to pack updates into the persistence, leaving the persistence just for data files
20:19 < ExoUNX> has anyone tried wireguard?
20:20 < Alabaster> I've found a couple solutions online about this adaptors problem with linux they all show the same package that has worked for them
20:20 < azarus> ExoUNX: yes
20:20 < azarus> i use it daily
20:20 < Alabaster> after updating and upgrading I can install the package
20:20 < ExoUNX> azarus, does it keep a pretty stable and persistent connection?
20:20 < triceratux> Alabaster: go the parrotsec way. only run it live. dont do dist-upgrade https://www.linux.com/learn/security/parrot-security-could-be-your-next-security-tool
20:21 < azarus> ExoUNX: yes, it supports roaming on both sides
20:21 < Alabaster> THANK you triceratux I hear about that. Great things
20:21 < jamiejackson> i'd prefer something more ubiquitous than gawk, but this does seem to work, ironus : https://stackoverflow.com/a/14085682/1026263
20:22 < ExoUNX> azarus, also, does it deal with multipoint well?
20:22 < Alabaster> I just don't know anything myself like how / where to even get it or exactly what it fully is
20:22 < bls> more ubiquitous than awk? it's part of POSXI
20:22 < azarus> ExoUNX: https://www.wireguard.com/#built-in-roaming
20:22 < azarus> dunno what multipoint is/does
20:22 < Alabaster> triceratux is parrotsec an OS or something for my linux distro?
20:23 < azarus> bls: gawk isn't quite POSIX is it, it has extensions
20:24 < bls> yes, but if you're not using any of those extensions, you can just call awk
20:24 < triceratux> Alabaster: its a security oriented os like kali but they dont make the same kinds of mistakes https://distrowatch.com/table.php?distribution=parrotsecurity
20:24 < azarus> Alabaster: http://lmgtfy.com/?q=parrotsec
20:24 < azarus> bls: yup
20:24 < _0x40_> Hi. My system will randomly freeze when doing gpu-intensive stuff. I can reproduce it somewhat consistently by changing the zoom level too fast in Firefox. How would I go about troubleshooting this?
20:25 < Alabaster> thanks guys I will look it up. Hopefully it will support my wifi adaptor
20:26 < jim> Alabaster, do you have the machine there?
20:27 < bls> and when you want to upgrade a live distro, upgrade the entire ISO so your persistent overlay doesn't get bogged down and essentially become your entire OS
20:27 < Alabaster> jim which machine. I have a PC and laptop
20:28 < jim> Alabaster, do they both have wireless net cards?
20:28 < Pusteblume> i searched for rkhunter.conf and it listed it as result but it also gave the error file not found: http://paste.debian.net/1031474/
20:28 < Alabaster> I just bought a new one. I'm maining my laptop
20:29 < Pusteblume> that gives me a strange feeling in my stomach. how do i proceed?
20:29 < Alabaster> it has a crummy intel but for learning security I just received a new usb adaptor
20:30 < jim> _0x40_, I dunno... I'd start by ruling out hardware problems, run a ram checker for a couple hours
20:30 < Alabaster> with the obvious two necessities
20:30 < bls> is rkunter.conf a broken symlink?
20:30 < ironus> jamiejackson: grep -Po '(?<=salt=\")(\w+)' Will that work?
20:30 < ironus> Took me way longer than it should have.
20:30 < _0x40_> jim: I'll try that. Thanks!
20:31 < Pusteblume> bls, its a text file
20:31 < jim> Alabaster, so, just the laptop has both wireless cards?
20:31 < searedvandal> _0x40_, what gpu are you using? check that you have correct and updated drivers and stuff too
20:31 < Alabaster> seems for Kal a lot of people have had a lot of problems getting it to show up or turn on even if they get pass the no kernel headers or driver issues
20:31 < bls> if he's worried about awk not being uqbiquitous enough, depending on PCRE is probably out of the question
20:31 < Alabaster> The laptop has one inside already internal intel
20:33 < jim> Alabaster, which os is it running?
20:33 < _0x40_> searedvandal: Nvidia GTX 760 with an up-to-date Nouveau driver.
20:33 < jim> (and which os do you want to run, I assume you want to install something?)
20:33 < bls> printf '    chili="abced"\n    salt="fghig"\n    pepper="lmno"\n' | sed -n 's/.*salt="\([^"]\{1,\}\)".*/\1/p'
20:34 < searedvandal> _0x40_, alright. haven't had any issues with nouveau myself, that's seems the most stable to me. but run a ram checker and stuff like that which jim mentioned earlier. could be a hardware issue.
20:35 < Alabaster> jim no no, I mean i am upgrading Kal live at the moment but thats is only because the new adaptor I have isn't working with it
20:35 < Alabaster> It worked fine in windows as a test
20:35 < jim> Alabaster, the new adapter being the usb?
20:36 < jim> kali?
20:36 < Alabaster> but they had many probs years back getting even stable Kal support than they added it and than somehow now on newer versions its back to kernel header and driver issues
20:36 < Alabaster> yeah, if you been reading up I am doing CCNP and Security certs this summer
20:37 < jamiejackson> ironus, i think it will, thanks. seems to require gnu grep, which is too bad, but it's more prevalent than gawk.
20:37 < bls> jamiejackson: prue POSIX: printf '    chili="abced"\n    salt="fghig"\n    pepper="lmno"\n' | sed -n 's/.*salt="\([^"]\{1,\}\)".*/\1/p'
20:37 < Alabaster> security is just added bonus to me but such an easy to get and well the Cisco videos and books have it in there now because they want you to learn the internal of your network and traffic and whats inside to see the big picture
20:38 < jim> Alabaster, ok... what kind of adapter (maybe specific model number?) is the laptop's internal card?
20:38 < Alabaster> as ONE GREAT man in here said earlier is abandon that Kal right now and jump on Parrot
20:39 < jamiejackson> works, bls, thanks. i'm now going to mentally parse that
20:39 < Alabaster> I might just do that but I am more knowledgeable about networking and R and S than security
20:39 < Alabaster> jim the laptops internal is an Intel AC 7265
20:40 < bls> and if you need *only* the first, things will get uglier still
20:40 < Alabaster> my new USB is a Realtek RTL8812au I believe
20:40 < jim> oh right, you're running linux now? (kali I guess?)
20:41 < ironus> jamiejackson: If you know a language using their regex implementation may be best  . It would be a lot easier for me to do in python for instance
20:41 < Alabaster> no I am on my PC connected to the TV in the same room while apt-upgrade is finsihing on Kali which seems like it soon might be done because it is unpacking SET
20:42 < jamiejackson> bls, it would be better just to get the first, but i may be okay with it as is. one thing though, is that i can have empty quotes, so i simplified the repetition: sed -n 's/.*salt="\([^"]*\)".*/\1/p'
20:42 < zapotah> realtek and linux, very brave. also very foolish and risky :3
20:42 < Alabaster> it took all of a night to upgrade for some odd reason
20:42 < Alabaster> zapotah why?
20:42 < zapotah> Alabaster: you cannot imagine the amount of firmware and _hardware_ bugs baked into those chips
20:43 < jamiejackson> ironus, yeah, i'm much better at this stuff in a full fledged language's script.
20:43 < zapotah> Alabaster: some had a hard_wired_ size for mtu for instance
20:43 < Alabaster> and or tell me in a chat if you want. Either way I'm following along with CCNA and CCNP and security vids and books, and as long as I don't make a mistake it should be my own network.
20:43 < zapotah> not coded
20:43 < jim> Alabaster, why do you say the internal is crappy? what are some of the symptoms?
20:44 < zapotah> _wired_, the chip was physically set to a fixed size mtu
20:44 < Alabaster> jim to learn security and you want to deconstruct your own personal network you need a couple of things for the adaptor to support
20:45 < Alabaster> or even if another company allows you to help check their security / leakage
20:45 < jim> I have a wireless card that does the same bands, it kept losing connection with the AP
20:45 < searedvandal> just because the internal doesn't support monitor mode and packet injection, doesn't mean its crappy
20:45 < zapotah> this was pretty much me when i figured out what was going on with the chip: https://i.imgflip.com/slxyb.jpg
20:45 < Alabaster> zapotah, they may heat up or drop out but is their a security risk?
20:45 < jim> I didn't know how to debug it, but turns out I didn't have to:
20:46 < zapotah> Alabaster: there is a functionality risk, it might do weird AF stuff
20:46 < Alabaster> seardvandal, I was just being silly
20:46 < searedvandal> :)
20:46 < bls>  sed -n -e :b -e N -e 's/.*salt="\([^"]*\)".*/\1/p' -e 't q' -e 'b b' -e :q -e q
20:46 < bls> first only
20:47 < Alabaster> @zapotah HAHAHAHA
20:47 < Alabaster> thats funny
20:47 < bls> hmmm, that's not optimal
20:47 < jim> at one point I wanted to upgrade my net to gigabit, and eventually I replaced the WAP, with an asus that supports gigabit speeds at least on the ethernet ports...
20:47 < Alabaster> my friend who uses a different brand/ same chipset for everyday use says his drops out every other day
20:48 < bls> cleaner: sed -n -e 's/.*salt="\([^"]*\)".*/\1/p' -e 't q' -e b -e :q -e q
20:48 < jim> my connection problems went away when I started using the new access point
20:48 < Alabaster> hopefully not the other AP causing the problems
20:48 < ironus> bls: I was able to keep up with you up until the cleaner version haha
20:49 < jamiejackson> bls O_o thanks. that's complicated looking
20:49 < jim> Alabaster, but in your case, it's because your wireless doesn't  support certain things you need for your studies?
20:49 < bls> ironus: it's sed's terse version of goto
20:49 < bls> so yeah
20:49 < Alabaster> zapotah as long as it doesnt make itself present while scanning scanning deep is all I am worried about
20:49 < jim> Alabaster, in my case the wap was pretty old
20:50 < Alabaster> I ain't doing nothing illegal but at the same time I don't wanna be seen as suspicious although I do live in a MAJOR city with tons of other nerds testing and learning
20:50 < ironus> bls: do you have any recommended resources for sed? I can do basic stuff but nothing advanced.
20:51 < fscale> so yesterday I came here asking which distro supports 125% display scaling & one person said gnome, then I asked which distro came with gnome, person said fedora. Now I'm within a fedora live image & I don't see no 125% scaling option. :(
20:51 < fscale> https://imgur.com/a/0uVsyUP
20:51 < Alabaster> fedora
20:52 < Alabaster> almost all come with Gnome
20:52 < triceratux> folks say a lot of stuff in here ;)
20:52 < Alabaster> the GUI kde whatever
20:52 < Alabaster> yeah
20:52 < jim> fscale, it's not the distro that supports that scaling (or not), it's the display software (like, the X server)
20:52 < bls> I learned regexp from the perl camel book and perl best practices, picked up most of the sed from studying the sed-1-liners text file that's been around for years
20:52 < Alabaster> I mean Ubuntu should be done with their stupid unity experiment and the other api
20:53 < bls> http://sed.sourceforge.net/sed1line.txt
20:54 < searedvandal> fscale, check the tweaks tool and the windows option. should be a hidpi thing at the bottom of that
20:54 < bls> that and vi
20:54 < Alabaster> dpkg warning: unable to delete old directory "" is that right during an upgrade or just needs to update the program and leave the directory in place???
20:54 < jim> yes, of course the x server is in the dist, but what you want to find out is whether the X server it comes with supports it (or, which part of the X server would support it specifically), and whether or not it actually does
20:55 < fscale> searedvandal: Ok I'll see what I can get from google with the gnome tweaks tool search option first.
20:55 < ironus> bls: Thanks. Studying that page along with the manual will likely do the trick
20:56 < jim> Alabaster, well probably it's that when it removed the parts that come with the package, the dir still wasn't empty, so it issues that warning to let you know. it's not fatal at all
20:56 < fscale> searedvandal: Is this what you're referring? https://www.maketecheasier.com/enable-fractional-scaling-gnome/
20:57 < Alabaster> thanks. I am just losing it because I am on  day two of this upgrade
20:57 < Alabaster> hehehehehe
20:58 < searedvandal> fscale, that was gonna be my next point, fractional scaling in gnome is still experimental as far as I know
20:58 < searedvandal> so I don't think fedora lets you scale 125% without doing what that article say fscale
20:58 < fscale> searedvandal: That's better than being non existant like in other distros. Choices are very limited.
20:59 < fscale> *DE..sorry
20:59 < jim> Alabaster, dpkg (the program that handles installing and removing .deb packages) noticed files in there it didn't expect... likely, you may have dropped some files in that dir?
21:00 < jim> (dpkg doesn't just go around removing stuff if it notices stuff like that)
21:00 < Alabaster> no I haven't used anything yet. I just restarted it to update/upgrade and test the solutions I have found to get the adaptor to work
21:01 < jim> (another thing dpkg won't do, is automatically upgrade  config files it thinks you changed)
21:03 < jim> so generally, dpkg/apt is careful to preserve changes you have made
21:05 < Elladan> fscale, there are two forms of scaling. One is that you ask the X server to scale the whole screen, which is essentially the same as just switching to a lower resolution and living with the ugly fonts.
21:05 < blinksy> newbieQ: how do i change dns servers while still keeping dhcp on cli?
21:06 < fscale> Elladan: And the other?
21:06 < Elladan> fscale, the second form (which is what you want) is where you ask every application to draw itself differently, so it scales up the bitmap graphics etc. while rendering the fonts larger.
21:06 < fscale> Yes that's what I want.
21:07 < Elladan> fscale, that mode is primarily a feature of the desktop environment and the applications themselves. Apps can get it sort-of free if their widget set (KDE or gnome/gtk, say) supports it.
21:08 < Elladan> fscale, you can also not scale everything, but instead just go into your font config and select larger fonts.
21:08 < fscale> Elladan: Both desktop environment must support it & app devs have to do separate work?
21:08 < Elladan> Yes. The app may get it for free if it's simple enough that it just uses the desktop environment's default widgets.
21:08 < Elladan> More complex apps have to explicitly support it.
21:10 < Elladan> Anyway, 2:1 scaling is a commonly needed feature, because these modern UHD display resolutions are pretty much illegible without it, but also because integer scaling like that looks pretty good. 1.25:1 doesn't look very good.
21:10 < Elladan> So even if Fedora supports 1.25:1 scaling in some mode, you're going to run into apps that don't and it will annoy you.
21:10 < Elladan> As an alternate option, in pretty much into distro / gui you're able to go into your font config control and simply pick bigger default fonts, or sometimes change the DPI.
21:11 < Yamakaja> Hey, does anybody know if the 3.10 kernel supports p2p addresses with ipv6? I'm getting an invalid argument error when trying to set the peer address (ip a add <x> dev <if> peer <THIS>)
21:12 < Elladan> fscale, I'd suggest just doing that if you're finding the 1920x1080 display to be a little bit squinty. It won't scale the whole apps / icons up, but usually the larger fonts work.
21:13 < Elladan> fscale, just as a bit of history here, the reason hidpi / 2:1 scaling was invented in the first place was primarily because applications didn't properly render themselves at higher DPI settings, using vector icons and so on.
21:14 < Elladan> fscale, just scaling the whole app up 2:1 and hacking the font renderer was seen as a more tractable hack-fix (across all operating systems!) than making the applications themselves be properly dpi-independent.
21:14 < Elladan> Some operating systems, e.g. Android, were built from the start with dpi-independence in mind and don't do things that way.
21:15 < fscale> Elladan: Android must be all vector graphics then..material design?
21:16 < barometz> not so much vector, but it's made pretty easy to include variously scaled copies of images
21:16 < Elladan> fscale, pretty much yes. All vector graphics isn't really possible of course -- you always have bitmaps -- but the real point is that apps are expected and required to deal with it correctly or they just don't work on various phones.
21:16 < barometz> scaled, or with details adjusted to be appropriate for a given DPI
21:17 < searedvandal> my windows laptop defaults to 125% scaling, and the software I need acts rather bonkers with that scaling. so it's the same issue "on the other side"
21:17 < markasoftware> there is no way to do a conditional in Make based on a target-specific variable?
21:17 < markasoftware> because they are evaluated at compile time
21:17 < Dagmar> There is.
21:18 < barometz> markasoftware: fortunately there's a flag in the compatibility options to fix most of that
21:18 < fscale> searedvandal: Which software you have problem with? Is it the blurry fonts?
21:18 < searedvandal> fscale, erratic mouse behaviour in Ableton
21:18 < markasoftware> a command line one?
21:18 < bls> you don't want what you think you do in this situation though. you need to start using sentinels because you're asking for make to track state between invocations without any artifacts available to do so
21:19 < searedvandal> so they haven't figured it out over there either. so I just set everything to 100% and sit close to the monitor :p
21:19 < Dagmar> Elladan: Apps that become illegible because of an HD display are failing programmers
21:19 < markasoftware> bls: I just have two different ways to build my app, each one in a different target, but they share many prerequisites
21:19 < Dagmar> X knows how large the physical screen is.  It also knows how many pixels high and wide the display is.  If someone uses a pixel-size measurement, that's their failure to plan ahead
21:20 < markasoftware> some of these prerequisites should be run slightly differently depending on which "top-level" target is being executed
21:20 < markasoftware> i am not trying to do anyting specifial between invocations of make
21:20 < bls> markasoftware: right, but you need a way to track which method was used to build your outputs so you can get rid of things built in the wrong way
21:20 < Elladan> Dagmar: yes... The whole thing is a hack.
21:20 < Dagmar> It's a hack that works
21:20 < markasoftware> yes, i suppose
21:21 < Dagmar> THe only time it can really be said to fail is when someone puts a display on a giant projector
21:21 < Dagmar> ...and there's a simple override to fix that
21:21 < Elladan> Dagmar: I mean, hidpi mode is a hack. Just using the x dpi setting is correct
21:21 < bls> the way I've done this in the past is have the mutually exclusive build targets put a special dotfile in the output dir
21:21 < Dagmar> ...where you basically just lie to X and tell it the display is really only a more reasonably-sized monitor
21:21 < Elladan> Dagmar: but there are a lot of broken programmers
21:22 < bls> and then add dependencies on that dotfile that will call clean if it doesn't exit
21:22 < Dagmar> Elladan: And unfortunately I have neither the budget nor the clearance to go beat them all with sticks
21:22 < markasoftware> that's a good idea
21:23 < Elladan> Dagmar: it's not really a lie if you consider the user's distance to screen :-)
21:23 < Dagmar> No, it's still a lie.  It's just a lie that's useful
21:23 < bls> you'll end up with a couple extra rules/redirects, but it'll provide that state tracking between invocations you need that you can't get solely from file presence and time stamps
21:23 < Dagmar> It also beats the hell out of seeing all your fonts rendering two-pixels high
21:24 < Dagmar> That having been said, there's a lot of stuff that's done on Android that's not precisely ideal either, just because of the irrational exuberance it has for GPU offload
21:25 < markasoftware> that doesn't alleviate the need for runtime conditionals though, does it?
21:25 < markasoftware> seems like an unrelated thing, albeit a useful one
21:25 < Dagmar> Like, you can tell it to rescale a bitmap by precisely 2x or 4x, but more often than not it gets handed off to EL/GL which just makes a mess of things
21:26 < bls> something like: `dev: output/.dev-build` `prod: output/.prod-build` `output/.dev-build: clean\n\ttouch output/.dev-build`
21:26 < Dagmar> Sure, they've got vector asset rendering now, but that's just starting with a much higher new baseline of GPU power eaten up
21:27 < bls> so if the sentinel isn't present, you'll call clean before creating it
21:37 < markasoftware> bls: that will run the output/.dev-build rule every time, and therefore run `clean` every time even if `prod` has not been run in the middle
21:37 < w1n5t0n> hi, I'm running Elementary OS 0.4.1 and I have a weird problem, when I boot normally I can log into my account but everything (except for actual windows) is invisible. The taskbar, the dock, the applications drop-down menu, even the terminal window are all invisible. I can blindly search for programs and launch them, but everything else I just can't see. It works when I boot into recovery mode, something to do with graphics drivers?
21:38 < bls> not if output/.dev-build already exists
21:38 < markasoftware> it still runs, I think it's not sure whether it's up-to-date or not because it has nothing to compare it to
21:40 < markasoftware> i'm pretty sure make stil runs rules on files that are already created, it only skips if its newer than prerequisites, but this rule has no "real" prereqs
21:40 < triceratux> w1n5t0n: did you try nomodeset or xforcevesa on the bootloader commandline ? one of them might get the video up long enough to keep debugging
21:41 < Dagmar> nomodeset is teh best
21:41 < Dagmar> I really wanted to believe modesetting would be 100% reliable, but that's yet to be the case
21:41 < triceratux> xforcefesa is only for bonafide buntus which eos seems to be
21:42 < w1n5t0n> triceratux: how would I do that? I'm a bit of a newbie
21:42 < bls> I'm probably leaving something out of the pattern, been a long time since I had to use it
21:43 < Dagmar> w1n5t0n: At the grub menu, press the key that allows you to edit the boot params, and just add those and maybe 'nosplash'
21:43 < triceratux> w1n5t0n: when the grub2 menu comes up you should be able to hit "e" to edit the line before selecting it. add "nomodeset" at the end & hit ^x to proceed
21:45 < w1n5t0n> triceratux: thanks, will report with the results!
21:49 < triceratux> Dagmar: nomodeset is no longer preferred practice since theyve made nodesetting the default in xorg https://tjaalton.wordpress.com/2016/07/23/intel-graphics-gen4-and-newer-now-defaults-to-modesetting-driver-on-x/ it now works better than vesa finally & doesnt have to be turned off all the time
21:49 < birdbolt1> i set up a volume with `volumes: -type:bind source:/home/username/dockervolumes/web1root target:/etc/nginx`
21:49 < triceratux> *modesetting
21:49 < birdbolt1> i run the container and the web1root folder on the host is empty
21:49 < birdbolt1> did i type something wrong?
21:49 < bls> birdbolt1: intending that to be for #docker?
21:49 < birdbolt1> thats docker by the way, sorry
21:50 < birdbolt1> bls, posted in there, its a mortuary
21:50 < birdbolt1> quiet as ever
21:53 < sadasaulna> birdbolt1: #docker peeps busy waxing their beards and starting a new pop up restauraunt probably
21:54 < birdbolt1> sadasaulna, rofl
21:54 < Egyptian[Home]> hi how come sometimes when i connect to my home wifi it says i got the worng password? the password should be saved on file. network manager settings show i have the correct passowrd so what giveS?
21:54 < birdbolt1> good news i posted there right after coming over here and theyve got like 3 people chiming in now
21:54 < birdbolt1> wonder where they suddenly came from lol
21:54 < sadasaulna> their ears must be burning :)
21:55 < bls> most people don't idle in support channels waiting on someone to ask questions
21:56 < esselfe> I do, no one has question for me
21:56 < sadasaulna> yeah this chan has a bit of life
21:56 < esselfe> Reading IRC is like reading a book, live
21:57 < esselfe> and sometimes I can learn something
21:57 < bls> I set an alarm on my client, if something triggers it when I happen to be near a computer I'll look in, otherwise I'm not sitting staring at a screen with no activity
21:57 < sadasaulna> i learn shit all the time from this channel
21:58 < sadasaulna> i've got nothing better to do bls, its between this and a game of freecell at the moment
21:58 < esselfe> my systemd journal is ugly and bloated with useless stuff, I miss the dmesg day
21:58 < searedvandal> freecell sounds great
21:58 < esselfe> man my neighbours are cooking meat on the BBQ, smells divine
21:59 < esselfe> I'm between reading IRC and reading "Fifty shades of grey", I have the 3 books
22:00 < esselfe> I notice the server doesn't disconnect me when I post messages
22:00 < esselfe> they used to disconnect me after 30-45 minutes being idle
22:00 < sadasaulna> IRC and a bit of spank and tickle, fair enough
22:00 < searedvandal> good mix
22:01 < sadasaulna> is that not something that keepalives would fix esselfe ?
22:01 < sadasaulna> what irc client?
22:03 < esselfe> I never read about it, this is a new OS, I used to set it to 600 seconds, I don't know how long it is now
22:03 < esselfe> 7200 seconds
22:04 < esselfe> 120 minutes
22:04 < esselfe> (/proc/sys/net/ipv4/tcp_keepalive_time)
22:06 < esselfe> irssi
22:06 < sadasaulna> so you gonna put it back at 600 seconds, or something lower?  Might fix your issue
22:06 < Dagmar> That won't have anyting to do with being disconnected from the IRC server
22:07 < esselfe> no I'm going to let it that way, I used to be paranoid on ip_conntrack stuff
22:07 < esselfe> (that's having some 600 connections)
22:08 < sadasaulna> Dagmar: what do you think is causing it?
22:09 < sadasaulna> brb, getting a bottle of red wine.
22:09 < Dagmar> *Inactivity*
22:09 < sadasaulna> Dagmar: i've stayed logged into IRC for days idling
22:10 < esselfe> I think the client holds an idle time and the server also
22:10 < Dagmar> They dont' really do inactivity disconnections on freenode much.  ...or at least not on the servers I use
22:12 < sadasaulna> Does irssi not send those ping messages to the IRC server even if nothing is going on?  I'm not very clued up on IRC, even if I have used it for 20 plus years lol
22:12 < esselfe> anyone know a good replacement to oidentd? it refuses to compile, I guess my compiler is too recent >.<
22:12 < Dagmar> Probably not unless you _tell_ it to send pings
22:12 < Dagmar> esselfe: pidentd
22:13 < esselfe> ping-pong is in the IRC rfc
22:13 < Dagmar> It's an identd.  Don't expect rocket science
22:13 < sadasaulna> ah ok.  anyway, wine, wine, brb
22:13 < esselfe> thanks
22:13 < Dagmar> I've replaced them with shell scripts before
22:13 < Dagmar> ...and a tcp_wrappers twist directive.
22:13 < esselfe> I know, I just want the tilde off my address :P
22:13 < Dagmar> Most things making ident queries aren't real picky about whether the answer is sent _before_ the query.
22:16 < Dagmar> ...nor do they really scrutinize the response all that much.
22:18 < birdbolt1> can you guys try hastebin.com
22:18 < birdbolt1> and let me know if it use reachable?
22:18 < markasoftware> birdbolt1: it is reachable
22:19 < birdbolt1> i ran my docker container and while it was running, it was somehow intercepting requests to pastebin...
22:19 < birdbolt1> i shut down the container, and pastebin shows as unreachable in my browser
22:19 < birdbolt1> i wonder why specifically pastebin, since other sites worked
22:20 < Dagmar> That depends on what derpy things you put into the host resolver
22:20 < Dagmar> Probably some ipv6 nonsense
22:20 < birdbolt1> errmm, my nginx config untouched, apart from pointing it upstream to django (which served the error message i saw)
22:21 < birdbolt1> what else but pastebin can i use, since hastebin doesnt work
22:21 < markasoftware> ix.io
22:21 < markasoftware> ghostbin
22:22 < esselfe> it doesn't work, I'll fetch some docs from the internet although it should be simple
22:22 < compdoc> lets not be hasty
22:22 < markasoftware> or you can use Anypaste to get a nice command line interface for several of them
22:22 < esselfe> birdbolt1: termbin as in 'cat this |nc termbin.com 9999'
22:23 < esselfe> birdbolt1: ideone is also quite popular
22:23 < azarus> esselfe: UUOC
22:23 < azarus> useless use of cat
22:23 < azarus> nc termbin.com 9999 < this
22:23 < markasoftware> echo this | xargs cat | nc termbin.com 9999
22:24 < jim> UUOC is telling a cat to differentiate my favorite function
22:24 < esselfe> my cat is just fine, it's just logic
22:24 < MrElendig> ptpb.pw
22:26 < jim> well another way to explain it is... if you have nc installed, you can pastebin the output of an arbitrary command, for example ls -CF if you run it like this: ls -CF | nc termbin.com 9999
22:27 < birdbolt1> can someone please explain this wizardry for me? https://ghostbin.com/paste/co7xy
22:28 < searedvandal> what's to explain?
22:28 < bls> what wizardry?
22:29 < birdbolt1> the missing nginx.conf* files
22:29 < searedvandal> 'cd ..' will probably help
22:29 < avis> n00b's
22:29 < birdbolt1> it copied everything else but those, which were specifically what i was interested in
22:30 < birdbolt1> searedvandal, they are all in the same directory
22:30 < searedvandal> ok
22:30 < avis> i do tattoo art.  don't mean to a griffith, tell them thank you
22:30 < birdbolt1> see the difference between the ls command in the container, and the ls command on the host
22:30 < revel> birdbolt1: You copied it to ~/dockervolumes/web1root/ and you appear to be in ~/dockervolumes/web1root/nginx/ ...
22:30 < searedvandal> ^ which is why I said 'cd ..' may help
22:30 < bls> you copied the file to a different directory. no wizardy here
22:31 < avis> save every days work in a new folder ?
22:31 < birdbolt1> oooh looks like i forgot to include the command prior to that revel
22:32 < birdbolt1> i am confused why i copied the entire directory and those files were included
22:32 < bls> or use a version control system so you don't have to have mulitple copies of the same thing lying around
22:32 < searedvandal> birdbolt1, does nginx.conf appear in web1root/ ?
22:32 < birdbolt1> i recognize copying th individual file worked
22:32 < BenderRodriguez> Dear Linux
22:32 < avis> i like bible verse and terminal colors
22:32 < BenderRodriguez> I need help
22:32 < birdbolt1> searedvandal, yes, but im confused as to why it didnt get copied along with everything else
22:33 < birdbolt1> there is a command prior to those, which i forgot to include
22:33 < avis> dumass.  no not dumb ass.  dumass
22:33 < lnnb> dear BenderRodriguez, please send snacks
22:33 < revel> lnnb: Do you have iron deficiency?
22:33 < lnnb> yes, actually  :\
22:33 < searedvandal> who doesn't these days?
22:33 < revel> Then you'll love the snacks.
22:33 < revel> I don't.
22:34 < searedvandal> you take too good care of yourself then.
22:35 < BenderRodriguez> Anyone use virsh/libvirt?
22:36 < BenderRodriguez> I think I found a bug but the libvirt room is dead to do anything aboujt it
22:36 < hassoon> i use nothing
22:37 < revel> BenderRodriguez: https://libvirt.org/bugs.html#bugzilla
22:38 < BenderRodriguez> but do *I* have to report it? :v
22:38 < revel> No, actually, it has to be your legal guardian.
22:39 < searedvandal> well, you could always take the backseat and wait for someone else to find and report the bug. but that's not the right attitude
22:42 < bls> searedvandal++ because the OSS community loves it when people complain about things but don't provide patches or at least bug reports
22:42 < searedvandal> I know, right
22:42 < searedvandal> if you find a bug, report it. doesn't take much effort really
22:43 < azarus> and patch it if you can
22:43 < searedvandal> sure, if you got the skills
22:43 < searedvandal> my skills stop at changing password
22:43 < azarus> in an ideal world users == developers
22:44 < bls> worst case, it's ignored and no one ever sees it. that's going to be rare. what most commonly happens is the report eventually shows up in a web search, enough other people pile on with more details, and it builds enough momentum to get attention
22:44 < searedvandal> yeah
22:51 < TR1950X> I am running about 5 VM's using KVM on my workstation. Is it possible to tell Linux to shutdown the vm's beofre shutting down the host machine?
22:52 < bls> TR1950X: yes
22:52 < compdoc> it should already be doing that
22:52 < TR1950X> okay, thanks
22:53 < bls> yeah, unless this is some fringe distro, it should already be happening
22:53 < compdoc> windows likes to install updates when it shuts down, so it might take too long and kvm wont wait forever
22:54 < bls> in which case you'd want to customize your shutdown above and beyond the default behavior
22:56 < markasoftware> is there a channel for Make?
22:57 < bls> markasoftware: ##workingset is a good place to try
23:04 < redrapscallion> I have a video, and I'm trying to use ffmpeg to get rid of 16 pixels from the right border, and 100 pixels from the bottom border to hide an unwanted part of the GUI that's showing up in the video.
23:05 < redrapscallion> I've tried following https://video.stackexchange.com/questions/4563/how-can-i-crop-a-video-with-ffmpeg but it looks like my crops are happening at the top left corner, and not from the bottom right corner.
23:06 < lnnb> whats the command for crop
23:06 < bls> crop what?
23:06 < lnnb> image
23:07 < bls> convert from the ImageMagick package
23:07 < lnnb> do you have redrapscallion on ignore?
23:08 < bls> no, don't know how to crop videos, just images. thought you were asking a different question
23:09 < Elodin> is there a way for me to stream some video to a smart tv from within my linux box in the same wifi?
23:10 < bls> Elodin: depends on what protocol the TV supports, but check out minidlna to see if it's what you want
23:10 < bls> guess it's called readymedia now
23:11 < bls> https://wiki.archlinux.org/index.php/ReadyMedia
23:12 < lnnb> redrapscallion: `man ffmpeg-all`  search for 'crop' or 'Crop the input'  your x/y offset should probably be zero, and use out_w/out_h to set the right/bottom cutoff
23:12 < lnnb> i think, but have not tried
23:12 < Elodin> bls: thanks
23:20 < redrapscallion> lnnb: well, I've tried : ffplay -i input.mp4 -vf "crop=out_h=in_h-50:out_w=in_w-60" (ffplay is like ffmpeg, but it shows a preview)
23:21 < redrapscallion> lnnb: but that takes off pixels from the right AND left, and from the bottom AND top. I read the man pages, but I can't seem to understand why it's taking from both sides.
23:22 < redrapscallion> lnnb: I think that's expected behavior though, because in their examples, a crop of "crop=in_w-2*10:in_h-2*20", will crop 20 pixels from top and bottom, and 10 from left and right.
23:23 < lnnb> why  mess with in_h/in_w
23:24 < lnnb> the manual states x/y defaults to (in_w-out_w)/2
23:24 < lnnb> well, x is that, y is (in_h-out_h)/2
23:31 < sklv> hi, i have this regex to remove brackets and contents from string https://regex101.com/r/Mo4iW7/1
23:32 < sklv> i'm trying to use it inside qtregexp like this https://dpaste.de/qbsD
23:32 < sklv> but it's not working onthe string 'test string [test]' with no matches - any ideas?
23:33 < bls> sklv: what tool are you using it with?
23:33 < bls> oh qtregexp
23:33 < sklv> qtcreator
23:33 < bls> I believe regex101 produces regexes for PHP that don't work with a lot of other tools
23:34 < sklv> yep which is why i'm asking here for ideas - the regex should match in principle
23:35 < phogg> sklv: do you have docs on what flavor of regex qtregex supports?
23:35 < sklv> https://doc.qt.io/archives/qt-4.8/qregexp.html
23:35 < bls> also note there there are #qt and #qt-creator channels on here
23:36 < sklv> that's a good idea
23:37 < phogg> looks like PCRE or something intended to be similar to it
23:37 < bls> yeah, PHP supports a modified subset of PCRE
23:38 < searedvandal> "QRegExp is modeled on Perl's regexp language"
23:38 < bls> heh, so just about as bad
23:38 < sklv> ah i found what was causing it to fail
23:38 < phogg> searedvandal: ah, good catch. So perl-alike things like \d, but not pcre exactly
23:39 < phogg> sklv: is it the ?
23:39 < sklv> it doesn't require \\ to escape \ even though it prints unknown escape sequence in isseus
23:39 < sklv> and  yes it doesn't support ?
23:39 < phogg> sklv: that was my next guess
23:39 < phogg> sklv: inside [] you should not need to escape [
23:39 < bls> was curious about that too
23:39 < sklv> i'm using double quotes in qt which sometimes require escaping
23:40 < phogg> inside C++ double quoted strings I would expect you to require an escape for \
23:40 < bls> some syntaxes require escaping ] though unless it's immediately preceeded by [ if memory serves
23:42 < searedvandal> sklv, the QRegExp docs are actually quite good if you ever get stuck in the future
23:44 < bls> heh, been a while since I messed with something this picky `echo 'foo[bar]' | sed 's/[][]//g'`, []s usually imply a grammar that has a better parser the regexps
--- Log closed Sun Jul 01 00:00:20 2018