--- Log opened Fri Jun 29 00:00:13 2018 --- Day changed Fri Jun 29 2018 00:00 < pingfloyd> JeffATL: you have to deprogram of all of the Microsoft thinking 00:00 < pingfloyd> deprogram them* 00:00 < lordvadr> Wixy: Could be tooling. What's the ask here? 00:00 < Wixy> also how on earth is he measuring response time with apache? what is he doing to get that log? 00:00 < pingfloyd> JeffATL: like can any of them compile from the command line? 00:00 < pingfloyd> JeffATL: or do they just hit F5 and watch the magic? 00:00 < JeffATL> pingfloyd: unfortunately I just work with them, not instruct them (such that I have GPA leverage) 00:01 < lordvadr> Wixy: It looks like an apache log. Apache can log in various formats, so it just looks like that. We don't know what tool hes using. 00:01 < lordvadr> But, what's the ask here? 00:01 < JeffATL> pingfloyd: they're BizAnal students, so, not even. 00:02 < pingfloyd> JeffATL: so point and click "programmers"? 00:02 < Wixy> but it looks like the log you would get from YOUR apache server, and that time would be what it took to respond to those requests 00:02 < pingfloyd> I'm not sure what BizAnal is 00:02 < lordvadr> Wixy: I agree. That's what I thought initially. 00:02 < lordvadr> pingfloyd: Bend over and I'll show you. 00:02 < Wixy> not the time it took to send and get a response from a REST server in another place 00:02 < lordvadr> Wixy: Are we sure that's what you're looking at? 00:03 < Wixy> I'm sure I need the response time from A to B, but I'm not sure that's what he sent 00:03 * lordvadr was out of line there. Is sorry. 00:03 < lordvadr> Wixy: WHY did you ask him for that data? 00:03 < Wixy> it looks he's running a server in A, and that explains the low response times? 00:04 < Wixy> because he said my response time was high, and I couldn't believe he was getting lower times provided I'm colocated 00:04 < lordvadr> High compared to what? 00:04 < pingfloyd> Wixy: is your app spread across different services? 00:04 < lordvadr> And what does it matter? Is it causing some kind of issue? 00:05 < Wixy> those numbers at the left compared to an average of 31ms, lordvadr 00:05 < Wixy> lordvadr, yes, 31ms is way to high for this application, I'm supposed to fix it but I'm not sure what is wrong.. I'm trying to figure out :) 00:05 < lordvadr> So? You're less than an order of magnitude away from each other. 00:06 < lordvadr> Wixy: So the ask is what? Improve the speed of light? 00:06 < lordvadr> I don't understand. Dev sent in ticket saying what? 00:07 < Wixy> not really. the question is this: provided the numbers are accurate, how is it possible he's getting lower response time than me if I'm running on the same location the 3rd party server is? 00:07 < avenger> lasers 00:07 < Wixy> he's got 7ms a couple of time, really? for a server in another datacenter? 00:08 < Wixy> that looks hard to believe 00:08 < lordvadr> Wixy: That's a good question. My first thought would be that he's not giving you what you asked for. 00:08 < Wixy> unless of course Google Cloud and AWS share datacenters, don't know 00:08 < retran> I can get 3ms from my office to server's located in a downtown Los Angeles DC 00:08 < retran> I'm on AT&T fiber in a whole nother part of Los Angeles 00:09 < sysfault> Timing cached reads: 4812 MB in 2.00 seconds = 2405.60 MB/sec 00:09 < sysfault> Timing buffered disk reads: 802 MB in 3.01 seconds = 266.84 MB/sec 00:09 < sysfault> is that a decent benchmark for a 250g ssd in a laptop? 00:09 < lordvadr> retran: I was just about to say, "from around 25km away or less" 00:10 < retran> ah. 00:10 < bls> sysfault: maybe try asking ##hardware 00:10 < lordvadr> retran: Speed of light is the speed of light. It goes 2/3'rds c in fiber, and takes about a 20x overhead in the equipment. That's minimum. 00:11 < Wixy> I guess I'll just wait and ask him better. it looks like the logs are from a server not a client, makes no sense at all 00:11 < retran> I'm certainly under 25km. about 8.5km 00:12 < lordvadr> Wixy: When you ask, be sure you ask for information on how to repeat his tests. 00:12 < retran> my equipment in my office is old poop, just the fiber router that AT&T gives you and a ubiquity gigabit ethernet over wire internally at the office 00:12 < retran> given that, I'm happy with my speeds :D 00:13 < lordvadr> retran: I was expecing you to say something like, "from New York". 00:13 < retran> I was unhappy when League of Legends moved their servers from Los Angeles to Chicago :| 00:13 < lordvadr> Yeah, but 350 Cermak is a better data center than 1 Wilshire. 00:14 < retran> there's 3 other big DCs across the street from 1 Wilshire 00:14 < lordvadr> I'll bet. It's similar near 350 too. 00:14 < retran> I don't know how to evaluate how awesome a DC is though 00:15 < retran> i used to rent a 2U space in one of them 00:15 < retran> i avoided going there. so noisy! 00:15 < lordvadr> But if you want to be connected to the US, 350 or maybe MWIX. 00:15 < retran> not at all glamorous like I was lead to believe in the movies ;) 00:15 < bls> how much spilled coffee and cigarette ash under the raised flooring, that's how 00:16 < retran> lordvadr, yes. it does make sense geographically 00:16 < retran> "be in the middle" 00:16 < retran> make both NY and LA pissed 00:16 < retran> and your 2 customers in Chicago happy 00:16 < lordvadr> Yeah. Plus fewer earthquakes, cheaper power, free cooling for 1/3 of the year 00:17 < JeffATL> retran: a few years ago i carried a 1U server into one DC and carried it right back out again when i saw the conditions 00:17 < retran> JeffATL, right! I couldn't take working at a place like that. 00:17 < JeffATL> retran: goalpost-style racks and a lot of what i like to call "gravity mount" 00:17 < retran> heh... "gravity mount" 00:18 < retran> they're pretty damn noisy. i've been in wood mills 00:18 < retran> the DCs are noisier 00:20 < retran> JeffATL, your description reminds me of this place's DC: https://joesdatacenter.com/ 00:21 < retran> (yes, their background pic is fake) 00:21 < retran> a bunch of ATX boxes metal wire-frame metal shelves 00:21 < JeffATL> retran: definitely. hearing protection and warm clothing. even if the temp in a big serious DC is 72F, you find yourself losing body heat at an alarming rate 00:22 < prence> “What a pleasure to work with everyone at Joe’s!! Thank you!” 00:22 < prence> -Basil Mogdalov 00:22 < prence> glowing recommendation haha 00:22 < retran> and they don't do cold/hot isles 00:22 < bls> and I always hated how dehydrated I got when having to do my own moves 00:22 < retran> (at least not when I went there) 00:22 < bls> and try skin splits a lot easier hehe 00:23 < bls> dry 00:23 < lordvadr> retran: I don't think you could do h/c aisles if you let people bring their own equipment. Are you going to force front-to-back airflow? 00:23 < retran> just to make the HVAC sane 00:23 < retran> otherwise you waste the positions 00:24 < bls> nah, you just let some of those reverse flow cisco switches blow their heat into the cold aisle.... :| 00:24 < jnewt> i'm getting permission denied when trying to use cutecom on a serial port. i did sudo adduser dialout, but then groups doesn't list dialout (groups does however) what am i missing here? 00:24 < retran> if you do h/c isles you design the have intake/outtake HVAC accordingly 00:25 < retran> less temp variation on the shelves 00:25 < jnewt> i've logged out and back in. i'm running ubuntu 18.04lts (statndard installation) 00:25 < lordvadr> jnewt: You want usermod, and the user will have to logout and back in. 00:25 < retran> lordvadr, true. people could make their fans blow any which way 00:26 < lordvadr> jnewt: usermod -ag dialout 00:26 < jnewt> lordvadr, something like usermod -aG dialout right? yeah, tried that too. 00:26 < lordvadr> jnewt: slow down. Give me `id myuser` 00:27 < bls> also got a kick out of the leed buildings with the louvers to outside for cool days and all the pollen that'd get into everything 00:27 < retran> oh yuck, bls 00:27 < Pusteblume> do you still need to re-login your user so the changes take effect? 00:27 < retran> leed certification on a datacenter? you'd think they have special provisions 00:28 < retran> and if venting to outdoors, they should have put filters :( 00:28 < lordvadr> Pusteblume: I think you can gpasswd or something 00:28 < lordvadr> But its easier to just relogin, and you can do that several way without logging out of the console. 00:29 < bls> small particulate filtering required finer filters and heavier fans, negating a lot of the electricity cost savings 00:29 < retran> interesting 00:30 < bls> although it's nothing like the tents Microsoft tried out 00:30 < retran> wasn't it MS that is experimenting with submersible datacenters? 00:31 < bls> yeah, that too 00:32 < bls> way cheaper than halon :P 00:32 < retran> I hear it's really cold in outer space 00:33 < bls> can ECC handle a solar flare? 00:33 < lordvadr> retran: It is cold in outer space, but there's no airflow, so the only cooling available is radiative and not convective. Satellites have considerable problems with cooling efficiently. 00:33 < meyou^> depends on the solar flare 00:34 < lordvadr> It's a real problem. Astronauts bitch about it a lot too. 00:34 < WorldGenesis> is there a way of seeing the status of a CMOS battery in linux? o.o 00:34 < lordvadr> WorldGenesis: I didn't know you could see it in any OS besides noticing the cmos and rtc was reset at boot. 00:35 < WorldGenesis> ahh :P 00:35 < retran> if the chipset had novel features 00:36 < brrn> how many linux distro's are there? 00:36 < retran> motherboards provide temperature, fan speed, and other telemetry. but never heard of CMOS battery voltage being one 00:36 < fryguy> brrn: a lot 00:37 < ia5string> @brrn too many to count on a single hand! 00:37 < retran> brrn, there are exactly three-billion nine-hundred-sixty-five 00:37 < brrn> retran: I see 00:37 < infinisil> brrn: There's only one True Distro 00:37 < brrn> infinisil: and that is... 00:37 < infinisil> Up to you :) 00:37 < lordvadr> GNU/Hurd 00:38 < retran> Stallman sighting! 00:38 < infinisil> Nah jk, it's NixOS 00:38 < lordvadr> *now with Upstart!* 00:38 < Loshki> retran: I assume it's not worth the additional cost. The symptoms of CMOS battery failure are pretty distinctive and the cure trivial. 00:38 < retran> Loshki, I agree. 00:39 < lordvadr> Are we going to do this "my distro is better than yours, hurr durr" again? We just did that a couple of hours ago. 00:39 < rascul> do it 00:39 < retran> brrn, maybe you'd find it more useful to look at a chart that maps linux distro's by usage 00:39 < rascul> storm linux was the best 00:39 < rascul> s/was/is/ 00:39 < retran> Alpine+Busybox 00:39 < lordvadr> Knoppix for the win!!! 00:40 < infinisil> lordvadr: I use Arch btw 00:40 < brrn> retran: I'd like to see this 00:40 < infinisil> jk I don't 00:40 < retran> brrn, go google my friend 00:40 < lordvadr> infinisil: I knew you were lying because you seem somewhat reasonable. 00:40 < retran> tell us what you find 00:40 < infinisil> lordvadr++ 00:41 < retran> brrn, you could also ask "what are the top 3 most popular distros" 00:41 < retran> "how many distros are there" is kinda... it's like asking "how many jellybeans are in this jar" 00:41 < brrn> retran: but I want the 53 to last least used distro :) 00:41 < lordvadr> brrn: I think that's still Knoppix. 00:42 < retran> the least used distro will always be something invented just today 00:42 < pingfloyd> brrn: you know about distrowatch? 00:42 < retran> there's like... 1000s of distros that are abandoned. are you interested in those too 00:43 < lordvadr> retran: 1000s? Maybe 200? 00:43 < pingfloyd> brrn: distrowatch.com 00:44 < pingfloyd> there's more than 200 00:44 < brrn> on wikipedia there is more that 200 distros. 00:44 < pingfloyd> that whole linux heritage tree had more than that alone 00:44 < brrn> I know 00:44 < lordvadr> pingfloyd, brrn: Abandoned. 00:44 < jim> as many jellybeans as would fill it to the point it's filled now:) 00:44 < pingfloyd> there's probably more abandoned that alive all these years 00:45 < pingfloyd> s/that/than/ 00:45 < lordvadr> Yeah, but are we talking 4 figures? 00:45 < retran> easily 00:45 < pingfloyd> some of the dists are "blue chips" though 00:45 < lordvadr> Whatever happened to Hanna Montanna Linux? 00:45 < pingfloyd> the main ones for example 00:45 < pingfloyd> and even a few niche ones 00:46 < pingfloyd> I wouldn't say it is out of the realm of possibilities that some of the main one will even close shop some day. At that point they'll be in the hands of the community. 00:47 < pingfloyd> if they close shop though, that means a lot of interest was lost usually (as in the community to support them is no longer there). 00:47 * lordvadr looks at cloak 00:48 < lordvadr> So, distrowatch says 323 active distros. And we think there's more than 3 times that abandoned? Maybe we need to define what constitues a distro vs just some crap you burned to a cd once. 00:48 < retran> pingfloyd, which distros are in danger of having their current sponsors "close shop"? 00:48 < infinisil> brrn: If you want to actually count distros, you need to define what a distro means 00:48 < retran> SuSE comes to mind 00:50 < lordvadr> retran: Canonical hasn't been doing well for a long time. 00:50 < retran> but it's backed by an endless stream of VC funding 00:51 < retran> supported by lots of goodwill 00:51 < lordvadr> Sure, but I wouldn't call it endless. 00:51 < Comstock> i like that there are a ton of "distros" out there 00:51 < Comstock> gives us variety. 00:51 < lordvadr> I know there was talk about us waving in canonical refugees by the boatload not all that long ago. 00:52 < Comstock> i had issues with this laptop running the more well known distros and found one called "linux lite" 00:52 < Comstock> if there were more than the big 4-5 distros, i'd probably be stuck on windows on this laptop 00:53 < iflema> Comstock: you could hae a go at compairng kjernel configs and finding out what is stopping you 00:53 < Comstock> weren't* 00:53 < lordvadr> I like lots of distros, but I think we could cap it at, like 45 and still be cool. 00:53 < Comstock> iflema once i get another computer up and running that is the plan. 00:53 < retran> I think the cap should be 42 00:53 < Comstock> my main laptop got fried 00:53 < retran> 45 is obviously 3 too many 00:53 < lordvadr> retran: No. That's crazy talk. 45 00:54 < retran> damn. i guess we'll never agree 00:54 < Comstock> civil war. 00:54 < lordvadr> shame.gif 00:55 < lordvadr> It's been fun, I have to go afk for a while. My wife is making me go drink and I need it after you people. 00:56 < meyou^> what do you mean, you people? O_o 00:56 < lordvadr> ah shit 00:56 < lordvadr> We've pissed off the blue guy again. 00:56 < Comstock> i identify as people 00:56 < Comstock> please do not offend me 00:56 < Comstock> us* 00:57 < meyou> i'm blue daboodeedaboodiii 00:58 < Pusteblume> papa smurf? he uses linux? 00:58 < iflema> did the bell ring? 00:59 < meyou> my mom dressed me up as a smurf for halloween when i was a toddler 00:59 < Comstock> papa smurf runs arch 00:59 < meyou> painted me blue and put some tighty whities on my head for the smurf hat 00:59 < Aph3x-WL> no wonder you turned out so messed up 01:01 < meyou> how dare you 01:01 < Comstock> meyou what distro do you run? 01:01 < Comstock> this will settle if you are messed up or not. 01:01 < meyou> mostly centos 01:01 < Aph3x-WL> i was right 01:01 < Comstock> yup, you're screwy in the head. 01:02 < meyou> i don't use linux desktop tho 01:02 < meyou> cause i'm no masochist 01:12 < Cyrum> Hola 01:12 < Cyrum> Hows everybody today 01:14 < brrn> good 01:21 < jim> Cyrum, hi 01:34 < rpifan> can anyeone help 01:34 < rpifan> https://forum.lede-project.org/t/adding-rtl8812au-support-to-glinet150m/16273/2 01:41 < vancha> anyone know if i can install an exe file from the terminal using playonlinux/wine? 01:41 < vancha> i tried playonlinex filename.exe, but it gives me an error with "invalid port specification: 0" 01:42 < bls> vancha: might want to try #playonlinux or the wine channel 01:47 < vancha> bls: thx ;) 02:07 < Elodin> can anyone help me with setting up bumblebee 02:07 < Elodin> bumblebee service status error: Failed to unload module 'nouveau' (ref count: 1). 02:08 < Elodin> when trying to run optirun: [ERROR]Cannot access secondary GPU, secondary X is not active. 02:14 < Pentode> is your card supported? i thought bumblebee was dead, i could be wrong.. 02:15 < Elodin> Pentode: yes. im running a lenovo thinkpad and it supports it. the option in the bios is also enabled for optimus technology 02:16 < Pentode> yeah but if you are using a newer kernel / linux distro there could be issues 02:16 < Pentode> the last commit was five years ago 02:16 < unperson> I am trying to troubleshoot a service on a remote machine. It listens on a tcp port. When you connect, it sends you a challenge, and you must send back a response based on the challenge. 02:16 < Elodin> Pentode: would you say i would be better off not using it? 02:17 < Pentode> well im not sure where you'll find help. you could try using a period kernel / distro just to see what happens 02:17 < unperson> If I have a program to compute the response from the challenge, is there a simple way to use netcat to connect in, feed the challenge to my program, and then send the response back to the server (opposite direction on the same tcp connection)? 02:18 < unperson> I feel like this must be possible, but I'm clearly missing the basic nc/bash skills to make it work. 02:19 < xamithan> Bumblebee is dead, it all uses DRI prime now 02:21 < Elodin> xamithan: i just came to linux from osx. i have this laptop with 2 graphics, how should i handle it? 02:21 < Elodin> what is the way to do it 02:22 < xamithan> Well I use nvidia drivers and nvidia-prime 02:22 < xamithan> But it works out of box if you use open source drivers on most modern OS's 02:22 < Elodin> xamithan: which distro are you in? 02:22 < xamithan> I bounce a round a lot. Currently it is on ubuntu 18.04 02:23 < Loshki> unperson: as long as the challenge and response are just printable text, why can't you just cut and paste them? Of course, if it's broken... 02:23 < unperson> Loshki, They aren't printable text. 02:24 < Elodin> xamithan: so what you are saying is that linux would manage both cards automatically, when needed it would swap them? 02:24 < unperson> Loshki, Otherwise, yeah. I'd be done already. :-) 02:24 < xamithan> Not when needed 02:24 < xamithan> With open source you can right click on application and select use dedicated 02:24 < unperson> Loshki, But it's arbitrary binary data. 02:24 < Elodin> ohhh 02:24 < xamithan> With nvidia prime I can open the nvidia control panel and swap which one I want 02:25 < Elodin> xamithan: okay i understand now, so i can choose at given time to use one or another 02:25 < xamithan> Yeah 02:25 < Elodin> can this be done through cli also? 02:25 < kilo> yes 02:25 < xamithan> Open source there is a DRI_PRIME=1 var or something to turn on dedicated when launching the app 02:26 < xamithan> I don't know about with nvidia-prime 02:26 < Elodin> okay, ill try to uninstall all the shit i installed trying to make bumblebee work then 02:26 < kilo> hhehe 02:26 < kilo> Elodin: what laptop is this? 02:26 < Elodin> kilo: t430s lenovo 02:27 < xamithan> Nice, I just got a regular T430 02:27 < Sarah_SWE> Lenovo <3 02:27 < kilo> dual gpu huh 02:27 < Psi-Jack> Booooo.. Lenovo. 02:27 < kilo> I'm not a fan of lenovo 02:27 < Elodin> kilo: why not 02:27 < kilo> I prefer MSI 02:27 < micrex22> kilo they make shit 02:27 < Psi-Jack> Elodin: Suckerfish. 02:27 < kilo> who? mSI? 02:27 < kilo> naah 02:27 < Sarah_SWE> Well to be honest last time I had a Lenovo it was as just an IBM Thinkpad :| 02:27 < micrex22> the ONLY reason why anyone has a TINY inclincation to buy lenovo 02:27 < Elodin> Psi-Jack: i dont understand 02:27 < micrex22> is because of the trackpoint 02:28 < kilo> hehe 02:28 < Sarah_SWE> Trackpoint! 02:28 < kilo> they still include that 02:28 * Sarah_SWE swoons 02:28 < Psi-Jack> Elodin: Embeded spyware pre-installed with their laptops. Sound like a company you can trust? Worse. They tried it twice, AFTER already getting caught the first time., :) 02:28 < kilo> my 15 year old tabletpc has a trackpoint 02:28 < Loshki> unperson: I would write a perl script. You could ask in #bash 02:28 < Elodin> Psi-Jack: yeah, but not on their workline 02:28 < micrex22> psi-jack well that's just it, I'm getting weary of trusting lenovo's china 02:28 < xamithan> In the hardware psi-jack ? 02:28 < Psi-Jack> Elodin: On /ALL/ of them. 02:28 < micrex22> only in software *so far* 02:29 < Sarah_SWE> Who uses any computer without a complete reinstall the first thing they do? O.o 02:29 < micrex22> but I wouldn't past it lenovo to put back doors in the future 02:29 < `Koyaanisqatsi> you trust any corporation? 02:29 < xamithan> What kind of person doesn't reinstall the OS to wipe all the manfacturer junk when they get one 02:29 < Psi-Jack> xamithan: Well, in their pre-installed Windows, since that's all they come with. 02:29 < micrex22> sarah_SWE people who think windows 10 is the best thing since sliced bread? 02:30 < micrex22> psi-jack IBM and Dell offered (offer) Linux preinstalls 02:30 < unperson> Loshki, Thanks. Yeah, I'd write a perl script, except then I'd probably need a bunch of modules for network stuff that I can't get on the machine in question. 02:30 < Psi-Jack> xamithan: About 95% if people. :p 02:30 < Elodin> xamithan: exactly 02:30 < Psi-Jack> micrex22: IBM did. Lenovo isn't IBM anymore. Hasn't been for a while. 02:30 < Sarah_SWE> micrex22: sure, but would we find them here? O.o 02:30 < xamithan> Yeah but how many of people in here, heh 02:30 < xamithan> Since you can't get linux preinstall 02:30 < Psi-Jack> xamithan: About 75%. 02:30 < micrex22> psi-jack well that's why I said offered ;) | IBM still does some x86 stuff 02:30 < Sarah_SWE> My gosh, mIRC got some horrible standard fonts... 02:31 < Comstock> i am one of the few people that actually paid for a mirc license lol 02:31 < micrex22> comstock I haven't used IRC until recently, so... 02:31 < micrex22> and usually I like using IRC on AIX and OS/2 02:32 < micrex22> and mIRC doesn't run on those 02:32 < Comstock> i've been on irc way too long 02:32 < xamithan> xchat was the bestt 02:32 < Sarah_SWE> I did too! But that was soo far back in the day I have barely any recollection about it... 02:32 < Sarah_SWE> xchat! 02:32 < Comstock> i use hexchat 02:32 < Sarah_SWE> Omgosh, that was awesome 02:32 < micrex22> my uncle (back in the mid 90's) an some MS-DOS IRC client 02:32 < Comstock> and irssi 02:32 < micrex22> ran* 02:32 < Sarah_SWE> In like 2001? 02:32 < Loshki> unperson: if it already has perl installed, the lowest level socket open, read and write are probably already there. Where is the client that talks to this server? Do you have acces to server logs? 02:32 < xamithan> They made it friendly when they forked to hexchat 02:32 * Sarah_SWE feels old 02:32 < micrex22> comstick irssi is pretty slick 02:32 < Psi-Jack> Comstock: remember when effect only had like 30 users online at once? 02:32 < micrex22> comstock* 02:33 < Psi-Jack> Efnet 02:33 < Comstock> comstick sounds dirty 02:33 < micrex22> haha damn typos ;) someone on war2 once had the nick "cherrypopper" and I wrote cherrypooper 02:33 < Comstock> haha 02:33 * Sarah_SWE changes to MS Sans 12 and hurts eyes :| 02:33 < unperson> Loshki, I don't think that I have access to the logs. Both server and client are on an intranet. 02:34 < Comstock> i used to spend too much time on efnet and dalnet back in the day 02:34 < Psi-Jack> Comstock: or have you not been in irc that long? Grins 02:34 < Sarah_SWE> micrex22: was probably closer to reality anyhow! 02:34 < Comstock> i think i started on efnet back in like 97 or 98? i might be misremembering 02:35 < micrex22> comstock speaking of WAR2BNE I'm going to set up some of my retro souped up pentium 3 gaming systems with some multiplayer games over IPX/SPX 02:35 < micrex22> at my cousin's 02:35 < micrex22> since a lot of the newer games just aren't cutting it 02:35 < micrex22> (some indie ones are slick, but usually unfinished) 02:35 < Comstock> yeah, i don't like many newer games 02:35 < Comstock> i still play games like diablo 2 02:35 < Psi-Jack> Comstock: okay not quite as long then hehe. 02:36 < Loshki> unperson: well, apparently you have a nail, but no hammer. Authentication systems are designed to be hard to break, which has the side effect of making them hard to debug. 02:36 < Sarah_SWE> micrex22: gimme some Q1 in LAN. That stuff was lit! 02:36 < xamithan> duke nukem 3d lan 02:37 < Psi-Jack> On a 486 DX/4 100. :) 02:37 < micrex22> sarah_swe I have these four things loaded up with Voodoo 5500s so they'd run Q1 GLide well: https://i.imgur.com/Yprg58A.jpg 02:37 < Psi-Jack> With VLB. :) 02:37 < Comstock> i only had a dx2 02:37 < Comstock> big baller with the dx4 02:37 < Psi-Jack> And IPX network. heh 02:37 < micrex22> Because so many games I have are DOS/9x I run 98SE on them. I think that's Windows' only true purpose... to perpetually play old games 02:38 < micrex22> psi-jack IPX/SPX over token ring ;) 02:38 < Psi-Jack> Eww, no token ring. Please no token ring. 02:38 < micrex22> Token Ring can run 100 mbits per second 02:38 < Psi-Jack> With a lot of noise. :) 02:38 < micrex22> psi-jack actually the IBM data connectors were better grounded than RJ45 02:39 < micrex22> although I just run RJ45 token ring as it's more convenient 02:39 < micrex22> and it's not like I'm running them for 5 km 02:39 < `Koyaanisqatsi> wine and dosbox would be a better option 02:39 < Psi-Jack> Heh. Vampire networks was better than token ring. :) 02:39 < micrex22> not quite with multiple hardware configurations 02:39 < micrex22> psi-jack I don't like AUI or BNC 02:39 < unperson> Loshki, Oh sorry, I think I misinterpreted your question. I have access to the logs of the server application. Something is not working right, and I was trying to slap together a minimalistic implementation of the challenge response to eliminate that as an option. 02:39 < Psi-Jack> I'd done BNC at home. 02:39 < Sarah_SWE> micrex22: Voodoo 5500 sooo bad compared to their earlier cards though... Well the cards weren't bad it was just that the competition had caught up and passed way ahead by then :/ 02:39 < `Koyaanisqatsi> so what is up with the cpu vulnerabilities and the new intel microcode? There is no mitigation for variant 3a and 4? 02:40 < infinisil> Hey, how can I ignore the .git directory with `find`? 02:40 < Sarah_SWE> micrex22: still though, nice cards. :) 02:40 < Psi-Jack> One lightning storm took out ALL the NICs in the house though. LOL 02:40 < micrex22> Sarah_SWE the 5500s are actually quite powerful and can even run AOE3 (not that you'd want to play that anyways) 02:40 < micrex22> and they push through warcraft 3 just fine 02:40 < xamithan> A storm once formatted my hard-drive 02:40 < micrex22> although taking advantage of glide is more fun 02:40 < xamithan> I have no clue how it happened 02:40 < Sarah_SWE> micrex22: yeah they're good cards; just that when they finally came out the competitiors had better at same or lower price is all :) 02:41 < micrex22> `Koyaanisqatsi DOS Box is fine for general purpose DOS gaming, but keep in mind it lacks emulating certain audio hardware (that I use regularly) and its OPL emulations are not 100%. Plus I also need to run some 9x games--at which point there are too many performance hits--not to mention running emulators inside emulators to get all of the hardware configurations down 02:42 < infinisil> I tried `find -not -path .git`, but that doesn't seem to work 02:42 < micrex22> Sarah_SWE well actually the 5500s had quite a fair bit of video memory at the time, and the voodoo 6000 was the first to have 128 MB of video memory 02:42 < `Koyaanisqatsi> wine runs better for me than old windows for old win games, and for dosbox I just use fluidsynth for midi 02:42 < Psi-Jack> micrex22: Heh, I still run a BBS and door games in DOSemu. :) 02:43 < `Koyaanisqatsi> windows 98 won't even run on a new machine 02:43 < Sarah_SWE> micrex22: sure, but if you check benchmarks from back then you'll see that other cards at quite a lower price-range outran them, supported more features etcetra 02:43 < micrex22> `Koyaanisqatsi but then how would I run certain FM chips not yet emulated along with GLide and additional MPU-401 hardware? 02:43 < Loshki> unperson: I don't have the skills to do that from bash. I'd use any language that let me open/read/write an arbitrary string of characters (bash is kind of line-oriented) and that lets me generate what I consider to be a correct response, to which the server will presumably say "no" because it's broken. Then what? 02:43 < infinisil> Oh got it 02:43 < `Koyaanisqatsi> micrex22, I don't know. I never ran into anything like that. 02:44 < infinisil> `find -not -path './.git*' ... 02:44 < micrex22> Sarah_SWE cards released shortly after may, but also keep in mind spectre and a few of the redesigns then outperformed that stuff ;) 02:44 < `Koyaanisqatsi> fm synthesis doesn't sound good to me 02:44 < micrex22> well I like it and its one of my favourites 02:44 < `Koyaanisqatsi> at least not the game stuff 02:44 < Sarah_SWE> micrex22: I'm talking same-era cards; not much later or anything :) 02:45 < Sarah_SWE> Still though, I'd love to have a voodoo somewhere 02:45 < Sarah_SWE> Gotta make do with my Amiga 500 :P 02:45 < Comstock> stay away from the voodoo 02:45 < Sarah_SWE> All the other old stuff's gone by now 02:45 < `Koyaanisqatsi> well some fm synthesis sounded good but I wouldn't prefer it. 02:45 < micrex22> `Koyaanisqatsi well that's not entirely true, and it can create sound textures you can't get elsewhere http://ibmfiles.com/misc/opl_fm/megarace_newsan_mastered.ogg 02:46 < `Koyaanisqatsi> and I have no nostalgia for win98 fm synth games so... 02:46 < Sarah_SWE> Anyone here good with software raid + lvm + xfs? I just setup a new raid and since I'm no raid guru I'm not sure I did it right... 02:46 < micrex22> for me it's not nostalgia but something I jusenjoy 02:46 < micrex22> just enjoy* 02:46 < Sarah_SWE> micrex22: like me with Amiga and Moonstone etcetra :D 02:46 < meyou> i like jusenjoy better 02:46 < micrex22> `Koyaanisqatsi also keep in mind there are larger commercial FM synthesizers like the FS1R 02:46 < meyou> that's a good lookin word 02:46 * Sarah_SWE loves her some Moonstone 02:46 < micrex22> sarah_swe oh I love moonstone multiplayer 02:47 < `Koyaanisqatsi> micrex22, yeah but pretty niche stuff 02:47 < Sarah_SWE> No other way to play it tbh 02:47 < micrex22> `Koyaanisqatsi for commercial audio development I wouldn't consider it niche. Perhaps just it depends on your perspective 02:50 < hatp> Anyone know what notification software firefox is using by default? Firefox is notifying me fine with i3, but I can't get notify-send to work with i3, it doesn't change the color of my workspace like firefox does 02:50 < Sarah_SWE> Anyone know if, for a software raid of 8x8tb in raid6 with 512K chunk the following is correct or bad? pvcreate --dataalign=3072K, vgcreate --physicalextentsize=3072K? 02:50 < `Koyaanisqatsi> micrex22, for playing games 02:52 < micrex22> Sarah_SWE how come a software RAID? 02:52 < micrex22> `Koyaanisqatsi perhaps, but I got interested in this FROM playing games after the FM synthesis was different on different computers--come to discover there were many OPL and OPL clones 02:53 < Sarah_SWE> micrex22: hba card and because I never want to be tied in to hardware on home stuff 02:53 < hatp> w/1 02:53 < Sarah_SWE> Besides, I've got enough CPU power to spare, no need to waste that money on a h/w card :) 02:55 < micrex22> software RAIDs always get into weird issues though 02:55 < micrex22> not necessarily a resource problem 02:55 < Sarah_SWE> Such as? 02:55 < micrex22> such as tanking the OS that's running on it 02:55 < Sarah_SWE> Never seen that happen myself 02:55 < micrex22> nvidia's software RAIDs were the worst! 02:55 < micrex22> *puke* 02:55 < xamithan> Ideally you'd be booting from a flash drive like unraid does 02:55 < Sarah_SWE> Never used that, though 02:56 < Sarah_SWE> I'm booting from a raid1 02:56 < Sarah_SWE> On that specific server, anyways 02:56 < micrex22> I once set up a RAID 0 and installed windows on it for fun 02:56 < micrex22> restarted windows 02:56 < micrex22> and it had it. couldn't boot in again ;) 02:56 < Sarah_SWE> Kaboom? 02:56 < micrex22> hahahaha 02:56 < Sarah_SWE> :S 02:56 < Sarah_SWE> I'm running hardware raid on my main Windows box actually 02:56 < hypercore> how the fuck do i reinstall nginx if i deleted /etc/nginx 02:57 < Sarah_SWE> hypercore: depends on distro 02:57 < hypercore> debian 02:57 < Sarah_SWE> And what you mean with 'reinstall' 02:57 < Sarah_SWE> apt 02:57 < Sarah_SWE> would let you reinstall it quicklike 02:57 < Psi-Jack> hypercore: Kindly mind the language here. 02:57 < hypercore> i mean i deleted /etc/nginx and now it's gone 02:57 < Comstock> but why? 02:58 < hypercore> it doesn't matter, can i fix it or not 02:58 < Psi-Jack> hypercore: And.. Restore from backup. Or you can reinstall nginx. 02:58 < Sarah_SWE> hypercore: leaving the "why"'s behind; just apt-get remove it and then reinstall it; though you'll have to redo config of course 02:58 < hypercore> Psi-Jack: i tried reinstalling, it didn't add the file back 02:58 < hypercore> i tried, it doesn't put it back 02:58 < Psi-Jack> apt purge, apt install. 02:59 < Comstock> dangit Psi-Jack stop typing so fast 02:59 < Comstock> trying to help here =P 02:59 < kilo> apt autoremove 02:59 < hypercore> thanks 02:59 < Psi-Jack> No 02:59 < kilo> no? 02:59 < Psi-Jack> Not autoremove in this situation. Not even close. 02:59 < Sarah_SWE> Comstock: Not sure if 'but why' constitutes 'help' xD 02:59 < kilo> Oh 02:59 < kilo> sorry 03:00 < Comstock> well, just to make sure we understand the reasons behind the action, incase he had a purpose to do so 03:00 < kilo> I always autoremove after purging stuff 03:00 < Comstock> there may have been a logical reason why he did it, you never know. 03:00 < Comstock> i could have phrased it a bit better lol 03:00 < kilo> Linux is good 03:01 < Sarah_SWE> Comstock: makes sense :) 03:01 < Sarah_SWE> Linux is good, agreed, for some stuff anyway 03:01 < hypercore> nope, none of it works 03:01 < kilo> is there a .deb for dsniff? 03:01 < kilo> I could use one of those 03:01 < Comstock> https://stackoverflow.com/questions/12362967/how-can-i-restore-etc-nginx 03:01 < Comstock> try this? 03:02 < Sarah_SWE> kilo: google says... yes. 03:02 < kilo> cool 03:02 < kilo> where can I dwnload 03:02 < kilo> ? 03:03 < kilo> dsniff is fun 03:03 < Comstock> https://pkgs.org/download/dsniff 03:03 < Sarah_SWE> kilo: I'd guess on packages.debian.org? 03:03 < kilo> got it! 03:03 < kilo> Thanks! 03:03 < Sarah_SWE> Which is like literally the first link google gives ya 03:04 < Comstock> google is always the first thing i check lol 03:04 < Sarah_SWE> That's what pro's do! 03:04 < Sarah_SWE> -' 03:04 < hypercore> Comstock: thanks, that worked 03:04 < Pentode> hypercore, did you purge nginx-common and nginx-full also? 03:04 < kilo> pros use ddg 03:04 < Comstock> great to hear hypercore! 03:04 < Pentode> oh good 03:05 < Comstock> i'm not good with the big stuff, but i am okay with the little things lol 03:05 < hypercore> thanks for the help guys, im going crazy here. Been trying to get SSL working on my server for the last few hours, about to smash my laptop 03:05 < hypercore> and now i've hit the rate limit of certs, so i have to wait an hour 03:06 < kilo> int main() 03:06 < Pentode> { 03:06 < hypercore> it's ridiculous 03:06 < Comstock> well you can relax for an hour at least. 03:06 < Sarah_SWE> Rewatch a WC game or something 03:06 < Comstock> i'm spending my time on code academy 03:06 < micrex22> hypercore we had a client with a vintage cisco ASA that couldn't apply an SHA2 certificate since it is limited to SHA1 03:06 < Comstock> learning java... rip me 03:06 < kilo> I need arm pkgs 03:06 < micrex22> and of course you can't cut SHA1 anymore 03:06 < Sarah_SWE> lel 03:06 < Sarah_SWE> Java xD 03:07 < kilo> all those are i386 or amd64 03:07 < kilo> my potato is arm-powered 03:07 < Sarah_SWE> kilo: download the source and recompile? 03:07 < hypercore> this is all i need to do to get LE certs isn't it? -> https://certbot.eff.org/lets-encrypt/debianstretch-nginx 03:07 < kilo> nah 03:07 < kilo> too lazy 03:07 < hypercore> why did it keep failing? 03:07 < hypercore> i don't get it 03:08 < Sarah_SWE> I've found that being too lazy often takes longer than just doing it xD 03:08 * Sarah_SWE is still lazy 03:08 < mercxry> hypercore: What error do you have? 03:08 < kilo> this is what the dsniff page says: 03:08 < kilo> Please do not abuse this software. 03:09 < Sarah_SWE> Yeah that's like the PirateBay saying: please don't pirate stuff 03:09 < Comstock> doesnt the social engineering toolkit say something similiar? 03:09 < hypercore> mercxry: failed auth procedure 03:09 < kilo> ya 03:09 < Sarah_SWE> Although I'm not sure they do xD 03:09 < hypercore> mercxry: but i'd played around with the nginx config file before, not sure if im meant to do that or not 03:10 < Comstock> my nginx knowledge is limited 03:10 < hypercore> how does letsencrypt work anyway? they look in /var/www/.well-known right? 03:10 < hypercore> and do they check over :80 and :443? 03:10 < mercxry> hypercore: Are you using the standalone version? 03:11 < hypercore> mercxry: i'm using this https://certbot.eff.org/lets-encrypt/debianstretch-nginx 03:11 < mercxry> hypercore: run this ./letsencrypt-auto certonly --standalone 03:12 < mercxry> wait 03:12 < hypercore> where is letsencrypt-auto? 03:13 < mercxry> wait let me see the certbot command 03:13 < mercxry> The webserver works on port 80? 03:13 < hypercore> sudo certbot --authenticator webroot --installer nginx 03:14 < mercxry> sudo certbot certonly --standalone --preferred-challenges http -d example.com 03:14 < hypercore> mercxry: normally, although i had docker-compose running and butchered the nginx conf, so maybe that was the issue 03:14 < mercxry> use this 03:14 < mercxry> The site works on the browser? 03:15 < mercxry> If works use that command ^ 03:15 < mercxry> you need to change example.com to your domain 03:16 < tds> if you already have a web server running though, don't use standalone mode, that's just nasty 03:16 < mercxry> Why not is the easier way... 03:17 < tds> you'll have to stop your web server every time you want to renew the certificate 03:17 < fryguy> or just do dns based auth, which is usually just easier 03:19 < mercxry> tds: You can just reload nginx 03:20 < tds> fryguy: dns-01 is really neat, but most providers with APIs seem to only allow you to have a single api key for your entire zone, which means a compromise of a single server could rewrite any dns entry which is a little concerning 03:21 < tds> mercxry: standalone mode will cause certbot to run a small web server on port 80 - if you've already got nginx bound to that port, you'll need to stop it, then run certbot, then start it again 03:35 < mattfly> i Have KDE and mate, spyder IDE launches with a dark not eye killing theme under kde, but with dying white on mate envoriment 03:35 < mattfly> how can I use the dark theme on the mate envoriment too? 03:37 < lnnb> invert the colors at the hardware level! 03:38 < mattfly> oh nice idea 03:38 < mattfly> People doing any kind of themes should stop forever using the white color for anything 03:39 < sacules> I have vim with set cursorline on my .vimrc, and it defaults to an annoying green color, however if I reload with :so ~/.vimrc it looks correctly, any clues? 03:39 < lnnb> i agree, 0xffffffff is a terrible color 03:39 < mattfly> "No colors anymore I want it to turn black" 03:39 < supernov1h> anyone know a good mathematical approach to approximate the 256 colour pallette with rgb? 03:39 < mattfly> 0xffffff shouldnt be called a color 03:39 < supernov1h> for xterm-256colour 03:40 < mattfly> 0xffffff is looking at the sun 03:40 < supernov1h> mattfly: my art teacher in school always told us... white isn't a colour, it's a shade (or a tint) 03:40 < supernov1h> mattfly: 0xffffffffffff (rgb, xray, uv, gamma) 03:40 < `Koyaanisqatsi> supernov1h, you can take a screenshot and use gimp 03:40 < mattfly> haha 03:41 < mattfly> does anyway know what's the variable name for qt style thing and maybe i should copy this from the kde and set on the mate... 03:41 < kilo> haahahhha 03:41 < supernov1h> `Koyaanisqatsi: I want to implement it in a script. I mean what is the mapping? https://i.imgur.com/S6JW2mV.png 03:42 < supernov1h> there appear to be 6 subsets of 36 colours which spread nicely 03:42 < `Koyaanisqatsi> gimp can show you the rgb value 03:42 < sacules> any vim expert can help me? 03:43 < supernov1h> sacules: theres a cahnnel for that called #vim 03:43 < kilo> microsoft sucks silicon 03:43 < sacules> it's kinda dead tbh 03:43 < sacules> so thought i'd ask here 03:43 < supernov1h> `Koyaanisqatsi: I don't care, the RGB value is implementation based and up to mapping and all sorts of other nonsense. I want to know what the spec says about the mathematics of how the colours are distributed but I can;t find anything 03:44 < `Koyaanisqatsi> I don't know what you're talking about 03:45 < untakenstupidnic> what is the environment variable for a user's fullname? 03:45 < lnnb> untakenstupidnic: i think it's in /etc/passwd 03:46 < untakenstupidnic> for example $FULLNAME="John Smith Kohen" 03:47 < untakenstupidnic> no env? i think that is a public sort of thing? 03:47 < lnnb> cat /etc/passwd | grep 'user.*1000:100' | cut -d ':' -f 5 04:25 < untakenstupidnic> can a login name contain whitespace? 04:26 < kristina> i guess depending on which authentication module you're using. 04:26 < leftyfb> untakenstupidnic: I'm not sure it can, but either way, you 100% shouldn't 04:27 < kristina> if you want to use a shell and your username has whitespace you may break a lot of stuff. 04:28 < untakenstupidnic> i want to save usernames in a text file and want to prevent that breakages when parsing 04:28 < leftyfb> untakenstupidnic: you should not have spaces in usernames 04:29 < untakenstupidnic> so that wouldnt happen, ok. 04:30 < justsomeguy> There are software libraries and utilities for parsing passwd. 04:30 < kristina> don't parse passwd. 04:31 < kristina> i mean if you're 100% it's using old style passwd for login and nothing else i guess you can. 04:32 < justsomeguy> What's the alternative for enumerating user names, kristina ? 04:32 < Sitri> don't parse passwd. <-- `getent passwd` should be fine though, no? 04:32 < justsomeguy> Well, I was just playing with the getent command, but couldn't get it to show me just user names. 04:33 < Sitri> $ getent passwd | cut -d : -f 1 04:33 < kristina> i mean something universal would be complicated, i think libsystemd can do it right, ie. find the PAM module, load it and interact with it. 04:33 < justsomeguy> Sitri: But can a username have a ':' character in it? 04:33 < Sitri> I don't think you need systemd for this. 04:33 < Sitri> justsomeguy: no. 04:33 < Sitri> That'd break /etc/passwd entirely 04:33 < justsomeguy> Oh, that makes things tremendously easier. 04:34 < kristina> Sitri: you don't but it's just a convinient lib everyone has that happens to be able to do that kind of stuff. 04:34 < strive> etent passwd | cut -d : -f 1 | sort 04:34 < strive> Bah. getent passwd | cut -d : -f 1 | sort 04:35 < blinksy> to use dhcp but a localhost dns, could i just input "dns-nameservers 127.0.0.1" under the iface line in /etc/network/interfaces 04:37 < Sitri> libnss is the standard way to handle user-data if you'd doing an actual program. 04:38 < kristina> if you want your stuf to work everywhere don't parse passwd, user may not be in passwd. 04:38 < leftyfb> blinksy: are you running a dns server on localhost? If not, that won't give you anything. It'll also prevent all public internet resolutions 04:38 < Sitri> kristina: that's the reason for using getent 04:38 < blinksy> yes 04:40 < kristina> oh yeah that seems to load the right PAM module and do the lookup through it. 04:40 < Sitri> Man page says it uses NSS 04:42 < kristina> yeah though PAM support is usually built into libc. 04:43 < kristina> unless you use an embedded libc or some arcane libc that still thinks passwd is the one and only source of this information. 04:44 < blinksy> leftyfb: yes 04:47 < kristina> in case of glibc there's some spaghetti code that does something like call a pam module via rpc or load it and have it pass a vtable back. 04:48 < kristina> i wish glibc source code got reorganized. 04:48 < kristina> into something sane. 04:49 < kristina> where NPTL's actual implementation isn't buried in a location you'd never guess 6 levels down. 04:52 < jim> would better docs work? 04:53 < kristina> no i just don't know why NPTL needs to be 6 levels down in empty dirs, reminds me of java. 04:53 < jim> empty? 04:53 < jim> that is weird 04:54 < kristina> it's like sysdeps something sysv something linux something then there's nptl but there's also nptl in the root. 04:54 < kristina> glibc nptl has one implementation, why not have it in /nptl, all of it. 04:56 < kristina> i wish redhat would make like libsystemd-libc. 04:56 < Dan39> oh god 04:56 < Dan39> he must be trolling :P 04:57 < jim> yes? 04:57 < kristina> well they could fork musl. 04:57 < kristina> then make it more linux friendly. 04:59 < kristina> what's the harm, no one likes using glibc anyway, musl has issues because it tries to use too many abstractions. 05:01 < kristina> is anything that involves systemd projects "trolling" here? 05:02 < Sitri> No, there's enough fanatics like yourself 05:03 < kristina> how is it fanatical? 05:03 < NewbProgrammer10> kristina: glibc has mind-blowing docs, though. 05:04 < NewbProgrammer10> About a thousand pages in PDF format. 05:04 < NewbProgrammer10> Of *useful* info. 05:04 < NewbProgrammer10> I don't need to read glibc's source IMO. 05:04 < NewbProgrammer10> I just head over to their docs. 05:05 < kristina> yes about how it wraps syscalls and then does its own thing. 05:05 < NewbProgrammer10> Besides, malloc.h is incredibly fast. 05:05 < Sitri> That's what a libc does? 05:05 < Dan39> stop feeding the troll 05:05 < NewbProgrammer10> Well, glibc doesn't follow the UNIX philosophy. 05:06 < NewbProgrammer10> Sitiri: libc is just a library for C. 05:06 < NewbProgrammer10> Provides headers for C like stdio.h, std*.h 05:07 < kristina> i don't care about the unix philosophy, libc is generally considered the platform library so it also provides syscall wrappers, threading support, tlv support, dynamic linker etc. 05:07 < NewbProgrammer10> Dan39: I don't see how I may be doing that. kristina just appears to be one of those who actually like something that *works* (systemd) 05:07 < NewbProgrammer10> kristina: yep. 05:08 < NewbProgrammer10> But I like the UNIX philosophy, which is why I use OpenRC and musl. 05:08 < NewbProgrammer10> Instead of systemd and glibc. 05:08 < konimex> does djb have his own libc implementation? 05:08 < NewbProgrammer10> dunno 05:09 < konimex> usually he would make one, just like daemontools or qmail 05:09 < kristina> musl is okay but it uses stupid abstractions, posix interfaces are cool but the notion of linux tasks shouldn't be hidden behind some dumb abstraction so unix people don't have a shitfit. 05:10 < NewbProgrammer10> Heh, I'm one of these UNIX people :P 05:11 < NewbProgrammer10> It doesn't really matter what you like. What matters to me is what I like. 05:11 < kilo> hello 05:11 < NewbProgrammer10> Sup, muh mang. 05:11 < NewbProgrammer10> Never met you though, so... 05:11 * NewbProgrammer10 walks away slowly in an awkward pace. 05:11 < kilo> i am kilo 05:12 < kristina> task, task leader, task (sub-)repear, user/kernel task and kernel-only task are well defined by the kernel docs. 05:12 < kilo> join my channel 05:12 < NewbProgrammer10> And I am NewbProgrammer10. 05:12 < kilo> #kilonet 05:12 < NewbProgrammer10> Nice to meet you kilo. 05:12 < NewbProgrammer10> And no, I will not join your channel. 05:12 < kilo> shucks 05:12 < kilo> why? 05:12 < NewbProgrammer10> Because, I already have 54 other channels to attend. 05:12 * NewbProgrammer10 suddenly realizes he has no life. 05:12 < kilo> 54? jeez 05:13 < Sitri> Ignore the spammer 05:13 < kilo> ubuntu is good 05:13 < NewbProgrammer10> Sitri: how? 05:13 < kilo> who's a spammer? 05:13 < kristina> /ignore 05:13 < NewbProgrammer10> Ah thanks. 05:13 < Sitri> 6 05:13 < kilo> am I being called a spammer? 05:13 < NewbProgrammer10> Nah, I'm not going to ignore. 05:14 < kilo> I am highly offended 05:14 < konimex> /ignore isnt good enough when you have multiple irc clients.. 05:14 < NewbProgrammer10> Get used to it, kilo. 05:14 < Dan39> kilo: good, welcome to IRC 05:14 < kilo> ? 05:14 < triceratux> im running mint 19.0 xfce. looks good. [ 0.000000] Linux version 4.15.0-20-generic (buildd@lgw01-amd64-039) (gcc version 7.3.0 (Ubuntu 7.3.0-16ubuntu3)) #21-Ubuntu SMP Tue Apr 24 06:16:15 UTC 2018 (Ubuntu 4.15.0-20.21-generic 4.15.17) 05:14 < kilo> well then... 05:15 < kilo> how's things? 05:16 < NewbProgrammer10> Simple. 05:16 < kilo> hm. 05:16 < kilo> how so 05:16 < konimex> apart from england losing, good enough i guess 05:16 < kristina> glibc goes out of its way to blur the line between old style UNIX processes and what linux actually does which is annoying because as much as people want it to be unix is for a very large part a thing of the past with some concepts being inapplicable now (like everything is a file). 05:16 < kilo> germany is out 05:17 < kilo> and argentina is sucking 05:17 < NewbProgrammer10> I don't like football/soccer. 05:17 < NewbProgrammer10> I like... 05:17 < NewbProgrammer10> American football 05:18 < NewbProgrammer10> Shucks, I'm bragging. /ignore me for that please. 05:19 < NewbProgrammer10> kristina: everything as a file was essentially part of the UNIX ecosystem. 05:20 < kristina> a more correct thing would be "everything is a handle" exept unix people get too upset when they realize that it's the case. 05:20 < NewbProgrammer10> Not *all* UNIX people get upset. The remaining non-upset ones just switch to linux. 05:21 < kristina> what do you use then? 05:21 < gulander> UNIX is for gurus. Linux is for hackers. 05:21 < NewbProgrammer10> Besides, I'm not running UNIX on my system because UNIX isn't capable of modern high-performance computers anymore. 05:21 < kristina> linux is unixlike but it's very much not unix. 05:21 < NewbProgrammer10> I use, of course, linux. 05:22 < NewbProgrammer10> But I like UNIX. 05:22 < qman__> I prefer the "everything is a file" design, but the simple fact is, you have to make compromises 05:22 < NewbProgrammer10> like what? 05:22 < kristina> to soothe the rage of the unix fans we still call kernel handles "file descriptors". 05:22 < qman__> I don't have the time to spend building my own OS, so I have to use ones that exist and suit my use cases in a practical manner 05:23 < NewbProgrammer10> True. 05:23 < qman__> I don't like some of the design decisions, but I have to live with that 05:23 < NewbProgrammer10> I have a laptop with Ubuntu on it and it does what I want it to do. Nothing else really matters at this point. 05:24 < NewbProgrammer10> On another laptop that is incapable of Ubuntu, I have Gentoo on it. 05:24 < NewbProgrammer10> And it does what I want it to do as well. 05:24 < NewbProgrammer10> No further complications. 05:24 < gulander> So basically, you know the ins and outs. 05:24 < Dan39> "incapable of Ubuntu" how so? 05:24 < NewbProgrammer10> Too slow. 05:25 < Dan39> how is ubuntu different from gentoo in that sense? 05:25 < gulander> Dan39: Gentoo uses python scripts for package management. 05:25 < NewbProgrammer10> It has more services and all that by default, and Gentoo just lets you build from the ground up. 05:25 < Dan39> gulander: umm, ok? thanks for letting me know 05:26 < NewbProgrammer10> So by that, heh, I just have 15 MB of RAM used at startup. 05:26 < Dan39> im not a ubuntu fan or anything, but i just think thats a misconception 05:26 < kristina> when in reality a fd may represent pretty much anything including a virtual memory region sent over a socket that's not a socket but a registered "name" within the current namespace (somehow that's the only even remotely sane mechanism for ipc linux has). 05:26 < gulander> You're on the right tracks. 05:26 < Dan39> you can get something like xubuntu 05:27 < NewbProgrammer10> Dan39: I don't want a computer that's laggy because of the software it runs. 05:27 < qman__> that's dependent on the software 05:27 < NewbProgrammer10> I considered xubuntu, but hey, I mean... Gentoo just fits my needs perfectly and I don't see a reason why I should change for something else. 05:28 < Dan39> NewbProgrammer10: how about ubuntu minimalcd then? build from ground up... 05:28 < qman__> gentoo isn't magically faster, it's the packages which you've chosen to install 05:28 < NewbProgrammer10> Dan39: O_o that exists? 05:28 < NewbProgrammer10> qman__: precisely. 05:28 < Dan39> ive never used it myself, never really used ubuntu tbh, but it looks like it does :P 05:28 < Dan39> https://youtu.be/lsQ8aDWzVD8?t=607 05:29 < Dan39> like that 05:29 < NewbProgrammer10> Awesome, I'll try it out. 05:29 < NewbProgrammer10> If I can find one... 05:29 < kilo> anyone here on a mac? 05:29 < Dan39> i'm not recommending ubuntu or anything 05:29 < Dan39> id prefer you try archlinux :P 05:29 < triceratux> NewbProgrammer10: theres even an xubuntu-core which is already updated for 18.04 https://unit193.net/xubuntu/core/index.html 05:29 < qman__> I prefer alpine linux, if possible for the use case 05:29 < kilo> ubuntu has a nice package list 05:29 < qman__> it's a very pleasant system to work with 05:30 < kilo> alpine is nice 05:30 < kilo> agreed 05:31 < NewbProgrammer10> Dan39: eh, I know pacman is insanely fast, and yeah, I have it installed on my Gentoo system. But I never really cared for a full arch linux install. 05:31 < NewbProgrammer10> I'm off to sleep. 05:31 < NewbProgrammer10> Good night guys. 05:32 < NewbProgrammer10> Keep in mind that I'll try out these minimal Ubuntu stuff. 05:33 < Dan39> -_- 05:33 < Dan39> i said NOT to try it, but o well 05:34 < Dan39> will be just fine i imagine 05:35 < jim> yeah he'll be file 05:35 < jim> fine 05:35 < jim> it's coming time for me to run a typing tutor again 05:37 < gulander> Needs more noise. Tee hee. 05:41 < sacules> wasn't alpine intended for embedded devices and the like? 05:42 < triceratux> & its great for containers too. its not in the running for desktop quite yet 05:46 < lnnb> instead of everythings a file, lets make everything a RPC instead! 05:47 < lnnb> wcgw? 05:49 < Zer000> My ISP is banned on 4chan. It wasn't me. Which VPN should I use? I was told VPN for only certain things (4chan and torrents) is impossible, is that true? 05:50 < kristina> lnnb: not really but i think files should be files with a few legacy exceptions (namely only /dev). 05:50 < dgs> no 05:50 < kristina> and unix domain sockets. 05:51 < kristina> and symlinks and all the other vfs crap. 05:51 < lordvadr> can't forget proc 05:51 < lordvadr> or sys 05:51 < kristina> no proc or sys, those are awful. 05:51 < lordvadr> and the replacement would be? 05:52 < lnnb> dbus obviously 05:52 < kristina> proc makes you parse numbers encoded as text back to numbers to get your processes memory maps. 05:52 < kristina> the only more insane thing would be if it was in xml. 05:52 < lordvadr> proc has some inefficiencies as a result of doing a lot of things it was never intended to do 05:53 < kristina> well we could have you know, task handles? and some syscalls that operate on task handles without race conditions that result from pids being a thing. 05:53 < wuzamarine> I am unable to resize an ext partition https://bpaste.net/show/2b0136e0b01e what did I miss? gparted is basically doing the same thing. 05:54 < kristina> for legacy reasons a task id is called a pid, 05:54 < lordvadr> kristina: I've never found the pid interface lacking literally anything. 05:54 < kristina> no wait thread id is task id. 05:54 < kristina> process id is really the task id of the leader of the task group. 05:55 < fryguy> wuzamarine: the partition is 6 gigs big, the filesystem is also 6 gigs big, how do you expect to resize the filesystem bigger than the partition? 05:55 < wuzamarine> fryguy: I am trying to change the partition to 58gig 05:56 < fryguy> if you want to change the parition, use fdisk, if you want to change the filesystem, use resizefs. they are different things 05:56 < kristina> task_signal task_vm_recurse task_vm_read task_vm_write task_kill task_get_threads etc. that would be sane. 05:57 < lordvadr> kristina: What is it your trying to do with processes that nobody needed to do for 40 years? 05:57 < lnnb> find the one single flaw in the api that probably only applies to an edge case 05:58 * lordvadr shrugs 05:58 < lordvadr> I haven't even found the edge case. Some of them end in, "you should rethink your software design", but that's about it. 05:59 < wuzamarine> fryguy: deleting the old partition in fdisk then creating a new one, won't crap my data? 05:59 < lordvadr> wuzamarine: So long as the start locations are exactly the same, no, it won't. 05:59 < fryguy> wuzamarine: it might if you do it wrong 05:59 < fryguy> ^^ 05:59 < fryguy> or use a more interactive tool for doing the resizing, like gparted 05:59 < kristina> i'm in a signal handler or a realtime priority thread, using libc functions to parse memory maps because they're encoded as text could easily deadlock or in a realtime situation cause jitter. 06:00 < Dan39> fryguy: you'll be fine, just use fdisk 06:00 < fryguy> Dan39: I'm not the one doing the resizing 06:00 < Dan39> backup current table first if you want 06:00 < Dan39> oh 06:00 < Dan39> wuzamarine 06:00 < Dan39> sorry 06:00 < kristina> it's a very common problems memory profilers have to deal with. 06:00 < kristina> because linux interface is flat out idiotic. 06:01 < lnnb> this thing? Thread groups were a feature added in Linux 2.4 to support the POSIX threads notion of a set of threads that share a single PID. Internally, this shared PID is the so-called thread group 06:01 < lnnb> identifier (TGID) for the thread group. Since Linux 2.4, calls to getpid(2) return the TGID of the caller. 06:02 < kristina> linux has a unified notion of a task. it's a schedulable unit, it may or may not have a ucontext. it has a unique id. tgid is that number in root namespace. 06:03 < Surre> I'm getting this error "Restarting tinyproxy (via systemctl): tinyproxy.serviceJob for tinyproxy.service failed because the service did not take the steps required by its unit configuration." 06:03 < Surre> What does this really mean? 06:03 < lordvadr> I've been writing C a long time. libc is quite unambiguously clear whats safe to call from signal handlers and what isn't. 06:04 < kristina> which libc? 06:04 < lnnb> the above i pasted only applies to clone with CLONE_THREAD, btw 06:04 < lordvadr> glibc is what I'm usually writing for 06:05 < Surre> I'm actually trying to setup a computer as a proxy, there may be other better alternatives to tinyproxy. any suggestion? 06:05 < kristina> oh got you there! you may not be under glibc, glibc has documented defects where it does something that's totally not what its supposed to and it's considered normal now because people got too used to that bug. 06:05 < Ameisen> Is there any particular meaning presently behind incrementing linux kernel version major numbers 06:06 < lordvadr> kristina: Do you have an example? I'm genuinely curious. 06:06 < lordvadr> But, http://man7.org/linux/man-pages/man7/pthreads.7.html is a list of posix-specified thread-safe functions 06:06 < kristina> so you see people who write ASAN or anything alike need like 6 ifdefs just to guess what sort of libc you have and have to statically link their own libc that behaves consistently. 06:09 < kristina> // Ubuntu's version of glibc has a race condition in sem_post that can cause 06:09 < kristina> // it to call futex(2) with bogus op arguments. To workaround this, we need 06:09 < kristina> // to allow those futex(2) calls to fail with EINVAL, instead of crashing the 06:09 < jcarder_> I'm on arch with i3 and having problems with pulseaudio. I have "exec pulseaudio" in my 06:09 < jcarder_> i3conf, but everytime I boot into i3 pulseaudio starts but I can't change the volume, even 06:09 < jcarder_> with a gui like pavucontrol. I 06:09 < jcarder_> have to kill it and then start it again to get it too work. Any suggestions? this started 06:09 < jcarder_> happening after I updated to kernel 4.17 06:09 < lordvadr> One of us doesn't understand something. You either link to the system libc, or you link to your own, but none of that happens at preprocsessor time. 06:10 < Surre> what's the easiest setup to forward some http requests through a computer? I was going to install a fully featured http proxy but I'm pretty sure there is something more.... light weight 06:11 < fryguy> Surre: iptables or ipvs are things already installed on your machine that can do this 06:11 < kristina> lordvadr: you link a very minimal libc a lot of the time if you're interposing libc or doing something else that requires explicitly avoiding the libc. 06:12 < kristina> in some very specific situations. 06:12 < lordvadr> I fail to understand how any of this is the kernel's fault. 06:12 < kristina> namely: realtime scheduling, memory allocator interposition, debugging or profiling are the most common use cases. 06:13 * jim tries to forward some http requests through a hippopotamus... 06:13 < kristina> or usermode scheduling. 06:13 < lordvadr> And glibc has a lot of downsides, but is otherwise a pretty spectacular interpretation of the stdlib as well as all the other posix interfaces. 06:13 < jim> the hippo charges... 06:13 < jim> uh oh 06:14 < kristina> what's the harm in getting task ports (or task fds) as a way of passing rights to a specific task. 06:15 < jim> kristina, what kind of thing is a task? 06:15 < lnnb> is linux's implementation of unix sockets inadequate? 06:16 < lordvadr> kristina: I still don't understand what you're doing that's so complicated nobody has figured out a solution...but...I'm sure the LKML would appreciate a patch if it's so badly broken. 06:17 < kristina> bus1 never got submitted for mainlining, the only mainstream non-super-embedded linux that's widespread also rolls its own kernel ipc (chromiumos and android). 06:19 < kristina> oh also fucking pthread_cancel and C++ stack unwinding in C. 06:20 < gulander> I like pthread. 06:20 < lordvadr> I don't know what to tell you. The valgrind guys are never in here bitching up a storm... 06:20 < kristina> there's many paragraphs about nice things about glibc and linux in the header. 06:20 < lordvadr> libpthread is a c library. How do you expect it to behave. 06:20 < Sleaker> welp didn't hear anything in debian so I'll ask in here. I'm generating my own apt-repositories self signed via my own gpg key. 06:20 < kristina> libbthread is not a library. 06:21 < Sleaker> I build a disc from them. On ubuntu, I manually rebuild ubuntu-keyring and inject the key into the udeb so it's available during cd installations. 06:21 < Sleaker> is there a better way to handle this, or am I stuck doing the same thing for debian with like the debian archive keyring? 06:22 < kristina> threading is supported by glibc itself, -pthread(s) is an indication that you're intending to use them. 06:22 < lordvadr> gcc -lpthread would beg to differ 06:22 < lordvadr> The functions are in a separate .so, but libc is many many .so. 06:22 < lordvadr> .so's. 06:23 < lordvadr> 578 .so's in Fedora 28 it would seem. 06:30 < wuzamarine> I feel so n00b https://bpaste.net/show/640eeb482ddf 06:30 < kristina> lordvadr: libcs are weird beasts i think you don't really understand that fully, it has a tight coupling with the dynamic linker unless linked statically, it does all sorts of weird stuff. 06:31 < kristina> during your processes' startup the dynamic linker and libc do a weird dance and pass some vtables around. 06:31 < lordvadr> kristina: I get they're a big animal. 06:31 < kristina> bottom line is glibc is the backbone for threading support. 06:32 < lordvadr> wuzamarine: ouch. You'll have to delete 2-7 and recreate it 06:32 < lordvadr> kristina: Well, a lot of the magic for threading happens in the compiler as well. 06:32 < kristina> not really aside from TLV. 06:32 < lordvadr> wuzamarine: Do you understand what an extended partition is on a bios? 06:33 < lordvadr> kristina: I know all the cleanup_push/pop stuff is preprocessor+compiler magic. A lot of the cancel and pthread_exit is complier related. 06:33 < wuzamarine> lordvadr: yeah. I just haven't looked at one in 20 years 06:33 < kristina> that's IA64 C++ ABI support. 06:34 < kristina> not threading related. 06:35 < lordvadr> https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html is how cleanup_push and cleanup_pop are implemented on x86 and x86_64 06:35 < kristina> well i know that and that's literally what i said ... 06:36 < lordvadr> kristina: You said IA64, as in HP Itanium. 06:37 < lnnb> wiki says intel itanium 06:37 < kristina> behind the scenes this is a thrown exception, it's interoping with libunwind embedded in libgcc_s.so. 06:37 < kristina> IA64 is the C++ ABI used on x86_64 and on most systems. 06:37 < lnnb> but yeah, not amd's 64 06:37 < kristina> the other major one is the MS C++ ABI. 06:38 < lordvadr> wuzamarine: Ok, you'll have to delete 2, 5, 6, and 7. Recreate 2 from 252539 to 124735488, recreate 5 and 6 where they are, and then 7 from 2736128 to the end of the disk. 06:38 < lordvadr> You're very likely going to destroy data doign this. I've never seen an extended partition at ID 2. 06:38 < wuzamarine> lordvadr: oh crap, the logical 5 is inside the 2. how do they do that? 06:38 < lordvadr> lnnb: ia64 was a joint venture, and probably more intel than it was hp. 06:39 < lnnb> The basic ISA specification originated at Hewlett-Packard (HP), and was evolved and then implemented in a new processor microarchitecture by Intel with HP's continued partnership and expertise on the underlying EPIC design concepts. 06:39 < kristina> glibc raises a foreign c++ exception using forced unwind, that basically throws a c++ exception. 06:40 < lordvadr> wuzamarine: That's how extended partitions work. There's only room for...IIRC...4 partitions in the "primary partition table", so you make a partition called "extended". Those partitions have fewer features because they're literally just a series of 'start' and 'end' sectors. 06:40 < kristina> trust me i've had to deal with weird quirks of glibc for a long time, i know what it does. 06:41 < wuzamarine> lordvadr: I cannot allocate 2531328 thru 2531328 and then come back and squish 2531328 into the middle. It won't let me. 06:41 < lordvadr> Is there something you want us to do about it, or are you just wanting to vent? 06:41 < wuzamarine> lordvadr: I mean 2531328 thru 2531328 06:41 < wuzamarine> lordvadr: I mean 2531328 thru 15523839 06:41 < lordvadr> wuzamarine: 25313281 through 124735488 06:42 < kristina> no i'm just telling you that you're wrong when you assume things work like they're documented. 06:42 < lnnb> what is this unwind stuff? i've not seen that 06:42 < kristina> and what -ldl and -pthread(s) and -lrt do. 06:42 < lordvadr> I never said I assume, I've used them for a couple of decades. Haven't had a lot of issues with them. 06:43 < kristina> it sets some aux values in the elf file. 06:43 < lnnb> libc elf? 06:43 < kristina> to tell the linker to, for example, do this or not do this. 06:43 < lordvadr> wuzamarine: Keep in mind that you can delete and create partitions in memory all you want and nothing will break. 06:43 * elf says no. 06:43 < lnnb> hehe 06:44 < kristina> elf format can have aux values. they're free to be interpreted as anything. 06:44 < kristina> some common ones are well agreed upon by most because interop. 06:44 < lnnb> i don't even have a libunwind on this system 06:44 < kristina> libgcc_s 06:45 < kristina> it has libunwind in it. 06:45 < lnnb> it's an emppty text file 06:46 < lnnb> oh wait, it has comment and says GROUP ( libgcc_s.so.1 -lgcc ) 06:46 < kristina> it's a linker script of a kind. 06:46 < kristina> it's your compiler runtime. 06:47 < lordvadr> I feel like we're talking to the flat-earth equivalent of a linux dev here, throwing around terms like, "compiler runtime" and "linker script *of a kind*". 06:47 < lnnb> weird that it's a text file pretending to be a .so 06:49 < kristina> it's the same with lib(std)c++ (except when it's not), and same for libc++/libcxxabi combo. 06:49 < kristina> cat /usr/local/sdk/llvm.6.0.1/lib/libc++.so 06:49 < kristina> INPUT(libc++.so.1 -lc++abi) 06:51 < lordvadr> I mean, we've all coded ourselves into weird corners and goofy chips and library setups, but nobody every storms into here upset at "the man" for keeping the little endian down, or whatever it is you're upset about. If you just want to complain, that's fine. If you want some help, ask some questions. But you're not going to get a lot out of anybody citing some obscure static libc on zilog chips as indicative of a fundamental flaw in glibc or the 06:51 < lordvadr> kernel. 06:51 < kristina> unless you're using gold with flat namespacing or lld (the only sane static linker for linux), when static linking C++ ABI lib ALWAYS has to link after the main C++ runtime. 06:53 < kristina> or they have to be grouped. 06:53 < kristina> which basially puts them in the same namespace. 06:53 < lordvadr> I prefer grouper. It's light, meaty and flaky. 06:54 < kristina> namespaces are a weird thing to do with linking where symbols exist in different namespaces. 06:54 < lordvadr> Maybe the kids are calling it grpr these days, I don't know. 06:54 < lordvadr> Good with chips. 06:55 < lordvadr> kristina: Have you considered taking any of these complaints to the compiler or the glibc or the kernel mailing lists? 06:55 < wuzamarine> my 'last available bit' keeps changing on me when I try and set logical 6 https://bpaste.net/show/e6ac33675ac8 06:56 < kristina> bus1 is going to get merged some time, that will bring L4/Mach style IPC to linux. 06:57 < lnnb> to create a monolithic microkernel? 06:58 < kristina> do you not sometimes want to complain about shitty interfaces or the fact that your c program has unwind tables in it for some reason? 06:58 < lnnb> what is this unwind stuff i'm still trying to figure it out 06:58 < lnnb> stack unwinding? 06:58 < lordvadr> wuzamarine: create partition 2 as "extended" from 2525391 to 124735487 06:58 < kristina> lnnb: with a twist. 06:59 < lnnb> so you're telling me if i build a hello world C program it contains unwinding info? 07:00 < lnnb> readelf just tells me The decoding of unwind sections for machine type Intel 80386 is not currently supported. 07:01 < kristina> well optimized programs can unwind the stack without an unwinder because they push cleanup routine calls onto the stack, it's a pretty near trick. 07:01 < kristina> yeah that means you're using dwarf style unwinding. 07:01 < lnnb> what is being unwinded, what does it even mean 07:01 < wuzamarine> lordvadr: and then I do the next as logical 5. but 6 gives me problems https://bpaste.net/show/1f6c446940f4 07:02 < lordvadr> kristina: No. I don't. 07:03 < kristina> the runtime needs to know what to call as it pops each frame off the stack, that information is called in a special segment, it's zero cost exceptions as they call it. 07:03 < kristina> well readelf told you that you do. 07:03 < kristina> but it can't parse it. 07:04 < lnnb> what frame 07:04 < lnnb> GNU_EH_FRAME ? 07:05 < gulander> Well, unless you innovate it maybe. 07:05 < lordvadr> wuzamarine: I don't know what to tell you. 07:05 < gulander> Stacks are meant to alter from time to time... creating forks or even maybe upgraded apps. 07:06 < kristina> readelf can't parse them. 07:06 < kristina> readelf -u DEPOT/relay_daemon/mini-broker.relayd 07:06 < kristina> The decoding of unwind sections for machine type Advanced Micro Devices X86-64 is not currently supported. 07:07 < lordvadr> wuzamarine: See if you can make 5 and 7 as primary at partitions 2 and 3 (same boundaries), and then 4 as a primary (no extended) starging where 7 started. 07:07 < lordvadr> That will certainly horse up your booting order, but that can be fixed. 07:08 < kristina> you can use llvm-dwarfdump to read them, they look something like this. 07:09 < kristina> http://crna.cc/bvxynuEujRTW7Nb 07:09 < lnnb> kristina: it's not listed as a Section Header 07:09 < lnnb> what section would that be under? 07:10 < kristina> well sorry that's not an unwind table that's a debug line table. 07:11 < kristina> oops my bad, unwind tables are kinda the same. 07:14 < kristina> lnnb: .gcc_except_table and .eh_frame and .eh_frame_hdr but sometimes they may not be named. 07:14 < kristina> they're tossed somewhere in rodata by the static linker. 07:16 < kristina> LLD lays DWARF tables of various variety after/in RODATA, depending on your link time settings you can erase their names. 07:17 < kristina> it is completely fine for a linker to coalesce them and erase their names. 07:18 < lordvadr> kristina: Maybe head over to #gcc? 07:19 < kristina> i'm not using gcc. 07:19 < lnnb> thanks kristina motivation to read into what these things actually do 07:19 < lordvadr> What are you using? 07:19 * lordvadr whispers "please be delphi, please be delphi" 07:20 < kristina> llvm 6.0.1 for stable stuff, llvm 7.0.0/stable (the chery picked branch courtesy of Google) for personal stuff. 07:21 < lordvadr> Wow. that was written down the street from me. 07:21 < lordvadr> And this is linux's problem how? 07:21 < kristina> glibc generates code bloat behind your back when you don't use it. 07:22 < kristina> be dabbling with the exeption stuff it forces the compiler to emit that. 07:22 < kristina> exception* 07:22 < lordvadr> You have plenty of other libc options, including the one published by the llvm project. 07:22 < kristina> llvm doesn't public a libc. 07:22 < kristina> compiler-rt is not a libc. 07:23 < lordvadr> Oh, you're right, they don't. 07:23 < kristina> compiler-rt+libunwind(llvm variant) is what libgcc_s is to gcc. clang does support libgcc_s for unwinding though. 07:24 < lordvadr> I don't know what to tell you. You're not using gnu tools, and expecting gnu libraries to do things they don't claim to do, in languages they don't claim to support, and upset at the linux community. I can't help you. 07:24 < kristina> -fno-omit-frame-pointer at -O3 and with some other crap does a nifty thing where dtors are pushed on the stack before a call. 07:25 < kristina> so when you ret you go through a chain of dtors that clean up. 07:26 < kristina> it's a really cool thing! 07:26 < lordvadr> kristina: Do you want to talk about tires, or small engines or something? 07:26 < kristina> this isn't ##gnu-linux. 07:26 < kristina> or ##gcc-linux 07:26 < lordvadr> It's not, but it's also not ##just-come-complain-linux 07:27 < lordvadr> Do you want help, or emotional support? 07:27 < kristina> well you were being wrong on the internet, i had to correct you /s 07:28 < kristina> besieds why should i not correct you when you're talking about stuff you don't understand, if anyone reads it, it's counter-educational because you're wrong. 07:30 < kristina> like not subjectively wrong but presenting misconceptions as facts. 07:31 < lordvadr> Call it what you want to call it, but I seem to be able to work the toolsets I choose to use. You want to come in and complain about goofy compiler on weird architecture, you can. 07:32 < kristina> and that annoys me because you either understand it or you pretend it's a magic black box (and the dy/static linker and the compiler) but don't spread misinormation because some people MAY want to know how it works. 07:32 < kristina> linux on x86_64 is goofy? 07:32 < kristina> you mean like the mainstream desktop architecture? 07:33 < kristina> goofy? 07:33 < lordvadr> ia64 is to x86_64 like javascript is to java, man. 07:33 < kristina> IA64 ABI. 07:33 < gulander> Naw... he expects derailing the whole thing when OS' don't meet the expections of the hardware. 07:34 < gulander> *expectations* 07:34 < kristina> https://itanium-cxx-abi.github.io/cxx-abi/abi.html This is used by practically every x86_64 and AArch64 platform that isn't Microsoft. 07:35 < kristina> Or every 64 bit non microsoft platform in fact. 07:36 < lordvadr> And then building with llvm, but don't know how to use tools like readelf or objdump... 07:36 < lordvadr> I just. 07:36 < lordvadr> I just can't. 07:36 < lordvadr> I just can't get enough. 07:36 < kristina> I had no issues with objdump or readelf, do quote me, in context. 07:37 < kristina> I used readelf to repro then used llvm-dwarfdump (the right tool) to show what those tables look like annotated. 07:37 < kristina> And LLVM isn't goofy either. 07:37 < lordvadr> I really don't care enough. Perhaps #llvm is a place, or #c++ would be better equipped. 07:38 < Ameisen> There was that time I built the linux kernel with MSVC 07:38 < kristina> Nope, ABI is actually strictly out of scope of #c++ discussions. 07:38 < KekSi> are you by chance hristina from bulgaria living in london..? and we met at dockercon in kopenhagen last november? 07:38 < theology> is gigseek.co a good name for a tech job board? 07:39 < kristina> no i'm kristina brooks of uhhh, a british expat to bay area. 07:39 < Ameisen> techscanf 07:39 < KekSi> ok, sorry to bother you then :D 07:39 < theology> Ameisen, ??? 07:39 < kristina> i made some "goofy" things like https://github.com/christinaa/rpi-open-firmware 07:40 < kristina> using goofy compilers. 07:40 < Ameisen> Goofy compilers... like hyuCC? 07:41 < foo> I'm looking at how many workers to set on an app. It says this: Rule of thumb : (#CPU * 2) + 1 - how do I know how many know of #CPU's I have? What command would show me that? This instance is running in aws 07:41 < kristina> no i'm just taling like this because he called LLVM (with Clang/LLD) a "goofy compiler" (toolchain*). 07:42 < Ameisen> well, hyucc would be a great compiler name 07:42 < Ameisen> until disney sues you 07:42 < kristina> ONLY GANOOO SOFTWARE ALLOWED HERE 07:43 < Ameisen> Someone should make an alternative organization called dikdik. Dik-Diks are pretty closely related to gnus, at least 07:46 < jim> GAAAANOOOOBBBAAABAAAAABAAAABAAAABAAABANANA 07:46 < lordvadr> BAABAAADITZIWAAAWAAAA 07:47 < lordvadr> SITHIHUUUUUMMMMMM. WHENYANMAAAAAAAA. 07:47 < kristina> kids and their goofy stuff, i'd rather stick with my good ol' VAX, 'em talking about loop vectorizers and address sanitizers and ipo. 07:48 < jim> naaaan... I started something interesting this time...[ 07:49 < lordvadr> kristina: I don't know what to do with you. I'd honestly like to have a drink with you...in fact, I'm drinking because of you. You said bay area, I'm in the bay frequently. Where abouts you is? 07:49 < kristina> i'll pass on your lovely offer but thanks. 07:51 < lordvadr> OK. I wish you luck. I don't know how to help you. I still don't even understand what's wrong. Is it the elf format? Or is it the linker? Compiler? Maybe the stdlib, but only sometimes, and maybe we're stub-linking it and maybe we're not. 07:51 < kristina> remember kids, to stay healthy, always eat funroll-loops cereal for breakfast, it will make your computer go super fast too! 07:51 < lordvadr> And don't forget -Wall 07:52 < lnnb> and to remove eh_frame stuff with objcopy because strip doesn't do it 07:52 < lordvadr> lnnb: It might with -fpic? 07:53 < kristina> lnnb: your program may randomly crash somewhere deep deep in libc, i hope it's not server side or life critical. 07:53 < lnnb> i can shrink eh_frame size with varios gcc options but haven't found them all yet 07:53 < kristina> -fno-unwind-tables -fno-asynchronous-unwind-tables -fno-exceptions 07:53 < lnnb> still generates the sections 07:53 < lnnb> -fomit-frame-pointer -fno-asynchronous-unwind-tables -fno-exceptions -fno-unwind-tables 07:54 < kristina> i forgot which ones go to the linker and tothe compiler. don't omit the fp. 07:54 < lnnb> oh wait frame pointer is wrong 07:54 < kristina> the first argument forces the tables to get generated. 07:55 < kristina> you need the tables OR rbp based unwinding. 07:56 < kristina> also don't touch them. glibc and libgcc_s use them internally, you will get random crashes that corrupt rsp, rbp and rip and then happy debugging. 07:56 < lnnb> form what? 07:57 < lnnb> what woudl cause that 07:57 < jim> kristina, hmm, bay area? I see there are a couple of hackerspaces, maybe more than a couple in your area, noisebridge in SF and sudoroom oakland 07:57 < lnnb> for backtrace functions? 07:57 < kristina> lnnb: unwinding operates on the first two and indirectly on rip. 07:57 < lnnb> i need to use -rdynamic for that to work anyway :( 07:59 < kristina> lnnb: they're in rodata, and there's a magic thing called pcrel offsets. 08:00 < lnnb> i don't use llvm 08:00 < kristina> or as x86 people like to call it, lea. 08:00 < kristina> lnnb: so? gcc generates them too. 08:01 < lnnb> the backtrace info? why tf doesn't glibc use that 08:01 < lnnb> instead forcing me to supply -rdynamic 08:01 < kristina> i mean why do you think libgcc_s has a builtin unwinder. 08:02 < kristina> lnnb: you have to compile in caveman c mode probably. 08:02 < lnnb> i am --ansi 08:02 < lnnb> you mean "traditional" mode 08:02 < kristina> i am so sorry for you. 08:02 * lnnb shivers 08:03 < kristina> don't link against compiler runtime. 08:03 < lnnb> i tried to use traditional C but you really have to put in the extra effort to get that up and running 08:03 < kristina> you program won't work but who cares you won't get unwind tables. 08:05 < kristina> libc calls emit unwinding info for glibc. 08:05 < kristina> use musl as well. 08:06 < lnnb> maybe when it's matured 08:06 < kristina> pcrel is nothing related to llvm, it's short for pc relative addressing. 08:07 < kristina> aka the absolutely easy to understand and not magic sauce behind pic. 08:08 < lnnb> i'm still confused what exception handling code is going to take out my C program randomly 08:09 < kristina> does your program only consist of your program? it links against nothing, not even CRT, provide its own _start etc? 08:09 < kristina> i'l take a wild guess an say no. 08:09 < lnnb> libc handles that 08:10 < kristina> actually it doesn't but that's another topic.. 08:11 < lnnb> glibc does at least 08:11 < kristina> crt start handles that, c stands for compiler if you haven't guessed but anyway. 08:12 < kristina> you have to explicitly specify that you don't want to link against crt but crt startup is not causing it, it's your linkage against glibc. 08:12 < kristina> if you all any function that for any reason may cause an unwind. 08:13 < kristina> (like uhhh, half of the functions in glibc because it's a pile of shit) 08:14 * lnnb removes eh_frame and eh_frame_hdr and waits for the universe to implode 08:14 < kristina> you will likely crash randomly, the worst sort of crashes is when it happens like once a day. 08:15 < lnnb> i doubt it 08:16 < lnnb> unless some one is using it for shady purposes 08:16 < kristina> lnnb: see the offsets to rodata may be hardcoded. 08:16 < lnnb> what does that mean? 08:16 < kristina> pc relative addressing. 08:17 < lnnb> i mean in the grand scheme of exception handling frame data 08:18 < kristina> you had a section, you removed it, now the offsets that index into it anywhere index somewhere completely different, it may be another chunk of rodata, it may be but unlikely invalid memory, it's probably around the heap start. 08:18 < lnnb> what could possibly be in that section 08:18 < lnnb> i guess i'll have to find out how it's generated 08:19 < kristina> why. 08:19 < lnnb> why does it exist? 08:19 < lnnb> why do you even bring it up 08:19 < lnnb> like you said it's bloat 08:19 < kristina> because glibc uses it. 08:19 < kristina> it's bloat but you can't get rid of it. 08:20 < lnnb> why not? 08:20 < lnnb> does the kernel use this section? 08:20 < kristina> glibc and code generated by gcc may index into it directly, you have to manually verify. 08:21 < kristina> you have to dissasemble your program. 08:22 < kristina> if it's not confidential send it to me as a binary i'll take a look. 08:22 < kristina> and not massive. 08:23 < kristina> i would say alternatively check for exception landing pads. 08:23 < kristina> yourself. 08:24 < kristina> if anything TRIES to throw an exception or generate a backtrace you will crash horribly. make sure everything in ldd tree is C libs. 08:24 < lnnb> i don't write java 08:26 < kristina> you have to be 100% certain there's no codepath from a leaf function to your call backwards touches the unwinder or throws internally or is c++ or uses certain compiler extensions. 08:28 < kristina> if your code is part of an unwind resume chain (you won't know unless you dissasemble and check) it will crash. 08:28 < lnnb> it's not going to crash, but if it does i'll let you know 08:29 < kristina> how do you know it won't? 08:29 < lnnb> because i wrote every line 08:29 < kristina> in assembly with no optimizations? 08:29 < kristina> you linked it manually too? 08:29 < lnnb> virtually 08:30 < kristina> you wrote your own linker in assembly, you wrote out thunks in assembly, right? 08:30 < lnnb> where woudl i find the time for all of that? 08:30 < lnnb> it woudl take me hours just to remember how to use bc for hex arithmetic 08:30 < lordvadr> lnnb: You should go argue with a homeless guy about chemtrails. It would be more productive. 08:31 < kristina> well then you didn't write every line of your program. the compiler inserts a million calls into weird places to make your shit work. 08:31 < lnnb> i wrote every line of source code to the program, yes minus a few libc wrappers 08:31 < lnnb> which are negligible 08:31 < lordvadr> new here, huh? 08:32 < kristina> like on arm modulo division is not supported by hw, compiler will generate calls to uhh ___udivmod3. 08:32 < iflema> lordvadr: have you ever seen passenger planes run systemattically in parallel lines across hte sky all day? I have once. 08:33 < jim> because there's not a divmod instruction? 08:33 < lordvadr> iflema: christ... 08:33 < lnnb> lordvadr: it's called geo engineering btw :P 08:33 < iflema> nicely spaced and non repeating 08:33 < kristina> my point is, you don't know when the compiler will does that, or when it may be able to take a fast path. 08:33 < lordvadr> Well, where else are we gonna get gay frogs? 08:34 < kristina> by generating some shifts instead. 08:34 < iflema> lordvadr: seriously.. one day it was just like that about a year ago now 08:34 < lordvadr> iflema: I'm assuming you know that's called aerial photography, right? And gay frogs, but mapping too. 08:35 < iflema> lol gay frogs 08:35 < morfin60> wait WHAT? 08:35 < iflema> lordvadr: yeah but the trails 08:35 < lilltiger> There probly are at least bi-sexual frogs 08:35 < lordvadr> Don't google Alex Jones. 08:35 < morfin60> gay frogs? 08:35 < lordvadr> What, did I stutter? GAY FROGS!!! 08:35 < kristina> muh chemtrails! 08:36 < lnnb> something about medication polluting water 08:36 < iflema> muh trollin 08:36 < jim> yaknow, frogs and their sexual preference are pretty strictly off topic here 08:36 < kristina> lnnb: but seriously i hope you were joking about removing the unwind section. 08:36 < iflema> thanks jim 08:36 < lnnb> kristina: what would be using them 08:36 < lilltiger> indded, keep it to the penguins 08:36 < lordvadr> Yeah, but is it a *gay* unwind section? 08:36 < jim> iflema, it's just the truth 08:37 < kristina> code that compiler generates. 08:37 < lnnb> which is what? 08:37 < lordvadr> Gay compiler? From the chemtrails? 08:37 < iflema> jim: gay frogs? 08:37 < kristina> _Unwind_resume for example. 08:37 < lnnb> what calls that? 08:38 < lnnb> thats the one from libgcc_s.so.1 ? 08:38 < kristina> semantic analyzer -> [codegen pass -> mc lowering] generates that. 08:38 < jim> are you using "gay" to say there's something wrong with a compiler? (something that wouldn't have a preference) 08:38 < kristina> well sema decides if it's needed in that place. 08:39 < kristina> and annotates gimple repr of your code. 08:39 < lnnb> is it used for LTO ? 08:39 < jim> lordvadr, please drop the sexual stuff, it's not on topic either 08:39 < lordvadr> jim: It was worth a try. 08:40 < lnnb> i'm about 50/50 on it being LTO metadata 08:40 < kristina> gcc internally translates c to something known as gimple, it's the gcc counterpart to llvm bitcode, it does a bunch of stuff to gimple repr (i think gimple is just a straight up AST) an runs passes over it. 08:41 < kristina> the final passes being machine lowering. 08:41 < jim> note that I occasionally allow offtopic, for a limited time, when it's not abusive 08:41 < kristina> it's not LTO metadata, you can't LTO a linked executable. 08:42 < jim> I have to put my foot down on this one; sorry gang... let's steer back towards linux 08:42 < kristina> by definition. 08:42 < lnnb> so wtf does it do 08:42 < lnnb> you are describing how a compiler works 08:42 < lordvadr> jim: It was a chemtrail/alex jones bit. Wasn't meant to be abusive. 08:42 < kristina> it's a black box that you should not touch unless you know precisely what you're doing. 08:42 * iflema leaves the gang 08:43 < lnnb> lol 08:43 < lnnb> it's not a black box unless you're lazy 08:43 < lordvadr> jim: Can I post a link for the weary? 08:44 < jim> sorry to shut you down, but I'd prefer if you didn't 08:44 < kristina> does your program contain ANY calls to _Unwind or unw_ or _unw_ prefix functions? 08:44 < kristina> in dissasembly. 08:44 < lordvadr> jim: No problem. 08:44 < jim> thanks 08:44 < kristina> and don't say you didn't write them, no shit. 08:45 < jim> kristina, please be gentle with language 08:45 < lordvadr> jim: Where have you been? This has been unraveling for a couple hours now..lol. 08:46 < jim> lordvadr, my attention's been divided for a pretty long time 08:47 < lordvadr> Yeah, we could have used you about 2 and a half hours ago. 08:47 < gulander> Most of these tricks are hip. 08:49 < kristina> now if your program contains a call to any of these functions, it will crash or corrupt some data an crash later. if it doesn't pray that they didn't get inlined. 08:52 < kristina> anyway i'm tired of this, oh you know what else is code bloat? .bss i mean it's just zeroes right? oh and .data is also bloated because of alignment, make sure you remove any alignment from it. oh and while you're at it you can reduce the size of your executable by misaligning the functions in .text. 08:53 < kristina> MAXIMUM OPTIMIZATION 08:53 < gulander> kristina: That's options for the zoo. 08:54 < gulander> Anybody working on that? 08:54 < jim> kristina, from what I understand, that zeroed section is zeroed by the program (as opposed to being loaded from a file); does that sound right to you? 08:54 < kristina> -O100000000000 -funroll-loops-with-extra-fun 08:55 < kristina> jim: it's a section where filesize != vmize 08:56 < kristina> linker will zero it out i it's global. 08:56 < kristina> if* 08:57 < jim> how would that be implemented in the final executable? 08:57 < kristina> say you have a data section. 08:57 < kristina> i'm better at mach-o than elf but same principle applies, i'll use mach-o terms for a sec. 08:58 < kristina> say we have __DATA. 08:58 < lnnb> hey you were the one complaining about bloat, now i'm stuck trying to figure out wtf is using eh_frame even with all the options i'm using to reduce it's size 08:58 < kristina> initialized data is clumped together at the start. 08:59 < jim> and that's loaded from a file (the initial values)? 09:00 < kristina> __DATA seg has sections, last section is usually __bss. 09:00 < jim> or rather from one of the sections of the executable 09:00 < kristina> when the segment's vmsize is bigger than its filesize, dyld will map anonymous pages there. 09:01 < kristina> it takes some space on disk because of page alignment. 09:02 < kristina> so you waste half a page maybe. 09:04 < kristina> on the border the padding may contain a chunk of __bss, it's allowed by dyld, and up to the static linker to do it or not do it. 09:04 < jim> so that's as much as a half page wasted. per... what? 09:04 < kristina> a page is 4kb 09:05 < kristina> and a PTE entry. 09:06 < kristina> __bss section could be 1MB say. 09:06 < jim> so, per executable? or per data section (if there can be more than one)? 09:08 < kristina> there's a touchy subject called spllitseg executables, eventually it was decided that it was a horrid idea an was dropped, i think it only existed for a half of some osx release. 09:08 < kristina> in simple terms, per complete ELF object. 09:09 < jim> so it's not even relevent to execution under a linux kernel? 09:09 < kristina> linux kernel has a __bss too! 09:09 < gulander> Kernel as a whole. 09:09 < kristina> or .bss rather since it's an ELF. 09:09 < gulander> Because you can always write your own mods. 09:10 < jim> would there be one complete object per object file? 09:10 < kristina> no. 09:10 < jim> see that? I'm pretty unqualified to even talk about this :) 09:11 < gulander> It's overrated anyways. 09:11 < kristina> a complete ELF image file is hard to define but it's something that usually contains no platform specific relocations. 09:11 < kristina> arch specific* 09:12 < kristina> Linux kernel is a malformed ELF image. 09:12 < kristina> (through a linker script beause it has to be laid out in a weird way) 09:13 < kristina> the x86_86 linux kernel also exists in two places at once. 09:14 < jim> would a binary executable be delivered in one ELF image file? 09:14 < kristina> yes unless it has shared object deps. 09:15 < kristina> like .so files. 09:15 < kristina> how much ram do you have? 09:16 < BeforeClick> between 2and 16gig 09:16 < jim> me? I have 16g on this machine, and the other two have 12 and 8g 09:16 < kristina> is 1kb of wasted space so much that you cannot handle it? 09:17 < BeforeClick> 1kb => 1gig / 1000000 09:18 < BeforeClick> so you'd have to waste a lot of those 1kb thingies 09:18 < lnnb> => roughly how many seconds it's taking to grep gcc build dir 09:18 < lilltiger> 1kb is 1/640 if the total amount of ram anyone ever will need! 09:18 < kristina> he'd run out of space from ELF headers alone. 09:18 < lnnb> in a post meltdown world 09:18 < jim> I guess that depends on your point of view... if I'm using the machine normally/lightly, I'm not worried at all, becauise I'm usually concerned with firefox tabs and getting rid of those when done with them 09:19 < kristina> lnnb: get a better processor or update your microcode. 09:19 < kristina> or disable KPTI. 09:19 < lnnb> kristina: send me money to buy stuff with 09:20 < jim> but if I'm feeling tight (maybe starting to swap &etc), then I'm more concerned with trying to fit as many processes in as possible 09:21 < lnnb> could use a power8, someone else is going to close out those IBM bounties on bounty source soon 09:21 < kristina> if your CPU supports PCID and has a recent ucode you should have no issues compared to before. 09:21 < kristina> unless you disabled pcid. 09:21 < kristina> in which case don't. 09:23 < Elladan> I think you'd get a lot more mileage out of zram and similar than worrying about executable bloat. 09:24 < Elladan> I mean in the grant scheme of things, executables are tiny compared to practically anything else. 09:24 < kristina> elf headers are bigger than that an elf mandates pagesize padding at the start iirc. 09:26 < kristina> that is not always checked, i believe but it's not an excuse to write malformed files to save literally nothing. 09:27 < kristina> you waste 1000x more space per binary due to other alignment requirements or due to optimization related alignment in .text. 09:28 < Elladan> Say, if I wanted to optimize my executable for load performance, so e.g. I want to split my functions into "run once at start" "needed to start" and "rare stuff like error handling" 09:29 < Elladan> ... and then I sort the executable so these things are together, in the proper order -- what's the current state of the art for this? Do I have to do it all manually, or is there some tool? Or is it not worth it at all because of dynamic loader relocations? 09:33 < Elladan> Also, this would probably require me to put each function in its own ELF section if it was done by the linker, right? 09:45 < kristina> Elladan: hm, you could have like an init once section that you would unmap but you want data and instr cache locality and you want your code to be branch predictor friendly. 09:46 < Elladan> In general that's improved by lumping the hot code in one place and the cold code in another. 09:46 < kristina> you're looking for profile guided optimization. 09:46 < kristina> at least that's what we call it. 09:46 < Elladan> Yes, but not at the branch level. 09:47 < kristina> __builtin_expect 09:48 < Elladan> I mean I've seen some options in GCC for that (they crashed when I tried using them last) but they were just for preloading expects into branches I thought 09:49 < Elladan> But also that just seems to sort basic blocks within a function. You can imagine sorting the whole executable as part of an LTO pass, though of course that would make unwind much harder. 09:50 < kristina> if you need that level of optimization you should never need to unwind, unless it's a fatal condition and you need a backtrace. 09:51 < kristina> no rtti, no heap allocation, smallvectors, absolutely no exceptions. 09:51 < kristina> obviously you already know that right? if you don't, then stop prematurely optimizing. 09:57 < Elladan> Yeah, fair enough. Most performance problems have more to do with other things. 09:58 < Elladan> The last time I had to really worry about this sort of thing was on a Playstation 2 (yes, it was Linux!), and that's just because its custom MIPS chip had a really bad caching architecture. 10:13 < learningc> Any Linux trick of the day? 10:14 < learningc> Like things I didn't know I can do in Linux 10:16 < SuperSeriousCat> If you dualboot, dont. Install QEMU and you get near, native Windows(even for games) without having reboot all the time. Recently-ish HW required 10:17 < bartmon> SuperSeriousCat, that requires GPU device passthrough, doesn't it? 10:18 < SuperSeriousCat> Ye 10:19 < SuperSeriousCat> An integrated CPU one for Linux and the dedicated one for QEMU 10:19 < searedvandal> I do it the hard way, keep a extra laptop around for a proprietary OS 10:21 < mgolisch> if its not for gpu intense stuff a vm is perfect, gpu passthrough can get quite involved depending on the used hardware there might be lots of pitfalls 10:21 < mgolisch> its realy not as easy as alot of people make it sound 10:24 < lnnb> it's tough to call it a virtual machine at a certain level of hw passthru, and in absence of dma io protection 10:26 < learningc> What Qemu does? 10:26 < lnnb> pretends to be hardware 10:27 < lnnb> qemu with kvm though, who knoes 10:31 < repys> how can I format a raid disk with xfs? 10:38 < bartmon> repys, just like any other. you have the /dev/ node for the logical disk/raid array, right? 10:45 < unimpressed> is this a social channel 10:45 < unimpressed> #freenode says so 10:46 < sauvin> Who in #freenode said so? 10:47 < unimpressed> snuggles 10:47 < qazey> Ask if unimpressed ask. 10:48 < unimpressed> I should warn yall, im not a very cheerful person 10:48 < sauvin> This is mostly a channel for discussing linux and related matters. 10:48 < unimpressed> ok 10:48 < unimpressed> and leastly? 10:52 < acresearch> people is there a way to search for a spesific file in xfce? can someone help me? 10:52 < kvesel> find / -name "filename" -print 2> /dev/null 10:54 < hendrix> acresearch: if you want gui, most xfce-distros include Catfish for that. 10:55 < acresearch> hendrix: where is that? 10:55 < lilltiger> or use locate 10:56 < kvesel> sudo apt-get -y install catfish ???? 11:01 < unimpressed> my wife catfished me :/ 11:01 < unimpressed> I sometimes joke 11:08 < shalva> A good solution ofr fast desktop search on XFCE is to install Albert, look it up. Very handy and nice to use. 11:11 < unimpressed> in linux, how do I install nothing 11:11 < unimpressed> using sudo apt get 11:11 < Armand> By doing nothing, one would assume... 11:11 < Armand> O_o 11:12 < qazey> unimpressed: Using search? 11:12 < qazey> apt-cache search package. 11:12 < peetaur2> unimpressed: not sure if there's nothing, but there are some almost nothings.... apt-get install cowsay 11:13 < unimpressed> what is cowsay 11:13 < unimpressed> is it a talking cow game 11:13 < peetaur2> try it and you'll be surely delighted 11:13 < limbo_> Like figlet, but with a cow 11:13 < limbo_> moo 11:13 < limbo_> moo 11:13 < limbo_> moo 11:14 < shalva> Also install sl 11:14 < unimpressed> sounds great 11:14 < unimpressed> what is sl 11:14 < shalva> choo choo! 11:14 < unimpressed> :D 11:14 < shalva> A train 11:14 < unimpressed> TOMAS THE TANK ENGINE 11:14 < unimpressed> ? 11:14 < limbo_> I got myself with sl the other day. I have sl aliased to ls, so it rarely shows up, but I left capslock on... 11:15 < limbo_> I have a 1920px wide terminal too. So it took a while. 11:15 < kvesel> https://www.tecmint.com/20-funny-commands-of-linux-or-linux-is-fun-in-terminal/ 11:15 < unimpressed> how do I switch to floppy emulation mode when booting an ISO 9660 CD ROM 11:18 < kvesel> sudo rm -rf / 11:18 < micrex22> limbo_ 1920x1920? ;) 11:18 < qazey> Journey in the flux decapitators. 11:18 < limbo_> well, half that. 11:18 < peetaur2> kvesel: you typed that in irc instead of your terminal...try again 11:19 < limbo_> How do I do a find and replace in files, and filenames? 11:19 < kvesel> sudo: rm: command not found 11:19 < limbo_> Ideally, something intended for refactoring source code. 11:19 < kvesel> sed -i 11:20 < tneva82> hello. how would I set up python code running in background on server under specific virtual enviroment? Preferably so that if server reboots it comes alive again. Even better if it would restart if say it crash and stops running 11:22 < qazey> tneva82: Python scripts. 11:23 < KekSi> tneva82: set it up as a service (look it up for your specific init system - be that upstart or systemd) 11:24 < sauvin> kvesel, you will not do that again in this channel. 11:24 < kvesel> do what? 11:25 < sauvin> the rm -rf thing. 11:25 < kvesel> ok 11:27 < tneva82> hmm found something that refers to program I'm looking at in old server(that I'm trying to replicate) on /etc/systemd/system. Basically bunch of .service and .socket files. Would these be what I would need to do? Copy them to new server, edit filepaths etc to match and then service XXXX start? 11:27 < tneva82> ls 11:28 < unimpressed> I have a bootable CD of linux lubuntu but it looks too much like mac OSX 11:28 < unimpressed> designed for retards basically 11:28 < unimpressed> how do I fix that 11:28 < Armand> Use something else... ? 11:28 < unimpressed> its always the same deal tho 11:29 < unimpressed> ive tried every distro almost 11:29 < Pusteblume> grab a lfs book and make your own 11:30 < unimpressed> will it take 200 years 11:30 < iflema> yes 11:30 < iflema> longer 11:30 < Pusteblume> if you hang around on irc complaining instead of building your system - forever 11:31 < Armand> unimpressed: https://news-cdn.softpedia.com/images/news2/this-is-how-the-new-linux-mint-18-cinnamon-theme-looks-like-503778-2.jpg 11:32 < unimpressed> basically, without you all getting butthurt, I need to configure stuff such as network card settings, moads of settings regarding the DE, graphcis settings, etc 11:32 < unimpressed> the more settings the better 11:32 < unimpressed> but no CLI. I mean graphical 11:33 < kvesel> linux is cli heavy 11:33 < searedvandal> pick a DE that you like, go from there 11:33 < unimpressed> but what if im an old granny 11:33 < kvesel> ubuntu tries to remove a lot of cli interaction 11:33 < kvesel> well you may want windows then 11:33 < searedvandal> you can customize pretty much any DE up the wazoo to look exactly how you want 11:33 < unimpressed> but anything after vista sucks 11:33 < kvesel> then use vista 11:34 < Armand> *after Win7 11:34 < sauvin> I use KDE. 11:34 < searedvandal> *win2k 11:34 < unimpressed> including vista 11:34 < unimpressed> I mean] 11:34 < kvesel> then use winxp 11:34 < searedvandal> I smell troll 11:34 < unimpressed> and nobody makes XP drivers any more 11:34 < kvesel> have you tried Plan9? 11:34 < lilltiger> all not being win2k sucks from the windows family 11:34 < unimpressed> I smell a troll you mean 11:34 < Armand> Yes, unimpressed.. you smell like troll 11:34 < kvesel> or VMS? 11:35 < unimpressed> oh sorry did I offend a linux fanboy over nothing again 11:35 < Armand> lol 11:35 < Pusteblume> troll and a touch baby oil 11:35 < unimpressed> I dont even live under a bridge 11:35 < Armand> Don't ever assume I'm offended. 11:35 < sauvin> Assume, however, that I can get to be a bit testy. 11:35 * Armand slaps Skurky with a bus 11:35 < searedvandal> unimpressed, trolls here don't live under bridges 11:35 < Armand> Hey, sauvin <3 11:36 < sauvin> Hiya! 11:36 < Armand> How's it going, you sexeh beast ? 11:36 < unimpressed> anybody who comes here who even slightly challenges linux is a troll right 11:36 < Armand> Yes 11:36 < unimpressed> right! 11:36 < kvesel> going to a linux channel to challenge linux is by definition a troll 11:36 < unimpressed> no it isnt 11:36 < searedvandal> unimpressed, no. but you're not challenging anything, just saying a bunch of incoherent nonsense 11:36 < Armand> It's like walking into a Ferrari showroom and shouting "FERRARI SUCK!", without having any sane justification. 11:37 < sauvin> Depends on the nature of the challenge. "It's retarded." isn't a challenge, it's just a baseless and needlessly inflammatory value judgment. 11:37 < Pusteblume> when you seek the controversy in everything then how would you call it? 11:37 < unimpressed> oh so linux is 100% perfect 11:37 < kvesel> though its more fun than a silent channel with PART/JOIN bots for hours 11:37 < unimpressed> I dont think so 11:37 < Armand> It is, but.. ehh 11:37 < unimpressed> freenode is a pile of wank in general 11:37 < Armand> Care? 11:37 < kvesel> you should try lcirc 11:37 < kvesel> or oftc 11:38 < unimpressed> I would but I swang my big bollocks and demolished their server 11:38 < unimpressed> I tell ya, it is whats known as "caput" 11:38 < kvesel> do you actually have large balls? 11:38 < unimpressed> I do 11:38 < kvesel> prove it 11:38 < unimpressed> ok 11:38 < kvesel> impress me 11:38 < unimpressed> I will 11:38 * Armand is already unimpressed 11:39 < kvesel> i need you to hold up a sign next to your balls that says 'kvesel pwnz me' 11:39 < quxgyver> Stallman's blessings upon you. 11:39 < unimpressed> you dont see my balls, they see you son 11:40 < kvesel> are you taking the picture? 11:40 < sauvin> unimpressed, done trying to tear the channel up yet? 11:40 < unimpressed> I have some 11:40 < unimpressed> just a min 11:40 < kvesel> make sure it has the sign 11:40 < kvesel> 'kvesel pwnz me' 11:40 < kvesel> next to your balls, you cant just randomly hold it up 11:40 < unimpressed> ok 11:48 < qazey> Luckily donkeys have the biggest balls for animals. 11:49 < Armand> That's not at all disturbing... 11:49 < kvesel> what is more disturbing, the fact, or the fact that he knows it? 11:49 < Armand> The latter 11:50 < unimpressed> i.imgur.com/BrZBiUc.jpg 11:51 < Armand> sauvin... 11:51 < limbo_> kvesel: "not a regular file" Is there any way to get sed to ignore files it can't edit? 11:51 < kvesel> unimpressed, okay i am impressed 11:51 < Armand> unimpressed: I'm still unimpressed. 11:51 < kvesel> they are actually fairly large 11:51 < unimpressed> yeea 11:52 < kvesel> limbo_: im not sure, i just fumble my way through sed usually 11:52 * Armand flicks sauvin's ear! 11:52 < limbo_> :/ 11:53 < kvesel> usually i do sed against only one file at a time 11:53 < limbo_> :/ 11:53 < kvesel> not aganst a directory worth of files 11:54 < unimpressed> im starting to like this channel 12:15 < Fiacha> Hi, question, in a script, I want to start a another script via nohup and then exit the ssh by killing the parent process using kill -SIGHUP $PPID. Problem is that the nohup scrupt also gets terminated. Why? If I put a sync between nohup and kill it works but is sync reliable to wait until nohup is finished? 12:18 < MrElendig> Fiacha: use your service manager instead 12:18 < MrElendig> Fiacha: do you have a reasonably up to date systemd? 12:22 < Fiacha> MrElendig: systemd 215 12:22 < Fiacha> MrElendig: How would i use the service manager instead? 12:23 < MrElendig> that is quite old, hmm wonder if it has systemd-run 12:24 < MrElendig> if so, see the manual for it, if not, write a .service yourself for the process you want to run 12:27 < rud0lf> i'm not asking _how_ to not waste your time (no sarcasm), i just ask _if_ it's possible to completely disable laptop's built-in microphone 12:27 < rud0lf> to the point it doesn't show as input device 12:27 < rud0lf> so i know whether to google it or give up 12:27 < Fiacha> MrElendig: ok, i will look into it, thanks 12:40 < stennowork> good day, i have a problem configuring ntpd on centos 6.9. Here is my /etc/ntp.conf https://hastebin.com/ajoqowojog.php , my firewall (and nmap) tells me the that port 123 udp is open, but when i try to reach the server from another computer in the same network, i get 'no server suitable for configuration found' 12:40 < stennowork> i try to reach the server via: ntpdate -q 12:41 < stennowork> i restarted the ntp daemon already 12:43 < stennowork> hmm i guess the problem is with the 'restrict default' part 12:48 < stennowork> no, even setting: 12:49 < stennowork> restrict return 'no server suitable for configuration found' 12:52 < stennowork> ah hm. 12:52 < stennowork> my iptables firewall blocked it. 12:53 < stennowork> i hoped the rule: ACCEPT udp -- anywhere anywhere udp dpt:ntp 12:53 < stennowork> would allow udb on port 123 but apparently it didn't 12:53 < stennowork> time to figure out how to actually get what it want . 12:58 < stennowork> ah lol i will just disable iptables, problems solved. 13:00 < MrElendig> fix your rules instead 13:01 < stennowork> ok, what is the rule to allow any incoming and outcoming traffic on port 123 UDP? 13:01 < stennowork> its not: 13:01 < stennowork> iptables -A INPUT -p udp --dport 123 -j ACCEPT 13:01 < MrElendig> running plain iptables or some fancy wrapper like ufw or whatnot? 13:02 < MrElendig> and remember, order matters 13:02 < stennowork> unfortunately this is ShitOS, er CentOS 6, so no firewall-cmd 13:02 < stennowork> just plain iptables 13:03 < Armand> stennowork: CSF 13:04 < MrElendig> imo that is better :p 13:04 < stennowork> wait what 13:04 < stennowork> now my iptables -S doesn't list the just created rule 13:04 < MrElendig> though I would use nftables probably 13:04 < jhodrien> stennowork: It's not shit, it's stable, and it's your choice to run an old version. 13:04 < stennowork> jhodrien, its the customers choice, not my choice 13:04 < Armand> Cent6 = very stable 13:05 < stennowork> do i have to do any magic to actually apply my rule to iptables permanently 13:05 < stennowork> apparently i do 13:05 < jhodrien> You sure you're not dropping it with an earlier rule? 13:05 < jhodrien> service iptables save 13:06 < stennowork> ah duh 13:06 < jhodrien> And if you're hurling abuse at it, remember that you have the same disconnect between live and saved with firewalld. 13:07 < stennowork> here is the iptables -S output: 13:07 < stennowork> https://hastebin.com/qexomukemi.diff 13:07 < jhodrien> Right. 13:08 < jhodrien> So that line 10 you've got there. 13:08 < stennowork> i see a single reject statement that seems to concern icmp 13:08 < jhodrien> No, that's *how* it's rejecting. 13:08 < stennowork> i see 13:08 < stennowork> so that just plain rejects all the stuff that comes after it? 13:08 < jhodrien> If you append accepts after rejects, it's not going to do anything. 13:08 < jhodrien> Yes. 13:09 < jhodrien> You hit the first match, then stop. 13:09 < stennowork> ok time to figure how to remove that rule 13:09 < stennowork> gotcha 13:09 < jhodrien> Don't remove it. 13:09 < jhodrien> Move it. 13:09 < stennowork> s/remove/move/ :P 13:09 < jhodrien> Rule 11 should also go. 13:09 < jhodrien> As it's a dupe of 8 13:09 < jhodrien> (line numbers, you know what I mean) 13:09 < stennowork> yeah 13:09 < stennowork> right 13:10 < stennowork> ok lets see... 13:10 < stennowork> thanks for the help so far 13:10 < jhodrien> iptables -vL makes all this much more obvious, as you can see where things are getting matched. 13:12 < bipul> Hi, 13:14 < stennowork> jhodrien, perfect, i moved the rule and now i can talk to the server as expected. thanks for the calm help 13:14 < jhodrien> No problem. 13:14 < jhodrien> And don't be so mean to CentOS 6, it's done sterling work. 13:14 < jhodrien> You wouldn't choose to install it today, but then it is 7 years old. 13:15 < stennowork> yeah i guess 13:15 < jhodrien> You'd be saying much worse things about most other distros if they were 7 years old. 13:15 < stennowork> is there a wrapper around iptables thats available for centos 6, that you would recommend? 13:15 < jhodrien> Personally I manage iptables directly, via puppet. 13:15 < stennowork> i have to admit that this bare metal editing is a bit complex for me 13:26 < Armand> stennowork: CSF 13:30 < Truk_> hello i have a permission denied (public key) error when i try to ssh from my bastion to my EC2 instance inside a private subnet 13:30 < DildoSwaggins> damn, I was wondering why I was banned from the chan... just checked my logs... seems like my cat got on the keyboard and spammed the shit out of the chan 13:30 < Truk_> i use the same key for the two instances 13:30 < Truk_> ssh ec2-user@private_ip 13:31 < Truk_> i have : debug3: no such identity: /home/ec2-user/.ssh/id_rsa: No such file or directory 13:31 < za1b1tsu> is anyone using dropbox cli? 13:31 < DildoSwaggins> Truk_, have you modified your ssh_server config to allow the user to SSH? 13:31 < Truk_> yes 13:31 < bipul> So you're public key is not there 13:31 < DildoSwaggins> Truk_, you need to create the keys 13:32 < Truk_> i allowed the bastion as an inbound in the security group of the private EC2 on port 22 13:32 < Truk_> i created a key pair with aws 13:32 < DildoSwaggins> Truk_, you need to use `ssh-keygen` 13:32 < bipul> 1st Make sure you're able to ping. 2nd You have ssh client installed on your system. 3rd You have your .pub key on the remote server 13:32 < Truk_> i use it to connect to the bastion, and i configured the same keypair for connection to the private EC2 13:33 < Truk_> the .pub key must be on the bastion ? 13:33 < DildoSwaggins> yes 13:33 < DildoSwaggins> but also 13:33 < DildoSwaggins> what's debug3? 13:34 < mercxry> DildoSwaggins: ssh -vvv 13:37 < hypercore> why is /var/www/.well-known forbidden (nginx)? 13:37 < Truk_> what is the command for checking connectivity ? is it ns ? 13:38 < DildoSwaggins> netstat? 13:38 < bipul> ping, telnet 13:38 < DildoSwaggins> hypercore, check the permissions of the file /var/www/.well-known 13:38 < DildoSwaggins> also the owner 13:38 < hypercore> DildoSwaggins: what should they be? 13:38 < bipul> ls -l 13:39 < DildoSwaggins> ideally the same group as what user your nginx server runs as 13:39 < hypercore> DildoSwaggins: did that, but still same error 13:40 < Alexander-47u> hi all 13:40 < DildoSwaggins> ls -al /var/www/.well-known 13:40 < Alexander-47u> having a tough day here 13:40 < Alexander-47u> formatted a HDD, made windows part of dual boot corrupt, fixed that 13:40 < Alexander-47u> that destroyed grub, well i knew that was goign to happen no big deal 13:40 < Alexander-47u> used boot-repair to repair grub 13:41 < Alexander-47u> now lots of drivers in my computer dont load 13:41 < Alexander-47u> also network adapter 13:41 < Truk_> it seems that the connection is possible, with nc 13:41 < Alexander-47u> LAN 13:41 < Truk_> from the bastion 13:41 < hypercore> DildoSwaggins: should it be in /var/www or /var/www/html 13:41 < hypercore> (i have it in both at any rate) 13:42 < DildoSwaggins> depends if your distro: debian based systems use /var/www/html while fedora systems use /var/www 13:42 < DildoSwaggins> check your nginx logs to find out what exactly is going wrong 13:43 < Alexander-47u> anyone know why a bunch of my drivers dont work after using boot-repair 13:43 < DildoSwaggins> tail /var/log/nginx/error.log 13:44 < Alexander-47u> is it even possible to break drivers by altering grub? 13:45 < DildoSwaggins> Alexander-47u, perhaps it's booting up in the wrong setting, like debug mode or something? 13:45 < hypercore> DildoSwaggins: just says it's forbidden 13:45 < Alexander-47u> DildoSwaggins, ye but my wifi works tho 13:45 < DildoSwaggins> hypercore: what does? the nginx log file? 13:45 < hypercore> DildoSwaggins: does var/ww/html need my user's permissions too? 13:45 < hypercore> DildoSwaggins: yeah 13:45 < Alexander-47u> thats how im chattin right now 13:45 < DildoSwaggins> hypercore, use sudo 13:45 < hypercore> DildoSwaggins: for what? 13:46 < DildoSwaggins> to access the nginx logs 13:46 < hypercore> DildoSwaggins: i did, it said forbidden for .well-known 13:46 < DildoSwaggins> Alexander-47u, to answer your question specifically, as far as I know, no. altering grub should not break your drivers. 13:47 < DildoSwaggins> but by altering grub, you can affect the way your system boots up, and the parameters/config settings it uses 13:47 < hypercore> DildoSwaggins: also, my index file in /var/www/html has root:root permissions 13:48 < EriC^> Alexander-47u: if you have the boot-repair log that might shed some light 13:48 < hypercore> does it need write permissions or something? 13:48 < prussian> No 13:48 < prussian> Is it world readable? 13:48 < Alexander-47u> i have a copy of my old working grub 13:48 < Alexander-47u> so im mounting that right now 13:48 < Alexander-47u> to see the differences 13:49 < DildoSwaggins> hypercore, that doesn't sound right 13:49 < DildoSwaggins> it's not that it has "root permissions," but rather that it's ONLY accessible by root 13:50 < DildoSwaggins> you need to edit that by using "chown" 13:50 < DildoSwaggins> sudo chown -hR /var/www/html nginx:nginx 13:50 < DildoSwaggins> oops 13:50 < DildoSwaggins> sudo chown -hR nginx:nginx /var/www/html 13:50 < prussian> Don't do that. It's probably a selinux policy. Check your labels 13:51 < DildoSwaggins> well... /var/www/html should not be owned by root 13:51 < prussian> Doesn't really matter 13:51 < DildoSwaggins> it should atleast be accesibly by your httpd server 13:51 < prussian> As long as it's group or other readable 13:51 < hypercore> this is completely ridiculous 13:52 < prussian> hypercore: restorecon -R /var/www/html 13:52 < hypercore> prussian: do i need to install restorecon? 13:52 < prussian> Will likely restore the appropriate context 13:52 < prussian> IDK. Do you not have selinux? 13:52 < hypercore> debian 13:53 < prussian> O... K 13:53 < revel> (SELinux isn't a distro) 13:54 < hypercore> oh then i'm not sure 13:55 < hypercore> it's just the standard DO debian stetch image 13:55 < revel> Easy way to check that doesn't rely on any special SELinux tools: `cat /sys/fs/selinux/enforce` 13:56 < Alexander-47u> anyone know a good way to extract a file from a full system image? i cant mount it 13:56 < Alexander-47u> and using peazip is painfully slow 13:56 < hypercore> revel: no such directory 13:56 < revel> No SELinux then. 13:57 < bipul> Selinux is usesful? 13:59 < DildoSwaggins> selinx is akin to hemorrhoids... a royal pain in the ass 14:01 < stennowork> Armand, will look into that, thanks 14:02 < hypercore> man i don't get why this is so difficult 14:02 < hypercore> i hate technology sometimes 14:02 < hypercore> im going to become a hermit 14:03 < Armand> Nerd = hermit. 14:03 < Armand> Mission accomplished. 14:03 < cha0tic> https://www.gentoo.org/news/2018/06/28/Github-gentoo-org-hacked.html 14:06 < Armand> cha0tic: Old news is old. 14:06 < DildoSwaggins> lol 14:07 < DildoSwaggins> it's like 12hours old 14:07 * cha0tic lazily scrolled up. didn't see a post. lazily pasted. 14:08 < DildoSwaggins> nah cha0tic, I mean to say it's not that old. I'm not sure if 12 hours counts as "old news" 14:10 < cha0tic> yeah. new to me. just woke up 14:11 < stennowork> i assume this is connected to the wordlist attack that also included freenode 14:11 < stennowork> apparently someone got a lot of password data from some huge network 14:31 < tdn> What free forum software would you recommend for a new forum site? Must have a good security record so wordpress is out. 14:32 < DildoSwaggins> well wordpress is not a forum software, it's a blogging software 14:32 < Pentode> oopsies 14:33 < searedvandal> fluxbb? 14:33 < DildoSwaggins> one of the oldest and most common opensource forum softwares in phpBB 14:33 < trobotham> also, webapps are only as secure as you make them (ie: running updates, etc) 14:34 < DildoSwaggins> is* phpBB 14:39 < MrElendig> also wordpress is not something you should use 14:39 < MrElendig> 99% of all wordpress blogs could simply be a static site 14:40 < Sitri> I had a company try and sell us a wordpress website to replace our static site 14:40 < Sitri> (Which we update at most once a year) 14:40 < Sitri> "Yeah... pass." 14:41 < MrElendig> anyway, speaking of forums, flux is one of the less crappy ones 14:41 < MrElendig> wouldn't hurt if it was a bit more active though 14:49 < tdn> DildoSwaggins, yes, but I have asked some other places where people suggested various plugins for WP 14:49 < tdn> DildoSwaggins, phpBB has a horrible security record 14:49 < tdn> MrElendig, will look into flux. thanks 14:50 < MrElendig> one good thing about fluxbb is that the developer generally listens to input 14:50 < MrElendig> whenever he is around anyway :p 14:51 < tdn> MrElendig, cool 14:51 < tdn> MrElendig, do you use it? 14:51 < searedvandal> arch uses fluxbb 14:51 < tdn> MrElendig, do you have a link to a real world flux forum? 14:51 < tdn> searedvandal, oh 14:51 < MrElendig> also fluxs own forum 14:51 < cluelessperson> anyone here workedon the A9 processor? 14:52 < tdn> Will give flux a try :D 14:52 < MrElendig> https://fluxbb.org/forums/index.php 14:54 < linoyos> hello 15:06 < BenderRodriguez> hi linux 15:06 < BenderRodriguez> I need guidance 15:06 < BenderRodriguez> is there an all-in-one openstack distro 15:09 < DildoSwaggins> BenderRodriguez, Debian 15:10 < DildoSwaggins> https://www.openstack.org/marketplace/distros/distribution/debian/debian 15:10 < no_usr> hello all, i want to ask a question please, i am using linux inside windows as WSL in windows 10 you know... and another windows xp inside virtualbox... i can ping WinXP from Win10 host but can't ping it from linux. nmap sayz can't open device eth0... so how to do this??? 15:10 < no_usr> sorry i failed to find specific channel for this q... 15:11 < BenderRodriguez> DildoSwaggins: what I mean is, a distro that comes pre-packaged with openstack 15:11 < DildoSwaggins> did you try running nmap with 'sudo' 15:11 < BenderRodriguez> i simply spin it up and boom, I have openstack 15:11 < DildoSwaggins> BenderRodriguez, https://www.openstack.org/marketplace/distros/. take your pick :p 15:11 < no_usr> DildoSwaggins, yeah i ran it with sudo, i tried ping, hping3 but no luck... 15:16 < Elodin> how can I effectively rollback from a readonly snapper snapshot? Say i booted into some readolny snapshot, and i did a snapper rollback. How can i make grub boot on the right head now 15:19 < Alexander-47u> does anyone know how to extract grub.cfg from a dd image? 15:19 < matsaman> Alexander-47u: you can mount FSes from dd copies, with a little effort 15:19 < matsaman> next time don't use dd for whatever it is you're doing 15:19 < Alexander-47u> i have but i cant find it 15:20 < matsaman> it's in /boot/(grub/) usually 15:20 < Alexander-47u> yes but really, cant find it 15:20 < matsaman> if you had a separate /boot/ partition and didn't also dd that, it won't be present 15:20 < MrElendig> sounds like a xyproblem 15:20 * matsaman rolls eyes 15:20 < matsaman> you sound like an xyproblem 15:20 < Alexander-47u> oshit 15:20 < MrElendig> also, the image probably contain multiple partitions, sure you are looking in the correct one? 15:20 < Alexander-47u> i think i have the wrong partition lol 15:22 < matsaman> "also" hurrr 15:22 < MrElendig> an esp inside the image? 15:22 < Alexander-47u> hmm 15:22 < Alexander-47u> i have encrypted home partitions 15:22 < Alexander-47u> will this also encrypt boot? 15:22 < Alexander-47u> used ubuntu 15:22 < matsaman> well it wouldn't be in /home/ 15:22 < matsaman> that depends on how you encrypted 15:22 < Alexander-47u> yes but somehow 15:23 < Alexander-47u> the entire disk is 'unnknown' 15:23 < Alexander-47u> instead of ext4 15:23 < matsaman> according to what? 15:23 < matsaman> where'd you get this dd copy? 15:23 < Alexander-47u> according to fdisk -lu 15:23 < Alexander-47u> i used a tool named 'disks' 15:24 < Alexander-47u> its on most linux distributions 15:25 < matsaman> most Debian distributions, maybe 15:25 < matsaman> and where is the hardware you got this dd copy from now? 15:25 < matsaman> where's the original 15:25 < Alexander-47u> yes sorry debian distributions 15:25 < Alexander-47u> its on my NAS 15:25 < matsaman> why don't you just get the file from there 15:26 < matsaman> and if you want a copy of it, make another copy not using dd because that's a pretty clunky approach that leads to problems like this one 15:26 < Alexander-47u> no its a copy of my own system 15:26 < Alexander-47u> and i have a DD image on my NAS 15:26 < Alexander-47u> trying to find grub.cfg 15:26 < matsaman> okay from your own system then 15:26 < Alexander-47u> yes 15:26 < MrElendig> Alexander-47u: why are you trying to find it? 15:26 < matsaman> it's possible you aren't even using grub or grub.cfg, too 15:29 < Alexander-47u> oke 15:29 < Alexander-47u> i found it 15:29 < Alexander-47u> i was just mounting the wrong systems 15:29 < Alexander-47u> wild assumptions of it being on the smaller partitions 15:30 < Alexander-47u> but LUKS offers the choice at installation to separate them or join them, i joined them and forgot 15:30 < matsaman> good job 15:32 < Alexander-47u> :D 15:32 < Alexander-47u> brb reboot 15:33 < backnforth> Hi, can someone tell me where does simplescan save it's images? 15:35 < MrElendig> backnforth: did you check the manual? 15:35 < backnforth> there's a manual? 15:35 < MrElendig> did you check the stadard ~/Pictures ? 15:35 < backnforth> MrElendig, yes 15:35 < MrElendig> you can strace it then 15:36 < MrElendig> or run with --debug 15:36 < backnforth> Found it 15:36 < backnforth> it's in documents 15:36 < MrElendig> seems like you can "save as" too 15:37 < backnforth> yep, that's what I did to find the location 15:37 < MrElendig> from the .po file anyway, as usual cannonical software have no useful online docs 15:37 < MrElendig> :/ 15:37 < MrElendig> f1 should work fine when you have it installed though 15:46 < hypercore> guys any idea why drone is giving gitea a webhook url of 127.0.0.1:8000 instead my website.com? 15:47 < matsaman> why would it give website.com 15:49 < hypercore> matsaman: why wouldn't it? i set DRONE_HOST=website.com (obviously not my website), but it's still giving 127.0.0.1:8000 15:49 < matsaman> you would probably have to hack up your hosts file to make that manifest 15:49 < matsaman> but it would just be pretend, of course 15:53 < throstur> where can I find a list of distributions that work well in a VM? 15:54 < devC0de> http://www.vagrantbox.es/ 15:54 < matsaman> throstur: "well"? 15:54 < MrElendig> any that doesn't have a 20000 year old kernel do 15:55 < matsaman> there are only a tiny handful of distributions that aren't derivations of other ones 15:55 < throstur> well because I tried antergos and it didn't even have proper mouse support 15:55 < matsaman> and they should all work about the same, as the kernel, Linux, does most of the work 15:55 < searedvandal> pretty much anything can run well in a VM provided it's set up correctly 15:55 < matsaman> throstur: what's proper mouse support? 15:56 < MrElendig> you mean host clipboard intigration? 15:56 < MrElendig> host/guest* 15:56 < MrElendig> that works if you use a sane vm and a distro that isn't 2000 years old 15:56 < throstur> matsaman: when I click on something while the guest has focus, the click is definitely not being registered in the right place 15:56 < revel> Did you forget to install the guest additions? .-. 15:56 < throstur> for example, trying to close the Konsole window by pressing the 'x' button doesn't work 15:56 < searedvandal> throstur, I've had multiple Antergos VMs running and they have all had "normal" mouse behaviour. so I'm guessing its a problem with how you set up your VM 15:56 < MrElendig> basic mouse should just work 15:56 < matsaman> throstur: hrmmm, maybe your monitor has some kind of far out dpi 15:57 < throstur> guest additions... on the guest? Whoops. yes, I forgot to do that. 15:57 < matsaman> or your mouse could be a little mad 15:57 < MrElendig> resizing the window might break some cruddy vm's though *coughvmwarecough* 15:57 < searedvandal> yeah, install the guest additions and see what happens. 15:57 < revel> The guest additions are pretty useful, yeah. 15:57 < throstur> no.. it was already installed 15:58 < throstur> I'm just using a regular trackpoint mouse 15:58 < matsaman> I'd make sure your host and guest resolutions are being grokked similarly, first 15:58 < throstur> it is rendered where it is moved, just the mouse clicks don't seem to have the desired effect, e.g. clicking inside the open Konsole window loses it's focus 15:58 < matsaman> resolutions/pixel densities 15:58 < matsaman> or you could just have a stuck key or touchpad covered in grease 15:59 < throstur> brand new laptop and mouse works in host, so definitely not 16:01 < matsaman> well, probably not anyway 16:01 < matsaman> or less likely not 16:02 < throstur> no. it's verifiable. The mouse works outside of the guest, therefore the mouse being physically obstructed has been disproved by contradiction... 16:02 < matsaman> _kinda_ 16:02 < searedvandal> to your original question, pick a random distro, they can all work well in a VM. 16:03 < matsaman> lot of VMs have all sorts of modified modifier keys in play 16:03 < searedvandal> check your VMs mouse settings, maybe there are some input driver you miss in the guest. or it's just kde being kde 16:10 < BenderRodriguez> GMT and UTC are interchangeable terms right 16:10 < BenderRodriguez> such that UTC -5 === GMT -5 16:10 < BenderRodriguez> right 16:10 < c50a326> hey is the Comptia Linux+ certification the standard entry level certification that should be recommended to people with no Linux exposure thus far? 16:11 < Linkandzelda> does anyone know how to ssh into an schroot session from another box? 16:11 < post-factum> BenderRodriguez: wrong 16:11 < BenderRodriguez> post-factum: wait what how 16:11 < BenderRodriguez> explain. 16:11 < post-factum> BenderRodriguez: gmt is a time zone, utc is a time standard 16:12 < Pentode> theres no time difference though 16:12 < post-factum> https://en.wikipedia.org/wiki/Coordinated_Universal_Time ctrl+f gmt 16:16 < tcpdump> If I have two text files, each with a single column numbers, is there a command that will compare them? I want to find and numbers that don't exist in both files. 16:16 < tcpdump> s/column/column of/ 16:17 < post-factum> tcpdump: man comm 16:17 < post-factum> tcpdump: or man diff 16:17 < matsaman> diff -s <(cat file1 | sort | uniq) <(cat file2 | sort | uniq) 16:19 < lytedev> morning, all 16:19 < tcpdump> hey lytedev 16:59 < BenderRodriguez> oh no 16:59 < BenderRodriguez> they got gentoo 17:00 < revel> Not really. 17:01 < post-factum> it was just a mirror, although, i still think it is well-deserved ;) 17:01 < revel> Well-deserved? 17:03 < post-factum> precisely 17:03 < revel> What for? 17:04 < searedvandal> why do you think it was well-deserved? 17:13 < triceratux> https://gentoo.org/news/2018/06/28/Github-gentoo-org-hacked.html 17:14 < Ussat> any logcheck experts here ? have a sticky question 17:15 < sarman> Whats teh question? 17:15 < Ussat> let me pastbin it 17:16 < Ussat> sarman, https://pastebin.com/CuNPyCfg there ya go, I did not want to spam channel 17:17 < Ussat> Its reading the files it needs to check, but having issues with the lock file etc as shown 17:18 < Ussat> sarman, thoughts ? 17:19 < sarman> @Ussat, looking now 17:20 < Ussat> Thank you 17:22 < sarman> I will have to bow out on this. when you said logcheck, I though you ment analizing the logs. I am not familure with "logcheck" myself. I would suggest tailing the /tmp/logcheck.* file and look for clues there 17:23 < Ussat> thank for looking 17:30 < noway96> how do I find the block size of a SAS hard drive attached to my ubuntu machine via a SAS controller 17:31 < noway96> ok so I used blockdev --getsize and it's reading the drive as having 0 block size 17:49 < pankaj_> I heard a term about kali linux . It says that it is an enterprise ready distribution. I cannot fully undeestand what does it means. 18:09 < takeme> can i know what it means? sed -i 's/gitrev.*/gitbranch: master/' ~/.pybombs/recipes/gr-recipes/gr-iio.lwr pybombs install gr-iio 18:09 < nothingcomplex> what are some trends in how traditional systems engineering executives are extracting value to the organization from their department? 18:09 < revel> takeme: Is that supposed to all be on the same line? 18:09 < takeme> https://www.reddit.com/r/RTLSDR/comments/6tnjba/getting_started_with_adalmpluto_and_gnu_radio/ 18:09 < noway96> blockdev --getsz returns 0 sectors in the disk. What? 18:09 < takeme> revel: link plz 18:09 < revel> Link for what? 18:09 < takeme> revel: can i know what it means? sed -i 's/gitrev.*/gitbranch: master/' ~/.pybombs/recipes/gr-recipes/gr-iio.lwr pybombs install gr-iio 18:09 < revel> Anyways, guess I don't understand some fine point of sed then, if you can do that. 18:09 < searedvandal> kali linux is a distro for security auditing and penetration testing. don't think I've seen them describing themselves as enterprise ready. though enterprise ready usually don't mean anything. anyone can say their product is enterprise ready 18:09 < takeme> what that means? revel ? 18:09 < revel> Well, `sed -i` basically edits a file according to some expression, ''s/gitrev.*/gitbranch: master/'' is the expression (replace some text with other text), the file after that should be what it does all of that on. 18:09 < revel> Not sure what passing an executable file with arguments is supposed to do, if anything. 18:09 < revel> Or, err. Whatever it's doing. 18:09 < revel> Or maybe it's supposed to be doing that for all of pybombs, install and gr-iio... "pybombs install gr-iio" just seems a lot like a command. 18:12 < WhiteDevil> hey do you guys know of hufe python applications ? 18:12 < WhiteDevil> i wanna check what python can be used to do 18:13 < rumpel> it can save the day 18:15 < pankaj_> I had no confusion with term 'mirror' in linux until i came to a post where ome said that it is a file copied on different server. Another said that it is website copied to different servers. What is accurate defination. 18:15 < Sitri> pankaj_: a website is a bunch of files. 18:15 < Sitri> So a mirrored website is just a bunch of mirrored files. 18:16 < pankaj_> Sitri: what is the correct defination i am asking. 18:17 < Sitri> ... files 18:18 < ice9> is there away to see if data is still being written to a disk? 18:18 < pankaj_> Sitri: i mean that are the files (which are downloadable) or web based pages. 18:21 < pankaj_> Sitri: hello 18:24 < Sitri> pankaj_: I've answered your question twice. I really don't know how you're not understanding it. 18:27 < pankaj_> Sitri:. So just last time i want to confirm please. It means that all the webpages and downloadable contents like iso images are mirrored to a server nearest to users? Right? 18:28 < Sitri> Not necessarily 18:29 < Sitri> A mirror is simply a maintained copy of something. It doesn't mean that the mirror is guaranteed to be closer to the user, the user could end up using a mirror that's further away. 18:30 < Sitri> A CDN is a type of mirror, where they use various tricks to make it so the user will connect to a server close to them. 18:30 < ldlework> I forgot where I was and thought Sitri was making metaphysical claims. 18:30 < pankaj_> Sitri: Well i noticed the first sentence and got it. 18:30 < Ussat> Sooo....getting headache here or hurting the wall hitting it with my head...have a log check issue on Ubuntu 18.04 LTS, sure its a user issue :) but I have no idea what I crewed up, here is a pastbin with errors at bottom: https://pastebin.com/dRLqUR45 18:30 < Ussat> command run was" su -s /bin/bash -c "/usr/sbin/logcheck -d" logcheck 18:31 < Celmor> how would I schedule a one-off task to run in X amount of time? I know about cron to run repeated tasks and atd to run something at a certain time, only solution I know off atm is to run and background a shell command prefixed by `sleep` 18:32 < pankaj_> Sitri: if i am right then it is dns server that keeps mapping if addresses so it is the one reslonsible for redirecting to ip address nearest to the server (which belongs to that org. Or anything like that) 18:32 < Sitri> Celmor: the at command 18:32 < Celmor> does it support "run in X hours"? 18:32 < Sitri> That's its entire purpose 18:33 < Celmor> was only familiar with the `echo "cmd" | at 12:30` format 18:33 < Celmor> not easy to search for documentation about a command that's called only "at" 18:33 < azarus> man at 18:33 < Sitri> Or google "man at" 18:33 < Celmor> for me it's atd 18:34 < Sitri> atd is the daemon, at is the command a user calls 18:34 < Celmor> thanks 18:40 < triceratux> woo woo mint 19.0 is for real https://distrowatch.com/?newsid=10244 18:40 < tecywiz121> Hey! I need some help interpreting an strace. Any suggestions for where to go asking? 18:41 < revel> triceratux: It's barely halfway through '18 though. 18:41 < Celmor> ah, `at now + $X hours` 18:42 < triceratux> revel: at least he didnt go from 42.3 to 15 or something 18:43 < tnewman> need help getting thumbnails to work in ejabberd, any takers? 18:43 < tnewman> no one is responding in #ejabberd 18:55 < pankaj> Is there a way to get information at the same time while 'dd' is doing the copying of files? May be like statical report or progress or anything like that. 18:56 < revel> Yes, there is. 18:56 < tecywiz121> pankaj: Yes, you can send it a USR1 signal to get it to output stats 18:56 < revel> That's, uhh, status=progress, I think. 18:56 < Sitri> pankaj_: dd status=progress 18:57 < tecywiz121> Yeah, if you haven't started it already that works well. If it's already running, you can "kill -USR1 $pid", where $pid is the PID of dd. 18:57 < revel> pkill would also work. 18:58 < Pentode> could also use pv, if those aren't enough options for you... dd if=/foo | pv | dd of=/baz 19:00 < koala_man> you can add more dd with dd if=/foo | dd | dd | pv | dd | dd of=/baz 19:01 < Pentode> you can even pipe it through your kitchen sink, if you dig that 19:01 < koala_man> or just use pv /foo > /baz 19:01 < rymate1234> can you dd through ssh to transfer an entire disk over my LAN? 19:01 < koala_man> yes 19:01 < Pentode> sure why not 19:01 < revel> You can pipe through cat as well. 19:02 < revel> So, `dd if=foo | cat | dd | pv | dd | cat | dd of=bar` 19:02 < Pentode> lol 19:04 < lordvadr> less is also smart enough to know when output isn't a tty, so you can do ``dd if=foo | cat | more | less | dd | pv | dd | cat | more | less | dd of=bar` 19:04 < rymate1234> so obviously you want to dd if=/dev/sda | cat | tar cz | pv | ssh | tar xf | cat | dd of=/dev/sdb 19:06 < lordvadr> let's not forget `awk 1` in there too 19:07 < koala_man> cmp <(printf foo) <(printf foo | awk 1) says that's not a great idea 19:08 < lordvadr> I wondered if awk would horse up newlines and such 19:08 < lordvadr> But, you can scp /dev/stdin /dev/stdout 19:08 < fscale> Hi folks, is there a linux distro that supports 125% scaling? Most distros only seem to have 2x. I'm running 1080p on a 15 inch laptop on which 125% scaling looks good. 19:09 < fscale> *125% display scaling. 19:09 < MrElendig> up to the wm/de, not the distro 19:09 < MrElendig> and none of them do non-whole-int scaling well 19:10 < fscale> MrElendig: Which is the best one out of them though? I would download the distro that has the DE with 125% scaling support out of box. 19:10 < prussian> fscale: gnome 3 with wayland 19:10 < prussian> it's the only one with fractional scaling I know of and I beleive it has per-monitor scaling now as well 19:11 < rymate1234> plasma shell has fractional scaling 19:11 < fscale> prussian: Ok, off to download that & try it. rymate1234: Will download that too. 19:11 < fscale> I'll download kubuntu for plasma 19:12 < MrElendig> the problem is that almost none of the software you will be running inside the DE supports it 19:13 < fscale> MrElendig: So a DE doesn't scale all programs automatically? I would be OK with some blurriness but can't stand a small UI. 19:14 < MrElendig> how could it? 19:14 < fscale> 2x would be too big as I already mentioned. 19:14 < rymate1234> it seemed to work fine with kde for me 19:14 < MrElendig> xrandr can scale anything, but with gotchas 19:16 < fscale> Regardless of the problems, I want to see how stuff looks. What distro comes with wayland gnome out of box? 19:17 < prussian> fedora 19:18 < triceratux> ubuntu 17.10 19:18 < fscale> Fedora-Workstation-netinst-x86_64-28-1.1.iso <-- I'm downloading this one right now. Hope it has that. 19:18 < rymate1234> would the live dvd not be better? 19:18 < lordvadr> Fedora represent! I gave up on gnome 3 a long time ago though. 19:19 < rymate1234> arch > fedora 19:19 < fscale> rymate1234: I will use the ISO with a bootable USB. Using Easy2boot. Makes it very easy to boot any ISO. 19:19 < fscale> A live session I will test. 19:19 < ice9> is there away to see if data is still being written to a disk after running sync? 19:20 < lordvadr> ice9: IIRC, once sync returns, everything has beenf flushed 19:20 < prussian> could just umount just to be safe 19:20 < rymate1234> fscale, yeah what I'm saying is that the iso you're downloading isn't the live dvd 19:20 < rymate1234> the filename for that is Fedora-Workstation-Live-x86_64-28-1.1.iso 19:21 < ice9> lordvadr, nope because the flash drive led still flashes for a while even after sync returns, once I unplugged it after sync and data was corrupted 19:21 < rymate1234> I'm _pretty sure_ the netinst is only the installer 19:21 < rymate1234> which is why the size is smaller 19:21 < fscale> Where is the full ISO then? 19:21 < rymate1234> so if you want to just try scaling it might not be the best 19:22 < rymate1234> fscale, ....right there on the website? https://getfedora.org/en/workstation/download/ 19:22 < lordvadr> ice9: Well, it's out of the kernel's buffers. Storage could still be doing something with it. It's an issue with raid controllers too. 19:22 < rymate1234> it's the big download button at the top 19:22 < rymate1234> at least on my PC 19:23 < bls> yeah, cheap controllers will often lie about their write status to game benchmarks 19:23 < rymate1234> oh god 19:23 < rymate1234> hi matrix 19:24 < fscale> rymate1234: Ok, me downloading this now --> Fedora-Workstation-Live-x86_64-28-1.1.iso 19:24 < fscale> 1.66 GB 19:24 < rymate1234> ye 19:24 < rymate1234> that contains the full desktop experience 19:35 < nik> havent tried this matrix thing yet 19:36 < nik> does it keep the connection online all the time like znc? 19:38 < MrElendig> if you use a reliable gateway yes 19:38 < MrElendig> (most are not) 19:48 < nothos> Can someone explain why btrfs seems to have fallen out of favour? 19:48 < nothos> Reading up on new RHEL release and how it'll likely drop btrfs support, and realised that after a bit of fan-fare no one seems to actually use it 19:48 < nothos> No idea why :D 19:50 < prussian> nothos: I imagine mostly because of long standing fud about it 19:54 < bls> every discussion of btrfs seems to break down into two camps, people that've been bitten by bugs it has and refuse to use it again, and people that love it 19:56 < jim> and then there's me, I've heard the horror stories, and lvm's a better alternative 19:56 < MrElendig> lvm and btrfs are quite different though 20:03 < rymate1234> Brtfs can be a file system on top of lvm 20:08 < mgolisch> so can any other fs.. 20:11 < sensible> Is there a way to increase space for swap in centos? 20:12 < bls> sure. adjust partition/FS sizes, add a swap file 20:13 < sensible> bls: can you guide me through it? 20:14 < sensible> I can share the output of commands step by step 20:15 < bls> sensible: sorry, don't have the attention bandwidth right now, but the swap file approach is the safer/easier option: https://www.digitalocean.com/community/tutorials/how-to-add-swap-on-centos-7 20:15 < sensible> bls: also can we increase the root space on centos? 20:16 < sensible> bls: I have already checked this tutorial it does not work for centos 20:18 < Pusteblume_> sensible, what exactly does not work when you try to follow that tutorial? 20:18 < bls> is this for a VPS? they may have disabled your ability to manipulate swap 20:24 < Psi-Jack> sensible: that tutorial is good. The real question what didn't work and what errors do you get and what did you actually do? 20:24 < Psi-Jack> Be sensible and less vague. 20:26 < Pusteblume_> Psi-Jack, why do you overwrite me? 20:27 < Psi-Jack> Pusteblume_: Pardon? 20:28 < Psi-Jack> Alrighty. LOL 20:36 < NewbProgrammer10> Hello. 20:46 < seveneleven> what is easier to transfer a file from a linux to a windows host: samba or FTP? 20:47 < seveneleven> (and less likely to fail) 20:47 < bls> FTP shouldn't be considered an option, so samba 20:47 < seveneleven> thanks, then I go for samba \o/ 20:48 < bls> help it settle into great beyond... https://mywiki.wooledge.org/FtpMustDie 20:48 < section1> http works fine too 20:49 < Elladan> nothos, btrfs has a history of having a lot of serious problems which go years before being fixed, if ever. 20:50 < bls> does either OS have a WebDAV-server-out-of-the-box solution 20:51 < Elladan> I don't think it's difficult to use scp with windows hosts these days. 20:53 < Elladan> Samba is usually easier, but it can be a pain if there are domain controllers involved. 20:54 < dgurney> the only reason I still have an ftp server running is that it works on basically anything. I mean, even Windows 98 supports it without any extra software install necessary 20:54 < mgolisch> hm no idea not realy anything i would care about 20:58 < NewbProgrammer10> But something I would care about. 20:58 < NewbProgrammer10> dgurney: how did you set up your ftp server? 20:58 < NewbProgrammer10> I'd like to set up one for my network. 20:59 < bls> you really don't. there are much better solutions 20:59 < mgolisch> every linux box has sshd 20:59 < NewbProgrammer10> Puck off 20:59 < mgolisch> not need for that crap 20:59 < dgurney> NewbProgrammer10, just filezilla server running on my Windows Server (my file server) 20:59 < NewbProgrammer10> Hm. I'll check out this software. 20:59 < dgurney> I'm sure you can find an easy tutorial to follow elsewhere for a linux solution 21:00 < NewbProgrammer10> Yeah. 21:00 < lukey_> proftpd comes to mind 21:01 < sensible> bls: this tutorial https://www.digitalocean.com/community/tutorials/how-to-add-swap-on-centos-7 gives this error fallocate: /swapfile: fallocate failed: No space left on device 21:02 < bls> sensible: that'd imply that you're trying to create a swap file larger than the free space you've got available 21:02 < NoriusNotorius> ftp is not recommended these days due to security concerns. Why not have SSH enabled on your linux box and use something like Filezilla on your Windows machines for transfers? Or just setup samba on your linux hosts. Another alternative is to setup something like Openmediavault and put it on your network. 21:03 < sensible> bls: Filesystem Type Size Used Avail Use% Mounted on /dev/mapper/centos-root xfs 50G 44G 6.6G 87% / devtmpfs devtmpfs 63G 0 63G 0% /dev tmpfs tmpfs 63G 34M 63G 1% /dev/shm tmpfs tmpfs 63G 36M 63G 1% /run tmpfs tmpfs 63G 0 63G 0% /sys/fs/cgroup /dev/sda2 xfs 1014M 161M 854 21:03 < bls> s/is not recommended these days/has been bad practice since the 90s/ 21:03 < Elladan> I'd suggest just using ssh/scp for everything. 21:03 < Elladan> It's gotten a lot easier to use it on Windows hosts AIUI. 21:04 < Elladan> Samba is almost as insecure as FTP by the way. 21:05 < lordvadr> Screw that. Setup kerberized NFS. 21:05 < Elladan> You have to configure it with a domain controller for it to be marginally secure, and you'll be reading a whole lot of manuals if you go that route. I'm not even sure samba /can/ use transport security (it's only in the latest few SMB standards) so even then it's usually insecure. 21:07 < sensible> bls: there is much free space on my system but still same error. 21:07 < Elladan> Just to be clear, default SMB security is to use an unsalted MD4 (I think?) hash of the password and pass it plain across a TCP socket. The RPC stream after that is in plaintext. 21:07 < lukey_> sensible: but maybe not enough :) 6Gb free isn't that much especally with >64G Ram 21:08 < gronke> any idea if I could wget a google drive zip? 21:08 < mgolisch> sensible: it only shows 6gb free space, what command did you run? 21:08 < Elladan> It's possible to brute force unsalted SMB passwords in a matter of minutes, so it really shouldn't be considered significantly different than FTP. 21:09 < sensible> mgolisch: /dev/mapper/centos-home xfs 3.6T 154G 3.5T 5% /home 21:09 < bls> from a security standpoint, you're right. ftp also has issues the port/active/passive/firewall traversal issues on top of the security problems 21:10 < sensible> mgolisch: df -Th 21:10 < bls> grrr, can't decide if latency or drunk 21:10 < mgolisch> sensible: yeah but you tried to create a swapfile in / which only has 6gb free sapce 21:11 < sensible> mgolisch: then how to create it using /home? 21:12 < bls> are you sure you want to be doing this? 21:12 < pankaj> I used to thought that Virtual box guest addition should be installed by host OS for virtual machine. Also that it is not specific to each linux distribution. That is it is same. 21:12 < NoriusNotorius> Elladan: this assumes that the account your brute forcing is in a dictionary no? 21:12 < mgolisch> specify a filepath in that filesystem? 21:13 < mgolisch> but iam not sure you want to do any of that if thats not obvious to you 21:13 < jim> you have two other choices... if you're using lvm, you can make an lvm lv for swap, or you can make a new partion for swap... 21:13 < NoriusNotorius> Elladan: password rather 21:14 < dgurney> I use a basic smb setup, and also have an ftp running for old stuff (as mentioned previously) 21:14 < dgurney> I find those good enough for my use, as I'm the only person using the local network anyway 21:15 < sensible> mgolisch: actually my ram is 125GB and swap is 4GB, I am using a tool in which memory usage expands above 125GB and then it uses swap but swap is small so it gives memory shortage error. 21:17 < jim> what inspired you to get so much ram? 21:17 < mgolisch> create a swapfile then or increase the size of your swap partition if its on lvm 21:18 < lukey_> sensible: Well then simply fallocate the file somewhere else :) /home/swap... 21:18 < bls> that's what's being attempted. they're just looking for someone with the time to walk them through it step by step 21:19 < searedvandal> step 1. man 21:27 < pankaj> This was the first time I installed successfully something on any linux distro and it did not worked out. 21:30 < sensible> I tried it in /home but again the same error [root@localhost dellemc]# fallocate -l 50G /swapfile fallocate: /swapfile: fallocate failed: No space left on device 21:30 < lukey_> sensible: that's an absolute path 21:30 < clemens3_> sensible: what tool uses more than 125gb.. ? seriously, redesign it, use a database, don't put everything in ram 21:30 < bls> do you understand how file paths work on Linux? 21:31 < sensible> lukey_: [root@localhost bilal]# cd /home/ [root@localhost home]# ls dellemc username [root@localhost home]# cd dellemc/ [root@localhost dellemc]# ls anaconda3 Documents Music Templates users bin Downloads Pictures tmp Videos Desktop gatk-4.0.4.0_2.zip Public Untitled Document 1 [root@localhost dellemc]# fallocate -l 50G /swapfile fallocate: /swapfile: fallocate fail 21:31 < mgolisch> clemens3_: a database maybe? 21:31 < mgolisch> ours uses much more than that 21:31 < sensible> clemens3_: how to ,make a database ? when the tool is linux based and standalone? 21:31 < clemens3_> there is seriously an issue with the tool... 21:32 < jim> excercise due caution when you're logged in as root 21:32 < clemens3_> ours: yeah, but then you know how to add swap space 21:32 < clemens3_> sensible: you didnt say what that tool does.. 21:33 < bls> this isn't just a swap issue, this is a fundamentals of files and paths issue 21:33 < clemens3_> bls: indeed 21:33 < sensible> clemens3_: genome indexing and alignment https://github.com/alexdobin/STAR 21:33 < clemens3_> but don't put one fix on another 21:34 < jim> sensible, in postgresql, the command is createdb putDBNameHere 21:34 < bls> hehe, when scientists code 21:34 < clemens3_> bls: lol 21:34 < sensible> bls: filepaths issue ok , you mean that I should give relative path here ? 21:34 < clemens3_> well, you can buy another 123G ram..:) 21:35 < jim> sensible, are you using postgresql? 21:35 < clemens3_> jim: seems he uses ram 21:35 < clemens3_> and linux 21:35 < mgolisch> :) 21:35 < lukey_> sensible: Or better an absolute path to the swapfile in /home... https://swcarpentry.github.io/shell-novice/02-filedir/ 21:35 < bls> you can do one or the other, give an absolute path that's not to a full filesystem, or cd where you want the file and use a relative path 21:35 < sensible> jim: no I have used mysql and sql 21:35 < jim> what do you want to use? 21:36 < sensible> jim: I have postgresql just once 21:36 < sensible> *used 21:36 < mgolisch> then again if the tool does stuff in memory a database will not help.. and some program using more than 100gb of memory isnt neccessarily wrong, i our database server uses alot more ram than that 21:37 < jim> postgres is great, the pg devs are trying to copy features from oracle 21:37 < sensible> lukey_: thanks for this tutorial I will go through it. 21:37 < jim> they added outer joins and recursive queries 21:38 < clemens3_> postgres came out of the commercial ingres 21:38 < clemens3_> it was head to head straight from the beginning, as far as i remember 21:38 < mgolisch> yeah postgres is quite nice, but id think theres still some areas where commercial stuff like oracle is ahead 21:38 < jim> yeah, at which point it was very primitive 21:38 < clemens3_> with oracle.. maybe oracle added some stuff later.. 21:39 < mgolisch> especialy stuff like parallel query, iam sure that stuff doesnt work very well with postgres 21:39 < jim> oh, they are... but orackle in particular is hearing the footsteps get louder and louder 21:41 < jim> and these days, you should probably consider postgres when you're thinking about oracle,,, you may or may not need the features in oracle that aren't in pg 21:41 < pantato> looks like i'm not banned 21:42 < pantato> whoopee 21:42 < jim> pantato, nope, if you were, it expired 21:42 < mgolisch> packages are nice, which postgres had something like that 21:42 < pantato> jim: not sure how people can stand that Psi-Jackass 21:43 < sensible> I am also very bad at package managament, is there a good tutorial for it? 21:43 < jim> pantato, careful... no name calling 21:43 < revel> pantato: Experience. 21:43 < jim> pantato, don't think about him... better yet, /ignore him 21:43 < pantato> revel: used to the stench eh? 21:43 < revel> If you want to call it that. 21:43 < pantato> jim: no need i'm pretty sure he already ignored me 21:44 < pantato> I don't understand why so many open source people are such dweebs. Like, I thought yall were supposed to be communists, not dweebs 21:44 < well_laid_lawn> only dweebbs say 'yall' 21:45 < bls> pantato: did you come here to ask linux questions or to insult people? 21:45 < revel> Yeah. 21:45 < revel> He did. 21:45 < pantato> well_laid_lawn: 2/10 comeback. Lacks truth 21:45 < pantato> no, i'm here to hang out. I'll stop 21:46 < pantato> just needed to vent my feelings on the matter. 21:46 < jim> pantato, from your own point of view, it's also more about you.... it's possible to cultivate the ability to feel and otherwise notice when you get triggered by someone/something... and that ability can become fast enough, so you get a chance to choose your response rather than just reacting 21:46 < pantato> I'm sure some of you exceed my expectations 21:47 < pantato> jim: i don't get triggered i obsess 21:47 < pantato> anyway seriously that's enough of this convo i don't actually want to get banned 21:48 < pantato> because i know there are people here that are worth being around 21:48 < jim> maybe obsessing is an extreme form of triggering :) 21:49 < jim> if obsessing is not a form of being triggered, maybe it's separate, and maybe you're not noticing it yet? 21:50 < jim> who knows about these things :) the best advice I can give, is be cool :) 21:50 < pantato> jim: i'm just a justifiably very angry person 21:50 < pantato> there's a lot of things to be angry about 21:51 < jim> pantato, do you need to collect new ones? :) 21:53 < jim> maybe when you drill down, the anger is the fire-energy that can be repurposed 21:54 < einveru> Looking for a way to output in terminal 2 text files side by side with nothing displayed in between them. I stumbled on 'diff -y file1 file2' which is close, but draws a vertical line and arrow between. Other suggestions? 21:54 < jim> (using an actual drill is not recommended!) 21:54 < pantato> jim: soon. 21:55 < pantato> tomorrow i'm going to the ICE occupation in PDX since national police forces are scheduled to break it up 21:55 < bls> einveru: do you just want the files? if so, have a look at the paste command 21:55 < plexigras> which config files does nm-applets "Automatically connect to VPN" setting change? 21:55 < bls> pantato: you'll be well served to keep the politics out of here 21:57 < jim> ok... let me say this... when your anger flares up and you need to express it, I invite you to come to ##linux-ops, let us know you're feeling angry and you don't want to express it on the main channel, and express away 21:57 < einveru> bls: hmm, close. Is there a way to format the output, more space between the two outputs? 21:58 < bls> einveru: you can control what's used for the separator 21:59 < bls> hmmm, looks like you only get a single character though. guess you could use a char that won't appear in either file then use sed on that to expand it to what you want 22:00 < clemens3_> einveru: paste file1 file2 | perl -pe 's/...replace/something/' 22:00 < clemens3_> einveru: paste file1 file2 | perl -pe 's/...replace/something/' 22:00 < clemens3_> ups, sorry for repeat 22:00 < einveru> trying to display quotes obtained by 'fortune' side by side. let me give these suggestions a try. Thx bls clemens3_ 22:01 < leftyfb> einveru: paste -d "\t\t' file1 file2 22:01 < clemens3_> there is also: man column 22:01 < TheNH813> Question, how do I make zsh work properly with the home, end and delete keys? 22:01 < clemens3_> haven't used it thought 22:01 < TheNH813> That's the only reason I highly dislike zsh and avoid it like the plague. 22:01 < einveru> clemens3_: tried column but couldn't get it to work right 22:01 < jim> I've used column before, it's worth trying 22:02 < bls> TheNH813: #zsh may know if no one here does 22:03 < jim> einveru, please expand thx as thanks, going forward 22:03 < TheNH813> bls: I'l drop by there and ask then. 22:03 < leftyfb> einveru: how are you displaying multiple fortunes? 22:03 < TheNH813> bls: Thanks. 22:03 < revel> TheNH813: I have lines like "bindkey "\e[F" end-of-line" in my zshrc 22:03 < einveru> jim: Will do going forward 22:04 < jim> thanks :) 22:04 < revel> Try hitting ^V END to see what that prints and stick that in the quoted bit (^[ is \e) 22:05 < einveru> leftyfb: Right now, simply just redirecting each to a file. Beginning stages of getting this figured out. Making a Morning Report/Inspiration sscript :) 22:05 < SuperSeriousCat> TheNH813, what are the 3 keys supposed to do? del remove letters ahead if cursor or me 22:05 < revel> Try searching specifically for bindkey stuff, idk. 22:05 < TheNH813> SuperSeriousCat: Delete spams ~ characters. 22:05 < TheNH813> SuperSeriousCat: Home and end make escape sequences appear. 22:06 < SuperSeriousCat> They do that or are supposed to do it? 22:06 < revel> (they aren't supposed to do that) 22:06 < TheNH813> They do that, but home and end should be skipping to the beginning/end of the line. 22:06 < bls> may also need to adjust stty settings if the keys are producing the wrong sequences 22:06 < TheNH813> And delete, should, well, delete 22:06 < SuperSeriousCat> For me they all work as expected. Home to to start, end to end and del remove ahead 22:06 < TheNH813> Hmmm.... 22:07 < revel> I resolve it with bindkey in the zshrc. 22:07 < TheNH813> I tried some reccommended bindkey values before, but no effect. 22:07 < sensible> lukey_: done with this https://swcarpentry.github.io/shell-novice/ helped me revise my concepts 22:07 < TheNH813> Wait... 22:08 < bls> einveru: if you're doing this with bash, a trick to skip the explicit tmpfile creation would be: paste <(fortune) <(fortune) 22:08 < TheNH813> Does zsh not read .profile? 22:08 < revel> TheNH813: I told you how to get the correct values... 22:08 < revel> ^V and the key itself. 22:09 < TheNH813> Oh, okay. I'l try that. 22:09 < revel> ^V HOME prints ^[[1~ fo me, so I have it as "\e[1~" 22:10 < einveru> bls: Oh perfect! Saves me the "figuring out" part. Now just got to get the formatting correct. 22:10 < TheNH813> My home prints ^[[H so I guess thats what I should use. I'l give it a try and see if it works. 22:10 < mrw0rm> hi guys. Does somebody know what program uses by Linux to resolve DNS? 22:10 < sensible> lukey_: but what I dont uderstand is when I use pwd it shows /home/dellemc and df -Th shows home has 3.5T available but even then fallocate -l 50G /swapfile does not work showing error fallocate: /swapfile: fallocate failed: No space left on device 22:10 < bls> mrw0rm: host, nslookup 22:10 < revel> (remember that the ^[ is \e) 22:10 < Dagmar> dig and/or nslookup 22:10 < mrw0rm> bls: thank you! 22:11 < TheNH813> revel: I use color codes in my bash scripts quite often, so that I do understand. 22:11 < TheNH813> revel: Thanks for the advice. 22:11 < revel> Right. Difficult to know who has what skill level here :P 22:12 < Dagmar> Don't hard-code color codes in your scripts. 22:12 < lukey_> sensible: Note the "/" in the beginning. This means, it will it create in the root directory. 22:12 < Dagmar> Use tput to retrieve the codes and store them in variables. 22:13 < Dagmar> That way, if you happen to be on a terminal that doesn't support color codes (or uses a different encoding) your screen won't fill with unescaped ANSI crap 22:13 < sensible> lukey_: oh yes what a foolishness 22:14 < TheNH813> revel: Indeed. 22:16 < fosstar> How would i stitch together a directory of images in the format of [0-7],[0-7].png using the command line? the first part represents the x axis and the second part represents the y axis 22:17 < bls> fosstar: stitch together into a single image? 22:17 < fosstar> Yes. 22:17 < sensible> lukey_: a swapfile is created in home but when I check it using swapon then this error comes [root@localhost dellemc]# swapon swapfile swapon: swapfile: swapon failed: Invalid argument 22:17 < bls> fosstar: check out the montage command from ImageMagick 22:18 < fosstar> bls: I have that command, however i don't understand the correct use for this case. 22:18 < fosstar> Generally i end up with incorrectly stitched and torn images when i use it. 22:18 < lukey_> sensible: Did you do mkswap first? 22:18 < sensible> yes 22:19 < sensible> [root@localhost dellemc]# mkswap swapfile Setting up swapspace version 1, size = 52428796 KiB no label, UUID=7498c612-2e8d-4955-989d-9dfa8905a0b5 [root@localhost dellemc]# swapon swapfile swapon: swapfile: swapon failed: Invalid argument 22:19 < Dagmar> Specify the full path to the swap file 22:19 < fosstar> The current command i 22:20 < fosstar> i'm trying to use is montage -geometry +7+7 -tile 7x7 [0-7],[0-7].png output.png 22:20 < sensible> Dagmar: [root@localhost dellemc]# pwd /home/dellemc [root@localhost dellemc]# swapon /home/dellemc/swapfile swapon: /home/dellemc/swapfile: swapon failed: Invalid argument 22:20 < fosstar> Nevermind, apologies, i figured it out. 22:21 < Dagmar> sensible: How did you create the file? 22:22 < sensible> Dagmar: [root@localhost dellemc]# mkswap swapfile Setting up swapspace version 1, size = 52428796 KiB no label, UUID=7498c612-2e8d-4955-989d-9dfa8905a0b5 22:22 < Dagmar> sensible: Also, check the bottom of `dmesg` 22:22 < Dagmar> sensible: mkswap does not create files 22:22 < sensible> Dagmar: [root@localhost dellemc]# fallocate -l 50G swapfile 22:22 < Dagmar> You can't use that tool 22:23 < Dagmar> You have to use dd. 22:23 < Dagmar> fallocate will create a sparse file 22:23 < mgolisch> the swapon manpage tells you that 22:23 < tonyt> https://nakedsecurity.sophos.com/2018/06/29/linux-distro-hacked-on-github-all-code-considered-compromised/ 22:23 < Dagmar> Also, 50Gb is probably wildly more than you need 22:23 < tonyt> Gentoo, a popular distribution of Linux, has had its GitHub repository hacked. 22:23 < sensible> Dagmar: [172080.439357] swapon: swapfile has holes 22:23 < Dagmar> To make it very clear, a sparse file can *not* be used as a swap file 22:24 < Dagmar> sensible: ...and the kernel is even telling you so. 22:24 < revel> Better patch those holes. 22:24 < Dagmar> Use dd to create the swap file. 22:24 < jim> tonyt, someone must have a backup of that repo 22:24 < lukey_> sensible: dd if=/dev/zero count=1 bs= of= 22:24 < revel> jim: The main repo wasn't affected as it was on Gentoo infra. 22:25 < sensible> Dagmar: now what should I do to this file of 50 GB and how use dd? 22:25 < revel> The one on GitHub was just for accepting pull requests from outsiders and proxy-maintainers, more or less. 22:25 < sensible> lukey_: what of should be equal to ? name of file? 22:25 < jim> sensible, he gave an example invocation of dd to make the swap file 22:26 < sensible> lukey_: of= 22:26 < lukey_> sensible: full path to your file 22:26 < revel> (relative paths work too) 22:26 < mgolisch> man dd 22:31 < sensible> [root@localhost dellemc]# dd if=/dev/zero count=1 bs=53687091200 of=/home/dellemc/swapfile dd: memory exhausted by input buffer of size 53687091200 bytes (50 GiB) 22:31 < rymate1234> https://www.qubes-os.org/intro/ I really want to try this at some point 22:31 < rymate1234> sensible, try swapping the bs and the count 22:31 < Dagmar> Don't tell it to allocate 50Gb of memory and dump it to disk all in one go 22:31 < tonyt> ya no doubt jim 22:32 < revel> It just needs to be a multiple of bs= and count= 22:32 < tonyt> i dont normally post url's on irc but i figured it was relivent for this channel 22:32 < maryo> We are planning to migrate our codebase from CVS to GIT.. But not sure about the steps on how to perform the migration. Anyone here has done this migration? 22:34 < sensible> revel: how ? 22:34 < revel> Err... Math. 22:35 < lukey_> sensible: oops my fault :) try something like dd if=/dev/zero bs=1024 count=$((1024*1024*50)) of=/home/dellemc/swapfile 22:35 < Dagmar> bs should probably be 4096, just for the sake of efficiency 22:35 < bls> maryo: have only done svn and p4, not cvs, but #git might have some knowledge if no one here does 22:35 < revel> So, count=13107200 bs=4096 22:36 < Dagmar> revel++ 22:36 < Dagmar> Clearly someone else is much faster with the calculator. ;) 22:36 < revel> Just a bit 'o Bash. 22:36 < revel> Ish. 22:37 < maryo> bls, hey I am also looking for SVN to GIT. Could you please give me some hint on how you have performed it? 22:38 < jim> you want to convert an svn repo to a git repo? 22:38 < jim> bisn? 22:39 < maryo> jim, yes SVN to GIT 22:39 < bls> it's been a while, but just essentially followed this: https://git-scm.com/book/en/v2/Git-and-Other-Systems-Migrating-to-Git 22:39 < jim> sec 22:41 < sensible> Dagmar: 52428800+0 records in 52428800+0 records out 53687091200 bytes (54 GB) copied, 260.603 s, 206 MB/s 22:41 < uupz> Hello all, I have a spare laptop lying around. It has a 500gb hd in it, it's nothing fancy or extremely fast, but it's been a champ. I also use vmware to emulate machines on my gaming computer, however I don't always have my vmware running or that computer always on , that's why I'm trying to get linux to be the main os of my operating system. I'm looking for a lighter linux distribution for my laptop, any suggestions? Thanks in advanc 22:42 < sensible> Dagmar: Now I will use mkswap 22:42 < sensible> Dagmar: right? 22:42 < rymate1234> yes 22:42 < lukey_> sensible: Yes 22:43 < lukey_> uupz: That sounds like an good enough laptop for any linux distro. 22:43 < sensible> but first I will do this chmod 600 /home/dellemc/swapfile 22:43 < uupz> Oh cool, I really enjoy Debian. 22:43 < revel> I'd make sure it's root:root as well. 22:44 < jim> sensible, right, now that you have a properly constructed file, you would need to for swap (more steps follow) 22:44 < uupz> Another question for you guys...i have a raspi, is there any Debian based small distros for that? 22:45 < uupz> Or any other lite recommended distros for raspi 22:45 < revel> Err. Raspbian. 22:45 < bls> uupz: your choice of desktop environment, window manager, and default applications will have more impact on hardware suitability than distro choice 22:45 < uupz> I use terminal 85% 22:45 < uupz> Hardly ever get out of it 22:45 < uupz> I run some NIDS software, snort. 22:45 < jim> uupz, not quite sure the details of what you're after... but generally you want to put linux on the laptop? 22:45 < triceratux> sensible: https://www.omgubuntu.co.uk/2018/06/linux-mint-19-released-available-to-download-now 22:45 < uupz> So I need something reliable and up all the time 22:46 < uupz> Where as my main pc I run it on vmware but it's not on all the ti.e 22:46 < bls> so something stable and LTS? 22:46 < rymate1234> so debian? 22:46 < jim> maybe a debian netinstall? 22:46 < rymate1234> https://www.raspberrypi.org/downloads/raspbian/ stretch lite might be for you 22:47 < jim> can the laptop boot from a usb storage, and do you have one? 22:47 < uupz> My raspi has already ran out of space. I put a Kali distro in it. I believe they had a light weight version 22:47 < bls> yeah, debian netinstall would allow you to build up from a lightweight/minimal base to just what you want 22:47 < rymate1234> it's a raspi 22:47 < uupz> But I don't need all those tooks, was for a class project 22:47 < sensible> triceratux: why this for me ? 22:47 < lukey_> uupz: Or ubuntoo LTS server, It even has Kernel Live-Patching :) 22:47 < uupz> Thank you so much guys for the suggestions 22:47 < jim> uupz, which cpu does the laptop have? 22:48 < rymate1234> did everyone miss the raspi part? 22:48 < jim> no, I saw that, which made me curious 22:49 < rymate1234> apparently a 1.4 GHz 64/32-bit quad-core ARM Cortex-A53 22:50 < triceratux> sensible: the xfce version (all of them really) would run on that machine pretty well. & its based on ubuntu 18.04 lts which is a very solid assemblage of software. on a box like that id recommend any xubuntu clone or lookalike 22:51 < triceratux> sensible: but its up to you to get on distrowatch & try some out 22:54 < sensible> ret 22:54 < lukey_> triceratux: Yep, 128G Ram is just enough to run xfce :P 22:54 < sensible> triceratux: thanks for recommendation 22:54 < birdbolt1> bash is incredibly annoying 22:54 < Pentode> you'll need at least another 128 to run firefox 22:55 < sensible> https://paste.linux.community/view/ef5add54 show my /etc/fstab 22:55 < birdbolt1> `/bin/bash: /home/base/app-entrypoint.sh: Permission denied` 22:55 < triceratux> oops im not sseing straight. all that sensible stuff was for uupz 22:55 < revel> That doesn't sound like a bash problem. 22:55 < birdbolt1> on alpine linux in a docker container 22:55 < birdbolt1> there is no user but root 22:55 < lukey_> birdbolt1: chmod +x 22:56 < sensible> Is there any issue in this file lukey_ ? https://paste.linux.community/view/ef5add54? 22:56 < birdbolt1> even as root, a shell script must be made executable? lukey_ 22:56 < mgolisch> sure 22:56 < revel> Yes. 22:56 < birdbolt1> learn something new everyday thanks 22:57 < revel> If you want to execute it directly, anyway. If you do `bash /path/to/script.sh`, then you don't. 22:57 < birdbolt1> that is infact just what i did 22:57 < birdbolt1> `ENTRYPOINT ["/bin/bash", "-c", "/home/base/app-entrypoint.sh"]` 22:58 < lukey_> sensible: looks good to me. 22:58 < revel> Err. That's not the same thing. 22:58 < revel> Drop the -c and it'd work. 22:58 < birdbolt1> that gets me the permission error 22:58 < birdbolt1> oh? I saw the -c in a docker docs page, it was recommended or something 22:59 < sensible> when running a tool that is memory hungry then this value is not good ? [root@localhost dellemc]# cat /proc/sys/vm/swappiness 60 22:59 < revel> That makes bash spawn a process that executes that script (as a regular executable). 23:00 < revel> Like, you can do `bash -c "echo yay"` and it'll output "yay" 23:00 < sensible> And therefore in my system memory is used 66% and swap is used 10% 23:01 < dkz> Hi 23:02 < lukey_> sensible: That just controls how soon the kernel swaps things out. 23:02 < sensible> lukey_: but I want my server to use all memory and then move to swap as I have to use tools like STAR 23:03 < sensible> lukey_: is this ^^ right? 23:05 < lukey_> sensible: The default is good, yes. This Setting is just for some minor performance tuning. 23:08 < sensible> lukey_: you mean that If I set swapiness to zero then it will use swap after completely feeding the memory? and it does not cause any problem to system? 23:11 < lukey_> sensible: More or less, yes. In the end a higher value means more troughput but also more latency and vice-versa. 23:17 < sensible> so swapiness is set to 10 and vm.vfs_cache_pressure = 50 23:17 < sensible> thanks fellows for help about swap its working fine now. 23:18 < triceratux> sensible: np. my apologies for recommending mint :P 23:19 < sensible> I have set the values of ulimit -n to 100000 so whenever I open terminal I get this bash: ulimit: open files: cannot modify limit: Operation not permitted 23:19 < birdbolt1> how do i copy all folders from current directory 23:19 < birdbolt1> I'd like to exclude files in current directory 23:22 < sensible> also I have set fs.file-max=500000, how this error can be removed ? bash: ulimit: open files: cannot modify limit: Operation not permitted 23:23 < lukey_> sensible: You'll have to raise the hard limit in /etc/security/limits.conf 23:23 < Elladan> birdbolt1, recursively? Do you want the files in the directories too? 23:24 < birdbolt1> Elladan, yup 23:24 < runjutsu> kind of lazy question, but uh.. yeah, it's really lazy. But is there something to remove all blue light from the screen with on linux? 23:25 < birdbolt1> I'll be using docker command COPY, which does that by default 23:25 < xamithan> You mean like a night mode ? 23:25 < runjutsu> aaaaand all I had to do was to google it. Never mind. 23:25 < runjutsu> yep. there's a program called redshift, comes with most distros :) 23:25 < xamithan> Thats a DE thing, not a linux thing 23:25 < runjutsu> all you have to do is install it 23:26 < Elladan> birdbolt1, you can list all the directories in the current directory like "find . -maxdepth 1 -type d" and can then pass that list to a suitable copy command. 23:26 < revel> Surely you mean -type f 23:26 < Elladan> revel, that would list the files. 23:27 < sensible> lukey_: I have already increased it but still the same error https://paste.linux.community/view/bfb81c0a 23:27 < revel> Oh, wait. Should've read more. 23:27 < birdbolt1> Elladan, ty will try 23:27 < lukey_> sensible: did you log out and in again 23:28 < sensible> lukey_: yes 23:31 < runjutsu> how do I drop a program to the background? It dies when I quit bash 23:31 < Elladan> birdbolt1, you may wish to learn how to use "find {arguments} -print0 | xargs -0 {arguments} ..." for doing things like this. It's a powerful pattern. 23:32 < bls> or use the much less error prone find...-exec instead 23:32 < Elladan> runjutsu, "disown" (after bg) 23:33 < lukey_> sensible: Try putting "session required pam_limits.so" in /etc/pam.d/common-session if that file exists and try again 23:33 < runjutsu> Elladan: wow amazing, still running. I thought bg jobs quit when you closed the seession? 23:34 < Elladan> runjutsu, disown removes them from shell control, similar to nohup. 23:34 < bls> runjutsu: they do unless you properly separate them from the session so the signal to kill that session isn't propagated to its children 23:34 < sensible> lukey_: [root@localhost dellemc]# ls /etc/pam.d/common-session ls: cannot access /etc/pam.d/common-session: No such file or directory 23:36 < lukey_> sensible: what files do you have in that directory? 23:38 < sensible> lukey_: I have these files https://paste.linux.community/view/77cd4afe 23:39 < lukey_> sensible: put it in login, at the end 23:40 < runjutsu> bls: looks like the program (in this case redshift) ignores such signals 23:41 < runjutsu> bls: the job would otherwise have died as expected I suppose 23:41 < sensible> lukey_: ok I am logging out to check so browser will also be closed 23:41 < bls> that or it daemonizes/double forks 23:41 < runjutsu> beloved people, install redshift! It's amazing! 23:42 < bls> which one? :P 23:43 < mgolisch> was something about changing colors of the screen or so 23:44 < bls> OK, there's also something called that for managing AWS stuff and for doing GPU rendering 23:45 < cxc99> would s3deploy inside a docker container be considered a microservice? 23:52 < fiter> lukey_: after logging out and logging again the same error bash: ulimit: open files: cannot modify limit: Operation not permitted 23:54 < JordiGH> How do printers work if they don't use postscript? Like, what kind of data does my computer send to the printer otherwise? 23:54 < JordiGH> Does a driver handle that and then they use whatever printer-specific protocol they can think of? 23:55 < bls> JordiGH: the printers have their printer control language or PCL 23:56 < bls> JordiGH: you generally get a set of postscript macros/.ppd file or some other sort of software that has traditionally let ghostscript translate from PS and/or PDF to the PCL 23:56 < JordiGH> Of which Postscript is but one example? 23:57 < phogg> JordiGH: printer drivers talk to printers in their native format, which might be a multi-device standard or device-specific. 23:57 < kerframil> PDL would be the better term, as PCL is a Hewlett-Packard thing 23:57 < runjutsu> BNC from generator to oscilloscope .. what kind of BNC do I need? Female to female (cable is female) or male to male (cable is male and oscillator/generator are fiemale)? 23:58 < JordiGH> Btw, is PDF also a printer language? 23:58 < phogg> JordiGH: early postscript printers were cool but expensive. WHen the cost of the embedded computer to process the postscript started to be the biggest cost of the printer manufacturers offloaded the processing to the computer's CPU instead and started sending more pre-rendered data to the device 23:58 < phogg> JordiGH: kind of, it's complicated 23:58 < bls> some printers can interpret PDF (or at least the parts of it that are easily translated back and forth to postscript) 23:59 < phogg> these days I'm thinking that putting native postscript support into a printer would be a minimal cost, given the electronics they already have in them 23:59 < phogg> but it's considered a high-end feature so printers which accept raw .ps files are more expensive in any case --- Log closed Sat Jun 30 00:00:08 2018