IRC Logs for #crux-devel Tuesday, 2014-12-30

*** leo-unglaub has quit IRC00:03
*** Feksclaus has quit IRC00:17
*** tvaalen has quit IRC01:16
*** tvaalen has joined #crux-devel01:29
*** mavrick61 has quit IRC03:41
*** mavrick61 has joined #crux-devel03:42
Romsterjaeger, we should have a prt-get.git tree on crux.nu since cptn isn't active05:19
Romsterhow many have partial rewrites of pkgutils and/or prt-get?05:22
Romsteri know jaeger has a C pkgutils not published anywhere05:22
Romsterdon't know if jue started on anything05:22
jaegermine is vestigal at best, It hink teK_ has a lot more. Maybe we should just start developing his version in earnest05:26
jaegerfor now, going to sleep, night05:26
Romstershould start a git hub or something or on crux.nu so a bunch of us can work on the same one.05:27
Romsternight jaeger05:27
*** jue has joined #crux-devel10:17
juegood morning10:17
jueRomster: you got the mail sent to xorg-ports mailing list wrt libvdpau-32?10:19
jues/mailing list/mail alias/10:20
Romsteractually haven't been checking my email. does that now. and i did see your commit on libcdpau10:22
Romsterlibvdpau*10:22
Romsteri never had any documentation issues.10:23
jueI looked into that issue a bit and found out that libvdpau builds some docs if doxygen, graphviz and tex is installed ;)10:24
jueall togther10:25
Romsteri'm not sure if i'm getting emails from xorg-ports10:25
Romsteram i even on that list.10:25
juethe alias file on crux.nu says so10:27
jueis monster.romster@gmail.com still valid?10:27
Romsterits still valid.10:27
Romsterand i got crux@romster.me forwarding to that. gmail for now but i'll change that later on.10:28
Romsterah here it is i found one from xorg-ports jue10:28
jueok, so it works :)10:28
Romsterits in my inbox and didn't get on my filter to the crux directory.10:29
jueteK_: may I ask you to add frinnst to xorg-ports?10:29
Romsteris there a list somewhere on who's in what git tree access and mailing list?10:30
juewe have the usual mailman interface here -> http://lists.crux.nu/mailman/listinfo10:32
juedunno about a list with git users10:34
Romsteri keep forgetting who's in what group. and for compat-32 its jsut jaeger and myself, tek? xorg compat-32 is what i'm wondering. i did know but since forgotten.10:35
Romsterlibvdpau-32 fixed.10:36
juethanks10:36
Romsterand add a filter rule so it's in my crux directory.10:39
Romsterdocumentation can be a pain. only sows up if one has the necessary dependencies.10:39
Romsters/sows/shows10:39
teK_jue: done11:36
teK_I am  not in compat-32 or xorg; http://crux.nu/gitweb/?p=ports/compat-32.git;a=commit;h=c7ac6daabea85c57271bf7b2f873ceab7c81413c was an exception ;)11:37
teK_@Romster11:37
RomsterteK_, oh so you abused your powers to make that commit << lol11:42
teK_absolutely11:43
Romsterhehe11:49
Romsterjue i keep getting missing symlink files in vim adding-j2 to make install on both instances fixes that.12:05
jueteK_: thanks12:49
jueRomster: is your cccrap still worth the trouble you always have :)12:50
Romsterjue, which cccrap?12:59
Romsteroh my -j2012:59
Romsterit works fine12:59
diversedistcc?13:00
Romsterit's the crap of shitty Makefiles that arn't designed for high levels of parallel jobs.13:00
jueRomster: just kidding, no I meant your ccache/distcc setup13:00
Romsteryeah it works13:00
diverseI plan on using distcc in the future even13:01
Romsterpump mode is problematic so i stopped using that.13:01
Romsterbut distcc is fine.13:01
Romsterccache has only ever failed me once but not since i do export CCACHE_COMPILERCHECK="%compiler% -dumpversion; crux"13:02
Romsterif i had a xeon with 24 cores or so maybe i'd give up on distcc.13:03
Romsterall at 4GHz and not some crapy bogomips per a core.13:04
Romstersome cups have crappy per a core calculations.13:04
Romstercpus*13:04
Romsterin some cases some cpus work better with less cores.13:05
jaegerteK_: forgot to comment when we talked about it the other day - nvidia 343.36 is the "short lived branch" - it's stable currently but isn't always. the "long lived branch" still doesn't support the 900 series and that's what's in opt and compat-3215:14
teK_hmkay15:14
teK_343.36 does not work with 3.18(.1) ..15:14
frinnstyou guys should seriously consider buying amd/ati instead :)15:15
teK_sure..15:16
teK_dream on15:16
jaegerI've fought with amd/ati in the past, never again on linux15:17
jaegerIf you're using a card that's supported by the xorg ati driver it's fine... but those are mostly older than I want to use15:18
jaegerfglrx can get fucked15:18
teK_I do steam. And I do  it efficiently and  without hassle. So no thanks.15:19
jaegergot some ati cards in linux workstations here at work, they're still problematic... one of them (7750 I think?) causes a crazy hall of mirrors effect sometimes, have to remove and reinstall the driver to fix it. I have NO idea why that would matter but it works. Maybe it's an ubuntu/unity problem rather than ati, I haven't researched it deeply15:19
jaegerthe binary blob thing doesn't bother me like it does some FOSS evangelists... I just want my graphics to work properly and quickly :)15:21
diverseI'm currently using 343.36 with no problems at all16:23
diversebeen using the short lived branch ever since I got my gtx 980 2 months ago16:25
teK_butnot with the 3.18.1 kernel16:25
diversewhat happened on the 3.18.1 kernel? I'm currently using 3.17.716:26
teK_dunno. You never know if theyfixed nasty stuff ;)16:26
diverse?16:26
teK_they dont always mark commits that fixed security vulnerabilities16:26
diverseah it doesn't work...16:28
diverseweird16:28
diverseI guess I'm stuck on 3.17.x kernels16:28
*** Feksclaus has joined #crux-devel16:37
frinnstthe ati driver supports hardware quite quickly these days16:42
frinnsta few months or so behind16:42
diverseyou mean the open source driver?16:47
jaegerfrinnst: that's a nice improvement18:04
diversewell I guess open source drivers are better than their original proprietary version sometimes (*hint hint* Qualcom Atheros Killer)19:05
diversebut getting an ati card for jaeger's purposes won't serve him well for windows19:07
diversebut for linux only setups, ati would probably be a good deal19:08
teK_btw jaeger what's  the reasoning behind not using the short  lived branch? It's marked  as "stable"20:08
jaegerteK_: just because I personally favor the long lived stuff. Once in a while the short lived branch has been buggy, though its current incarnation is very stable20:28
jaegerI've also considered just adding an 'nvidia-sl' port or something20:29
teK_I  see :)20:32
diversethat would be nice21:01
jaegerhehe21:36
jaegerteK_: what's the current state of your pkgutils rewrite stuff?22:48
jaegerDoes anyone object to creating a prt-get git tree on crux.nu? I'll set it up if not22:48
teK_.irssi/logs/2014/Dec/#crux.log-10:37 < diverse> teK_: have you played on your new implementation of prt-get any?22:50
teK_.irssi/logs/2014/Dec/#crux.log-10:40 < teK_> not yet, REALLY busy right now22:50
teK_.irssi/logs/2014/Dec/#crux.log:10:40 < teK_> (until ~May '15)22:50
jaegerAh, gotcha22:50
teK_I did not magically finish the company project and my master thesis :\22:50
jaegerI missed that conversation, sorry22:50
teK_no problem at all22:50
teK_I hope things get traction only after that date, tbh :|22:52
jaegercrux things or life/work things? :)22:53
teK_the pkg stuff22:54
diversesigh22:54
jaegerdiverse: I've had no problem with ati/amd cards in windows, it's the shit linux drivers (the proprietary ones)22:54
teK_bute life will get traction after that date, too. I will work only one job at a time after May :)))22:54
jaegerteK_: I'd still like to develop my stuff in the meantime but I won't be offended if it doesn't get used22:55
teK_sure22:55
teK_nobody asked us to do that so there's not much right to be offended for either of us :p22:56
teK_yet I/you probably would like to see mine/your stuff succeed :)22:56
teK_man. Switching from 4k@27" to 1280@12.1" suuuucks22:57
jaegerI would like it to succeed, yes, but I'd rather see us get the best of all offerings from collaboration :)22:57
jaegerheh, I bet22:57
teK_especially if rdesktop sessions with 1080p were used22:57
jaegereven switching from my 2560x1440 at home to 1920x1200 at work is weird22:58
teK_I have a feeling that there will be no success without collaboration22:58
diversejaeger: I thought the window drivers were also terrible as been mentioned to me22:58
teK_I figure22:58
jaegerdiverse: I didn't have much trouble with them but I also haven't used them in years22:58
diversebtw, I've lost the motivation and interest in my project recently23:08
jaegersorry to hear that23:08
diversethanks for the sympathy. After the discussion with hhhhhhhh, I realized I got burned by Mozilla.23:10
diversewell it was fun while it lasted23:10
jaegerjue: was the change to /etc/rc.d/net for posix-compliance or just preference? :)23:11
teK_diverse: how did they do that?23:12
diverseit's somewhat complicated to explain, so to not bore you, it has to do with them inventing the stupid tool called "cargo"23:19
teK_you coded in rust, didnt you?23:20
diverseyeah23:21
teK_hm23:23
teK_reads like something optional23:24
diversehow do I explain it shortly?23:25
teK_you dont have to if you dont like23:26
diverseit's a package manager for rust which relies on a central repository called crates.io. The problem is, it uses a registry for pulling dependences on-the-fly. Most rust projects are blinding conforming to cargo and leaving mp room for building it based on what the distro already prebuilt for you. The rust devs even segmented a lot of the libraries from language into different git repos for them to be only23:34
diverseaccess by cargo on purpose. I find this stupid, because if the server goes down and I need to access/update a dependency for my project, I'm screwed.23:34
diverses/mp/no23:34
diverseIt makes things more complicated then it needs to be as well, but that's to be expected of Mozilla's over engineering like how hhhhhhhh said23:35
diverseso what was already a great upcoming language, ruined by inventing a language package manager because it's "trendy"23:37
diverseno wonder C and C++ lasted this long23:38
diverseanyway this is the devel channel, so I'm not going any deeper for the sake of staying on topic and for frinnst23:39
diversewell to mention one thing on topic, it's been fighting me on CA certs again, what a pain.23:45
teK_hm sounds a bit like what maven is doing but with the centralised aspect23:46
teK_on the other hand.. you dont have to use it for your code, right?23:46
diversefor mine, no, for others', yes23:48
diversebtw, what is maven?23:48
diversewho23:48
teK_a build tool for java projects23:49
diverseah23:49
teK_it downloads deps on the fly, too23:49
teK_it can help you deploy things to a Tomcat server or whatever, too23:49

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!