IRC Logs for #crux-devel Thursday, 2007-12-06

*** roliveira has quit IRC02:36
treach@seen sip03:24
clbtreach: sip was last seen in #crux-devel 6 days, 14 hours, and 45 seconds ago: <sip> yeah, I'm always lucky with this kind of things03:24
*** Man0l0 has joined #crux-devel04:38
*** DarkNekros has joined #crux-devel04:52
DarkNekroshi you all ;)04:53
clbUpdate from opt: 6 Dec 11:57 - [notify] openoffice: updated to 2.3.105:19
*** jaeger has joined #crux-devel09:05
*** jjpk has joined #crux-devel10:00
* jaeger kicks gstreamer 0.10.15 in the nuts10:09
Romsterjaeger, i can't compile it... 14 works but 15 seems broken10:17
Romsteror i did something wrong.10:18
Romsteri'll stay at 14 and wait for 16 to come out i guess10:18
jaegerit is broken but no solution yet besides downgrading10:23
tilmanbuild or runtime problems?10:24
jaegerbuild10:24
jaegerundefined reference to _priv_gst_info_start_time10:24
tilmanhttp://bugs.lunar-linux.org/view.php?id=31310:25
tilmanthey have given up on it10:25
tilman:D10:25
treachhaha10:25
jaegeryeah, saw that10:25
jaegerunfortunately there's not much more to be found than their bug :/10:26
jaegergstreamer's bug interface sucks10:26
treach"I've marked this as minor, because I'm uninstalling gnome anyway." ROFL10:27
jjpkOutsourcing the blame \o/10:28
jjpk"I don't use gnome anymore, you are on your own from now"10:29
Romsterand i thought i did something wrong and wasted time looking for a bug.10:29
treachthat's hilarious. "I don't use this sw, thus the bug isn't really bad."10:30
treachduh.10:30
jaegerthat really instills a sense of confidence10:31
treachyeah.10:31
jjpkA nice shit out of luck situation. :p10:32
Romsterwell that is not nice..10:33
Romsterwhat about everyone else that selfish bugger <<10:33
treachindeed.10:33
treachthe bug is *still* serious. that *he* isn't affected is completely irellevant..10:34
treach*irrelevant*10:34
Romstervery true..10:34
Romsterif it don't compile it's a serious bug if he uses the program or not.10:35
Romsteri don't use gnome but i use gstreamer.10:35
* treach walks wide of both10:36
Romsterlol10:36
Romstersomehow i think it's a bad idea to have one program bundle all codecs.10:36
Romsterits good being in one place but also gives a greater risk of breaking too.10:37
Romstermaybe renaming it back would work <<10:38
Romsteri never tryed..10:38
Romsterjaeger, what file was that error in i'm gonna have a look at the diff on there CVS10:42
*** jue has joined #crux-devel10:42
Romsteror do i have to try again to get the file that has that error from.10:42
jaegerthere's a log attached to that lunar bug10:43
jaegerhttp://bugs.lunar-linux.org/file_download.php?file_id=85&type=bug10:43
Romsterah10:43
juetilman: any schedule wrt xorg-server update?10:43
tilmanjue: not afaik :/10:43
jueso we should wait with rc1 for that?10:44
tilmanimo no10:44
jueok, fine10:45
* Romster would like to see the lzma patch in pkgutils for crux 2.410:46
tilmanno way10:46
Romsteror i'll keep using the hack i've done.10:46
Romstersuit yourself.10:46
jaegerjust out of curiosity, why are you so against it, tilman?10:46
juejaeger: .oO any news at the docu front?10:47
jaegerjue: which part? the new iso bit and PKGMK_SOURCE_MIRRORS? or was there something else I missed? I'm hoping to get a chance to do those today10:47
Romstersomething comes along betetr and tilman won't have support for it... that blows. yeah i'd like a reason for it being rejected.10:47
jaegerwell, better is subjective... I'm just curious10:48
tilmanshow me a single source tarball that uses lzma10:48
tilman:|10:48
Romsteri'm seeing lzma to be taking over bzip2 evently and crux is all about new technologys afterall.10:48
Romstertilman, off my site <<10:49
jueimo we should have the final ChangeLog and RelaseNotes in rc1, so, if everything goes well, we can just rename the iso10:49
Romsteri'm sure there will be mroe its the case of 'build it they'll come'10:49
Romstermore*10:49
juejaeger: yes, mostly the first one, the iso bit10:50
Romsterbut suit yourself i'll keep my own version and keep the hack in Pkgfile10:50
jaegerjue: no objection here, I'll try and get it done in the next couple hours10:50
juetilman: coreutils ;-)10:50
Romsterso much for crux being open to advancement and new technology.10:50
tilmanjue: good one, i've even seen that yesterday10:51
jaegerRomster: rather than complaining about how tilman doesn't like it, why not come up with a good list of advantages and disadvantages (if there are any) and some examples, and post it to the devel ML?10:51
tilmanif we add support to it in pkgmk, we'd need lzma in core10:51
tilmani think something like:10:51
tilmanif (source_is_lzma && have_lzma)10:52
tilman  try to use lzma to unpack10:52
tilmanis silly10:52
tilmanotherwise we'll get lots of confusion similar to the stupid pkgutils-needs-libarchive now situation10:52
Romster1) if i could use the ML they still haven't fixed it for Australia yet... 2) i thought the lzma site showed enough stastics but ok i'll get more good and bad points if that's what it takes,10:52
jaegerRomster: I realize the ML is a pain in the ass for you guys in .au but you could subscribe from gmail or something10:53
jaegertrust me, if I could fix the ML, I would10:53
tilmanactually10:53
tilmanlet me mail charlie again10:53
tilmanhis responsiveness this morning was excellent10:53
jaegerwell, that's good at least10:53
jaegerRomster: that's part of it but also, what's the intent? is it for source tarballs or packages? or both? what's the real benefit? is the iso too large now? etc. etc.10:54
tilmanaaaalso10:54
tilmanwe started release preparations in august, now it's december10:54
jaegerthe reason I ask is because we could have already added bz2 support but that hasn't happened either... so why is lzma better than bz2?10:54
tilmansneaking lzma in *now* doesn't seem like the best idea wrt release schedule10:54
Romsterk i understand that so later on then..10:55
tilmanjaeger: as i understand romster would like lzma *decompress* support10:55
jaegerI'm not trying to be combative, either. If lzma is the shit and we NEED it then I'd be happy to see it included10:55
tilmanjaeger: ie in pkgmk10:55
Romsteri'm only after the decompression support..10:55
jaegerI haven't seen any lzma source tarballs myself but if that becomes a common thing, then yes, it would make sense to add it at some point10:55
Romsteri'd like to use lzma for packages instead of gz but that's not important.. and is too close for 2.4 to even consider.10:55
tilmanwow10:56
Romsteri'm now using lzma becasue it compresses smaller and extracts faster.10:56
tilmancoreutils-6.9.90.tar.lzma vs coreutils-6.9.90.tar.gz10:56
tilman~3.5 mb vs ~9mb10:56
Romstersee10:56
jaegerwell, that's a good example :)10:56
jaegerbut if we wanted to use lzma sources right now we'd have to host them somewhere10:57
Romsterimagine all the man pages and packages using lzma?10:57
tilmanyou're confusing pkgmk and pkgadd i think10:57
juetilman: have you read this as well ? http://lists.gnu.org/archive/html/bug-coreutils/2007-12/msg00015.html10:57
tilmanlet's postpone it to 2.510:57
tilmani think it's feasable to aim for libarchive lzma support until the next release, too10:58
tilmanjue: reading now10:58
Romsterall iased was for decompression support now before it's too late to be in crux 2.411:00
jaegerhrmm... I think I should remove that crux 2.1 with udev doc from the wiki, heh11:00
Romsteri asked*11:00
jaegeror at least remove the link on the documentation page11:00
Romsterthat's quite old..11:00
jaegerindeed11:01
Romstertilman, so maybe you are impressed now?11:05
j^2what's going on guys?11:06
jaegerheyo, j^211:06
j^2my insulin pump broke yesterday :-O11:06
jaegerdoh11:06
j^2needless to say i was scared11:06
tilmanRomster: let's add it before 2.511:06
Romsteri'm wanting to tryal an idea for torrents for source files, but i dunno if i got enough programming expeareance todo that..11:06
Romstertilman, ok :)11:06
Romsteri realsie it's not sued alot now but i'm sure it will be.11:07
Romsterused*11:07
j^2Romster: thinking of using torrents to distrubute what again? (sorry i just jumped it)11:07
tilmanconsidering that bzip2 hasn't even become the standard i doubt it a bit :p11:07
j^2s/it/in11:07
jaegerhttp://crux.nu/Main/ReleaseNotes2-411:08
jaegerhow does that look for the ISO bit?11:08
tilmanjue: what's needed to make you agree with making bsdtar (from libarchive) _the_ tar on crux? :D11:08
Romstertilman, hmm bzip had to be renamed bzip2 for some stupid reason11:08
tilmanwhat? how is that related to anything?11:08
Romsterand bzip2 is slow too. but lzma can be slow at -9 compression.11:08
tilmanjaeger: looks good to me11:09
jaegerI will put the same text into the ChangeLog in git if there are no objections/suggestions11:10
j^2not that it means anything ;) but the notes look good to me too :D11:11
tilmand'oh, i need to update the x versions in ChangeLog, too11:11
j^2short and sweet11:11
jaegerand we'll need to remember to update the release date :)11:12
Romsterwill util-linux-ng make it into 2.4?11:13
jaegerj^2: give yourself some credit, input is appreciated :)11:13
tilmanRomster: already has11:13
Romsterah cool.11:13
j^2jaeger: :D11:13
Romsterjaeger, better write about util-linux-ng on the wiki too?11:14
jaegerwhat's vim's default tab size? 8?11:15
juejaeger: looks good to me11:16
Romsterhotplug in core hmm we don't need this anymore?11:16
tilmanjaeger: dunno, but you'll need ts=8 for CHangeLog :D11:16
jaegerthat's what I'm using11:16
Romsteri think it's 8 but i think 8 is insane and should be 4, although i use a tabstop of 311:16
jaegerI use 4 normally11:16
jaegerwanted to make sure it looked fine in a default session11:16
tilmanyou can use less to check11:17
juetilman: never looked at it, is it a complete replacement for gnu tar?11:17
tilmanjue: yup11:17
tilmanjue: well, i think it's missing some support bits for tape usage11:17
tilmanit's used in freebsd6, so it can't be totally broken11:18
jaegerit looks a tad odd, I spaced some of it instead of tabbing, looks better that way11:18
j^2jaeger: yeah it's 811:18
jueok, will have a look, do we have a port for it?11:18
tilmanjue: http://crux.nu/portdb/?a=getup&q=tilman11:19
jaegerpretty sure hotplug isn't necessary anymore but I'll leave that to the udev experts to figure out11:19
clbUpdate from opt: 6 Dec 17:01 - mysql: update to 5.0.5111:20
jaegerhrmm, git has done something very odd again and added random whitespace to my shit11:21
jaegerif I have committed but not yet pushed will git reset --hard remove my commit so I can redo it?11:21
tilmangit reset --hard HEAD^11:21
tilmanoh wait11:22
tilmanyou can also edit ChangeLog now, fix the errors11:22
tilmanand do:11:22
tilman  git commit --amend ChangeLog11:22
jaegergood to know, thanks11:22
tilmanit's much more likely that vim is to blame rather than git btw ;)11:22
juejaeger: please fix the kernel-version :-)11:22
jueI forgot that11:22
jaegertilman: in this case I think not... there was a change made between my edit and push... I checked for whitespace11:23
jaegerjue: ok11:23
jaegerok, pushed the updated changelog, let me know if there's a problem11:26
jaegergotta meet someone for lunch, I'll be back after11:26
jaegerwill add the bit about PKGMK_SOURCE_MIRRORS to the handbook when I get back11:27
tilmanjue: oops, i see you just added the missing x stuff11:29
jue?11:29
tilmanwondered why my patch wouldn't apply anymore after getting jaeger's change11:29
tilmanhttp://crux.nu/gitweb/?p=system/iso.git;a=commitdiff;h=8c948ecc352a08abbae4b6342395d479660640d311:29
jueah, yes11:29
jueI have some scripts to generate the changes11:31
tilmanah, i was looking through 'git log' output11:31
juethe package-{name,version,collection} are stored in a sqlite db for each crux version11:33
jueand than some SQL magic11:33
tilmanneat11:33
treachspeaking of sqlite, tilman.. my little paper isn't dead yet, I've just been kind of busy lately. :>11:35
treachhopefully I'll get around to revisit it soonish. :)11:35
tilmani also need to get my act together and work on pkgutils6 again11:37
treach:)11:38
treachRomster: btw, "Originally, bzip2's ancestor bzip used arithmetic coding after the blocksort; this was discontinued because of the patent restriction to be replaced by the Huffman coding currently used in bzip2.11:38
treach"11:38
treachsayeth the mighty wikipedia.11:38
treachthus, bzip2 != a renamed bzip211:39
Romstertreach, ah yes i had read that but i forgot...11:39
treachthus, bzip2 != a renamed bzip*11:39
Romstersame reason why libungif is gonna be now renamed to giflib11:40
Romsterbut i think lib<name> is the better format.11:40
tilmannot relaly11:40
tilmanbut oh well11:40
Romsterpersonal opinion.11:41
Romsterbut dosn't really matter as long as it works right <<11:41
tilmani wonder why anyone still uses gif the format11:43
Romsterno idea libmng should be supported in browsers, but it's not that welll supported yet.11:44
Romsterand firefox3 is trying to push there apng format11:44
treachhaha11:44
Romsteranimate png.11:44
Romsteranimated*11:44
Romsterand the png comunity don't like the hack.11:44
Romsterand why would they wanna mess with apng when there is mng anyways?11:45
treachI swear, romster's keyboard would die from shock if he ever managed to produce a sentence without errors or mistakes. :D11:45
Romstertreach, i've done a few rare lines of no errors.11:45
tilmanthere's one11:45
Romsterlike that one!11:45
tilmanand another11:45
treachwhee! :D11:45
Romsterheh11:46
treachsorry, getting very tired and silly over here, no harm intended.11:46
Romsteris the new_iso branch going to be merged into the 2.4 branch?11:46
Romsterit's ok treach11:46
tilmanRomster: i think i'll just drop 2.4 eventually, and rename new_iso11:47
Romsterk11:47
tilmanprobably *after* the release, to prevent strange errors11:47
Romsterk sounds good.11:48
sepenwhy I can't use 2.4 iso as rescues cdrom, normally I cant type at boot prompt something like CRUX root=/dev/hdax and now are a few options only11:51
sepenmany colors, less boot options11:52
jueit doesn't autodetect your cdrom drive?11:55
sepenjue, it can't boot my hard drive12:07
sepenI try on two machines, my laptop trying with CRUX root=/dev/sda1 and my desktop with CRUX=/dev/hda112:08
sepenIm trying to boot with the cdrom to use it as a rescue distro purpose too12:08
*** Man0l0 has quit IRC12:09
sepenthat feature dissapear on 2.4  http://crux.nu/gitweb/?p=system/iso.git;a=blob;f=iso/isolinux/roots.msg;h=900686d2948add4538b269b9fedceac9018c44b7;hb=2.312:10
juewouldn't it make more sense to boot the iso, have a running system and than look at the harddisk problems?12:10
sepenno hardisk problems here, I just want to say that roots.msg dissapear from the boot iso prompt12:12
sepenwhat was the reason?12:12
juehave a look here: http://crux.nu/Main/ReleaseNotes2-412:13
jaegershould a note about PKGMK_SOURCE_MIRRORS go into Handbook2-4-Package or -Ports? probably -Package12:14
sepenanother thing, take a look to http://crux.nu/gitweb/?p=system/iso.git;a=blob;f=iso/setup.dependencies;h=2c4befd1a31dd06514c0933a6d568a0d51b7acc2;hb=new_iso12:15
jueyep Package seems the right place12:15
sepensee line 72 prt-get: prt-get12:15
sepenshould be prt-get: pkgutils prt-get    no???12:16
tilmanHandbook2-4-Package is missing a nice paragraph about pkgmk.conf in general12:16
juesepen: no, see http://crux.nu/Main/PortGuidelines#ntoc212:17
sepenjue, well Release Notes are correct, cdrom way is fixed,  but other times I use crux cd to boot my hda partition when I reinstall or delete my master boot record due to windows installation12:17
sepenjue, yeah! but I think I can install only some packages from core, and select only prt-get without select pkgutils12:18
jaegerI could go back and add support for overriding the cdrom search for the next release12:20
juesepen: hehe, yes you can fortunately. But you will have a broken system afterwards ;-)12:20
jaegerbut even with this new iso you can mount up what you need and chroot into it12:20
sepenjue, also I think http://crux.nu/Main/PortGuidelines#ntoc2 are fine for Pkgfiles but not for a cdrom setup.dependencies file12:20
sepenxDD thats the pont12:20
sepen*pint12:20
sepen*point12:20
sepenjaeger, yeah, now Im forced to use chroot to pivot to my root filesystem, also its perfect for me12:22
sepenbut I think 'roots' cdrom option could be better12:22
juesetup.dependencies is autogenerated with prt-get. I doubt that we will find someone who is willing to do that manually.12:22
jaegeryou shouldn't even need to pivot if you're just fixing an MBR12:23
sepenohh true12:23
*** jue has quit IRC12:27
*** treach has quit IRC12:28
jaegertilman: do you want to add a pkgmk.conf section or do you want to try and come up with something useful?12:29
tilmanjaeger: if you don't mind... :D12:31
jaegerok12:31
jaegerI'm thinking just before the prt-get section12:31
tilmanyup, that's a good spot12:32
*** maro has quit IRC12:32
jaegeror perhaps it'd be more intuitive to put it after "Creating Packages"12:32
jaegerhrmm12:32
tilmansec12:32
tilmanoops12:32
tilmantrue, pkgmk isn't mentioned earlier12:32
jaegeryeah, that's what I was thinking12:33
tilmanmaybe before Package Guidelines?12:33
tilmanfirst explain how things work in general ("Creating Packages"), then talk about the adjustment possibilities w/ the config file12:33
jaegerthat seems reasonable12:33
sepencreating packages = 4 and whould be 4.1 or 4.x??12:35
tilmanit probably makes sense to make it a subsection12:35
sepenwhat about remove 'tabs' from the Pkgfile example in this section?12:36
sepenseems the header of this Pkgfile have some tabs12:37
sepeni.e # URL:         http://www.gnu.org/software/grep/grep.html12:37
tilmanyou mean the alignment12:37
sepenyeah12:37
tilmanwho cares ;)12:37
sepenhmm I need a dictionary12:37
sepenx)12:37
jaegerI'll go ahead and put the text in and we can move it if needed12:38
tilmanexcellent12:38
tilmanmaybe we should add a manpage for pkgmk.conf, too12:38
jaegerI think so12:39
tilmanit's not rocket science, but it would be good to have some docs easily available12:39
jaegersafe for me to add a bit of text that says "refer to the pkgmk.conf man page for more information" ? :)12:39
jaegerI plan to detail a few of the settings but not all12:40
sepen+1 for pkgmk.conf.512:40
tilmanjaeger: i think so. the handbook doesn't explain the other commands in detail either12:40
jaegerok12:43
jaegerI'm going to list CFLAGS, CXXFLAGS, PKGMK_SOURCE_MIRRORS, and PKGMK_{SOURCE,PACKAGE,WORK}_DIR and refer to the man page for the rest12:44
jjpkListing the basics should be enough imo.12:45
jaegeraye12:45
jjpkAlthough full documentation is also nice. Regardless, the more important variables should be first in line.12:45
jaegerhttp://crux.nu/Main/Handbook2-4-Package - added "Adjusting/Configuring the Package Build Process"12:53
jaegertweaked the text a bit on PKGMK_SOURCE_DIR12:54
jaegerhrmm... I need to figure out how to make it NOT linkify that example link12:55
jaegerok, that's fixed12:56
*** Man0l0 has joined #crux-devel13:01
tilmanjaeger: looks good13:06
tilmanjaeger: i wonder -- should we mention MAKEFLAGS there?13:09
jaegermy first inclination would be to leave that for the man page but I can add something if needed13:09
*** roliveira has joined #crux-devel13:18
clbUpdate from opt: 6 Dec 19:19 - firefox-flash-plugin: updated to version 9.0.115.013:21
Romsterjaeger, xerces-c gives a 40413:50
sepenoooh13:52
Romsteroh sorry tahts your port sepen13:52
Romsterthat*13:53
Romsterthats*13:53
sepenyeah, I'll fix13:53
Romstergah my crappy typing.13:53
sepenthanks Romster13:53
Romsternp13:53
Romsteri need ti to make kde4 >>13:53
Romsterit*13:53
sepenwell give some time please13:53
Romstersince the kde maintainer seems to not care to update anymore.13:54
Romsterand i know kde4 is in release candidate but i may aswell build and test.13:54
Romsteri might just clone it and throw it in my kde directory13:55
Romsterhave all the sources in the one directory not sure yet..13:55
Romstersepen, also could you edit the port so the version uses . and not _ see my boost port for how i did it.13:56
Romster_ in a version is dumb if you ask me, and i know it's not your fault.13:57
sepenI'll update to 2.8.014:00
sepenneed some time14:00
Romsterwhat a funky runConfigure line..14:01
Romsterno rush i'm messing with the port myself. and keeping it in my kde directory.14:01
Romsterbut if you find soemthing i'll edit this more.14:01
sepenok14:02
Romsteri might even see if i can take the kde repo if the other maintainer dosn't want it if i like kde414:02
sepen@seen alancio14:02
clbsepen: I have not seen alancio.14:02
*** Romster has quit IRC14:05
*** roliveira has quit IRC14:12
*** Romster has joined #crux-devel14:18
Romsteroh i hate power interuptions <<14:21
Romstersepen, i heard your working on a usb live version of crux?14:25
sepenyeah14:26
sepenbased on the released iso cd14:26
sepenusbdisks are great, we can store on it opossite to livecd where we can't14:27
jaegertilman: what path are we using for the source ISOs in the rsync repos?14:28
jaegercrux/crux-$version/sources/ ?14:32
tilmansounds okay14:35
sepenRomster, xerces-c up to date14:36
Romstersepen, k can you add -j1 to make too14:36
Romsteri find it spits out errors otherwise.14:37
Romsterand i compiled it here ok too but i'll take a look14:37
jaegertilman: ok14:37
sepenRomster, surely14:38
sependoing it, also I must delete a read line I fault14:41
Romsterwhat is the 'read'14:44
Romsteri was about to ask..14:44
Romsterand you don't need to use curly brackets on all the vars..14:44
Romsterbut that's upto you. and whats '-minmem' do?14:44
Romsteris that the size of the program binary or the amount of memory it'll use when ran?14:45
sepenRomster, Im fixing it, also Im doing something to eat14:46
Romsterk14:46
*** jaeger has quit IRC16:14
*** sepen has quit IRC16:35
*** Man0l0 has quit IRC17:00
*** DarkNekros has quit IRC17:37
*** jjpk has quit IRC17:51
*** roliveira has joined #crux-devel17:52
*** j^2 has quit IRC21:34
*** j^2 has joined #crux-devel21:45
*** jaeger has joined #crux-devel22:01
*** j^2 has quit IRC22:27
*** j^2 has joined #crux-devel22:30
*** j^2 has quit IRC22:53
*** j^2 has joined #crux-devel22:55
*** j^2 has quit IRC23:00
*** treach has joined #crux-devel23:01
*** j^2 has joined #crux-devel23:01
*** j^2 has quit IRC23:10
*** j^2 has joined #crux-devel23:13

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