[12:01] ogra: do you think you could do the xpdf-utils/poppler thing? [12:01] mdz, elmo: if kdpf is the last package holding xpdf, then it's worth the trouble of converting kpdf IMHO [12:01] pitti: it is [12:01] ogra: it did. but that was before cxx. anyway I'll try again now :-) [12:01] mdz, which xpdf-utils/poppler thing ? i didnt follow [12:01] So I was thinking. dhclient.conf should be configured to send his own hostname to the dhcp server by default. [12:01] jnc: when do you get that message? [12:01] ogra: creating a poppler-utils package which contains the xpdf-utils tools ported to libpoppler [12:02] wasabi: ++ :-) [12:02] ogra: I already did pdftops, but we also need pdfinfo, and we should port the other tools, too (and send them back to upstream) [12:02] pitti, what involves this "porting" ? [12:02] C ? [12:02] dholbach: i got that when i tried to start serpentine. after rm -fr ~/.serpentine; it works again [12:02] C++ for poppler, i think :) [12:02] ogra: take .c, rip it out of xpdf-utils, and hack on it until it compiles with libpoppler-dev :-) [12:02] ergh [12:03] proting C to C++ ? [12:03] elmo: if this works out and we can move it back, will it require a special elmo dance again? [12:03] ogra: they are both C++ [12:03] argl... [12:03] jnc: i don't get that funky message [12:03] I guess that would be properly done by dhclient.conf having a send-default-host-name option [12:03] wtf. it's burning the CD now === jnc blinks [12:03] ogra: false alarm maybe [12:04] i didn't change anything [12:04] oooooh [12:04] g-v-m [12:04] yep [12:04] ogra: so ... that's a bug eh? [12:04] ogra: if it's not something you feel confident about, no problem [12:04] jnc, in g-v-m [12:04] we ought to ask Riddell [12:04] pitti: can you talk to Riddell about it? your time zones are compatible [12:05] oh. i would have thought a bug in serpentine [12:05] mdz: yes, sorry [12:05] mdz, do i have time to ecide until tomorrow ? [12:05] i mean, GnomeBaker works fine with g-v-m running [12:05] decide even [12:05] night everybody [12:05] jnc, might be, but it doesnt use libnautilus-burn [12:05] ogra: it's a small thing and I just want to get it done so that we can be rid of xpdf again [12:05] ah [12:06] bye fabbione [12:06] ogra: maybe we can do some cow trading? I do the poppler stuff (I already know it a bit now) and you help me with the postgresql stuff? :-) [12:06] mdz, i know, but i have to decide if i actually want to learn C++ [12:06] mdz: sure, will do [12:06] ogra: if you would have to learn C++ in order to do it, it's easier for someone else to look at it, don't worry about it [12:06] Riddell should be able to do it [12:07] mdz, i'm ok with C and some OO langs so it should be a good task to learn it :) [12:07] mdz: creating a minimal poppler-utils with just pdftops and pdfinto is a matter of ~ 1.5 hours [12:07] pitti: yes, but you have too much work already, remember? ;-) [12:07] ogra: ok, let's discuss this tomorrow with Riddell [12:07] yep [12:07] mdz: -> that's why I proposed the cowtrading :-) [12:07] what a delightful package name: libxml-libxml-perl [12:07] :) [12:08] let's talk tomorrow [12:08] pitti: i think ogra already said yes :) [12:08] pitti: or did you invite somebody else to the pdf party? :) [12:08] bah, gtimelog doesn't really work well if you cross midnight... [12:08] mdz, vunz made clear he cant do the xscreensaver changes in breezy time, thats a huge bunch of work i have extra now....so i'm a bit careful about extra work currenly.... [12:09] dholbach: oh, my irc output is a bit shuffled here, 10 seconds latency (small net problem) [12:09] *nod* ok [12:09] ogra: fine for me if you want to do it :-) [12:09] ok [12:09] mvo: huh, why python2.4-apt? [12:10] Kamion: bugreporter-udeb was intentionally retired, right? [12:10] hmm, nm [12:11] elmo: it's build now for both python2.3 and python2.4 [12:12] good night guys [12:12] pitti: good night :-). see you tomorrow. [12:12] pitti: night [12:13] mdz: yes [12:13] mdz: replaced by installation-report and save-logs [12:14] jnc: what did you do to be able to burn? now I got that bug instead :-P [12:15] good night everyone :) [12:16] night dholbach :-) [12:21] jnc: nm. works. [12:23] err, KDE doesn't require gnupg2 anymore? [12:23] we went through all of that business for hoary just to have it go away? [12:23] kde deps seem to be in massive flux [12:24] leaving it alone for now [12:24] Nafallo, ?? [12:24] I did a bunch of obvious stuff [12:24] Nafallo, what was it ? [12:24] trying to get a handle on the set of packages which pitti will need to review in order to sync up [12:25] ogra: seems I had to kill g-v-m. [12:26] Nafallo, :/ [12:26] ogra: that shouldn't be the correct approach though ;-) [12:26] nope [12:26] elmo: how do you calculate the reverse-depends stuff in anastacia? [12:26] but i was suspecting its g-v-m === LeeJunFan [junfan@s64-186-37-120.skycon.net] has joined #ubuntu-devel [12:27] Heh. [12:28] If you select N/A in the Ubuntu device database creation thing for sound, the next page says "Video" "did you hear the text sound?" and offers no forward button or controls of any type. [12:29] ogra: ehm... will be nice to see if this disk works. the bar didn't move since it started writing in record speed. just fixated. [12:29] wasabi_: file bug :) [12:29] Nafallo, tell me if it worked then.... [12:29] ogra: works. [12:29] great... [12:29] anothe bug to fix [12:29] ogra: but an ugly bug ;-) [12:30] mdz: it's just looking at germinate 'all' output [12:31] elmo: ah, that explains why it's lying then [12:32] lying where? [12:33] ogra: could you fix elementtree to stop building python2.2 packages? it's trying to pull python2.2 into main [12:33] Kamion: in anastacia output [12:33] Kamion: it says it wants dialog because of alsa-utils, but it's an OR'd dep with whiptail, so it's actually something else === mvo goes to bed now [12:33] ah, ORed deps occasionally confuse germinate yes [12:34] Kamion: I think it's probably doing the right thing, just choosing an unintuitive package to list as the reason [12:34] mdz, oops, how could that get past the python transition ? [12:34] mdz: hum, that could be my fault [12:34] I had hints for germinate, and recently updated to 2005 germinate and may have lost the hints, lemme check [12:35] dialog | dialog | alsa-utils | Santiago Vila | 180578 | 1048 [12:35] hmm, nope [12:35] rdepends is unintelligible for the same reason [12:35] mdz, working on it [12:35] a ton of stuff depends on dialog | whiptail [12:35] ogra: thanks [12:37] mdz: reason selection is often dodgy; I could probably try to make it pick the first non-ORed dep it encounters [12:38] kamion; the problem is, given a or b, if b is already in main, it should favour b [12:38] and currently doesn't AFAICS [12:38] Kamion: I don't think it's worth it; doing the right thing seems Very Hard, and picking the first non-ORed dep doesn't sound much better than the status quo [12:38] germinate has no idea what's already in main [12:39] Kamion: eh, "is already in it's idea of main" (i.e. 'all') [12:39] especially not when running in the archive configuration where it just gets a single Packages file [12:39] Kamion: I would like to know what's going on in this particular case, though. it's starting to look like germinate is doing the wrong thing, actually [12:40] there are only 15 packages which depend on dialog-but-not-whiptail, and they're all in universe [12:40] elmo: it's encountering dialog | whiptail before it's had a reason to explicitly select whiptail [12:40] kamion: oh, right === elmo goes back to sleep [12:40] seeding whiptail would fix that [12:40] Kamion: does it matter which seeD? [12:41] does it have to be minimal (where the dep on alsa-utils is coming from), or can it be supported? [12:41] I think supported would do, but let me do some experiments to test that [12:41] germinate's begging for a test suite ;) [12:41] [he says helpfully] [12:42] germinate will spot that it can promote whiptail from an outer seed to satisfy the dep === Kamion introduces elmo to the Ubuntu seeds [12:43] in the same way the ubuntu archive is the test suite for katie? good call :P [12:43] it's like WORKSFORME, only WORKSFORYOU [12:44] I have lots of directories scattered around in ~/src/ubuntu/germinate/ called 'pre' and 'post' [12:48] +? Unknown dependency dialog by defoma [12:48] +? Nothing to choose to satisfy defoma [12:48] hmm, it doesn't entirely like that [12:48] I'll dig [12:54] duh, it helps to run against main,restricted,universe,multiverse not just main,restricted. spethial [12:56] kamion, the CoC suggests you should be sensitive to retards even when referring to yourself [12:57] sabdfl: nice -) [01:00] ah. he's ignoring trolls. sigh. [01:00] sabdfl: heh :-) [01:01] nah, I was just hunting down the libstdc++ porting howto for somebody on -users [01:01] what's he porting too? [01:01] from libstdc++2.10 (pre-standard) to anything vaguely current [01:01] btw, a mate of mine is trying to install hoary on a mini-mac, do we have any success stories on that? [01:02] well, pre-current-standard anyway [01:02] sabdfl: hmm, not that I know of, I think full support may need > 2.6.10 since it's pretty recent [01:02] his name's richard braine (really) and i pointed him in your direction [01:02] the "buy benh a mini-mac" campaign on debian-powerpc was after we froze, I think [01:02] does he have a friend called pinky? [01:02] that would be me [01:03] on bald days [01:03] sabdfl: yeah, I saw him on #-meeting, but he left before I got back; I'll mail him [01:04] cool, thanks [01:04] apparently it installed to console by xorg's not working, so it's not a totaly disaster [01:04] i'm surprised google has no fixes [01:06] I have a remark about http://udu.wiki.ubuntu.com/MountingHDDFilesystems [01:06] once fstab has the right entries, people can use the gnome disk mounting panel applet [01:07] it's easy to use but doesn't require auto-mounting of all NTFS/FAT partitions at boot time === diamond_ [~diamond@194.46.73.169] has joined #ubuntu-devel [01:09] oops, the snake got out, gave me a surprise :) === Kamion restores it to the tank === diamond_ hopes there is no metaphor involved here [01:10] JanC: you suggest making them all noauto by default; could do [01:10] diamond_: no :) [01:10] Kamion: that's ok then -) [01:10] Kamion: what kind of snake, if it's not to ot to ask? [01:10] amelanistic (i.e. orange) corn snake === pedro_ [~pedro@pc-200-74-115-220.megavia.pc.metropolis-inter.com] has joined #ubuntu-devel === diamond_ updates the kamion stalking wiki === nufan [~nufan@cpc2-hem14-4-0-cust73.lutn.cable.ntl.com] has joined #ubuntu-devel [01:11] diamond_: oooh! oooh! where is it? can I stalk Kamion too? [01:11] it's my fiancee's snake rather than mine [01:12] maswan: actually, i screwed up, i have to now officially deny it's presence here, but if you come around the back, i can let you know the dirty details [01:12] there is no kamion stalking wiki. [01:12] Heh, good, good. Fixing problems with the video output for now, looks [01:12] like DDC doesn't work with X, and not very well with radeonfb, because [01:12] of the stupid way Apple routed the DDC lines of the connector to the [01:12] "CRT2" DDC lines of the video chip... oh well... === maswan heads around back to beg acceptance from the non-existing Kamion staking cabal [01:13] sabdfl: ^-- benh having just got his mac mini in early March [01:13] although he said screwing the DVI cable all the way in helped ... [01:13] maswan: -) === nufan [~nufan@cpc2-hem14-4-0-cust73.lutn.cable.ntl.com] has left #ubuntu-devel [] [01:17] JanC: want to add that to a comments section at the end of that page? [01:19] this wiki uses different accounts than the normal ubuntu site I guess ? === astharot- [~isager@host71-134.pool8258.interbusiness.it] has joined #ubuntu-devel === Arrogance [~aks@CPE0050ba556e4b-CM001225423850.cpe.net.cable.rogers.com] has joined #ubuntu-devel [01:20] not sure, try your existing one first === astharot-- [~isager@host17-99.pool8252.interbusiness.it] has joined #ubuntu-devel [01:22] i've cut some things off the FrontPage of the wiki and put them in second level pages [01:22] just in case anyone misses something [01:22] back again ... [01:31] Kamion,mdz,jbailey: the reason of the buildability of OOo2 in hoary (and not in breezy) is the addition of gjdoc as a java-gcj-compat dependency. that one sucks in antlr, and then kaffe and sablevm [01:32] hi [01:32] I think antlr is fixable. In fact I had thought I fixed it. [01:33] Build-Depends: debhelper (>= 4), kaffe, fastjar, classpath, jikes-classpath (>= [01:33] 2:0.12-1), autotools-dev, gcj-3.4, libgcj5-dev [01:33] seems not. [01:33] Oh yeah. I remember now. I didn't fix it. [01:33] jbailey was going to work some cdbs magic on it. [01:33] ;) [01:34] Because it was a weird makefile build system. [01:34] jbailey! [01:34] wasabi_: Is that the only blocker? [01:34] mdke: thanks for the link to Kubuntu :) === wasabi_ shrugs. doko found it. [01:35] I remember skipping over it ages ago because it was an odd build system. [01:35] And I wasn't feeling up to it at the time. === diamond__ [~diamond@194.46.73.169] has joined #ubuntu-devel [01:36] Doesn't look like that big of a deal now that I look at it again. I probably learned some stuff about Make since then. [01:36] Riddell, fundamental i feel [01:36] mako, ping? === CarlFK [~CarlK@c-67-163-11-11.hsd1.il.comcast.net] has joined #ubuntu-devel [01:39] * Chose dialog to satisfy defoma [01:39] *good* crack [01:41] Think I've got it fixed. [01:42] Heh. Yay for circles. [01:43] is someone porting apt-cacher to apache2 or should I start working on that after I've been in bed for several hours? :-) [01:44] seems it might be fun to have for NetworkWideUpdates [01:46] doko, you aren't working on antlr, right? :0 [01:46] doko, I've got it fixed. === mkde [~matt@mdke.user] has joined #ubuntu-devel === doko [~doko___@dsl-082-082-189-098.arcor-ip.net] has joined #ubuntu-devel === OddAbe19 [~OddAbe19@68.82.202.24] has joined #ubuntu-devel === svenl_ [~luther@AStrasbourg-251-1-63-82.w82-126.abo.wanadoo.fr] has joined #ubuntu-devel [01:58] Kamion: very cool. surprised apple capable of braindamage like that. [01:59] hey guys [01:59] mako just found this site www.ubuntu.it, it is a redirect of the main ubuntu.com site but with a pubblicity banner at the top for a company doing linux training [01:59] crazy huh [02:01] oh wow time for me to get flamebasted on osnews [02:01] "mono 1.1 added to ubuntu" [02:01] news flash. [02:01] awesome [02:10] this poor soul will be using gnome 2.10 for the next 5 years since I am "adding" mono to ubuntu [02:10] im so mean. [02:12] hehe === blueyed [~daniel@iD4CC0360.versanet.de] has joined #ubuntu-devel === bradb [~bradb@modemcable141.58-131-66.mc.videotron.ca] has joined #ubuntu-devel [02:34] jbailey: do you have an archive for initramfs-tools? I'm pretty much guaranteed to have to make changes to it === shaya [~spotter@dyn-160-39-242-154.dyn.columbia.edu] has joined #ubuntu-devel === tim [~tim@cpe-66-67-139-238.rochester.res.rr.com] has joined #ubuntu-devel [02:44] does anyone know if there is are anymore ubuntu offshoots in development and where I can find info on them (like kubuntu) especially one for e17-ubuntu :P [02:46] jbailey: at a minimum I need a hook facility so that I can get some additional scripts run at boot time [02:55] morning boys and girls, and elmo [02:55] good morning jdub! [02:56] jdub: morning :-) [02:56] tim: kubuntu is the biggest, but guadalinux and gnoppix are ubuntu based as well [02:56] tim: im not sure there is a list. [02:59] tseng, thanks :) [02:59] mdz: I don't yet, although I've promised one to the guy who'll be doing the Debian merges. [03:00] jbailey: who's that, out of curiosity? [03:01] maks. He's expressed interested in the initramfs work and has been doing some bug fixing on the Debian initrd-tools stuff. [03:01] Not yet a DD. [03:03] Hmm, no debs for bazaar-ng. === jsgotangco [~jsg@202.57.71.235] has joined #ubuntu-devel [03:04] hi jsgotangco [03:04] morning! === Firetech [~Firetech@h53n1fls301o1100.telia.com] has joined #ubuntu-devel === Kamion [~cjwatson@host81-153-126-219.range81-153.btcentralplus.com] has joined #ubuntu-devel === stub [~stub@203-217-37-199.dyn.iinet.net.au] has joined #ubuntu-devel [03:12] mdz: I'm going to go out for a walk in amoment - do you have any questions about the setup before I go> [03:12] jbailey: nah, I'm going to have to do the first iteration by hand anyway [03:16] Cool. /me goes for a walk. === AndyFitz [~andy@wlanconf-nat-pool-brisbane.redhat.com] has joined #ubuntu-devel === shaya [~spotter@dyn-160-39-250-69.dyn.columbia.edu] has joined #ubuntu-devel === tim [~tim@cpe-66-67-139-238.rochester.res.rr.com] has left #ubuntu-devel ["Leaving"] === AndyFitz [~andy@wlanconf-nat-pool-brisbane.redhat.com] has left #ubuntu-devel [] [03:37] jbailey: so nfs root is working for you, right? having some difficulty here [03:37] "mount: No such device" "short read: 0 < 28" [03:37] after successful ipconfig [03:38] ah, need to specify the nfs module explicitly === timelyx [~timeless@c-67-180-65-89.hsd1.ca.comcast.net] has joined #ubuntu-devel === ikuyaLoqu [~ikuya@gnulinux.good-day.net] has joined #ubuntu-devel === mark_al [~markalit@82-32-64-135.cable.ubr01.bath.blueyonder.co.uk] has joined #ubuntu-devel [04:03] hi, can anyone tell me how to capture lots of useful info to add to a bug report? its for a problem with a kernel module (snd-usb-audio) which worked fine in warty [04:03] Man. THe Java stack that is going to come along with Ooo2 is really going to screw up the idea of putting it all on one cd [04:03] Hmm, or maybe not. [04:04] gcj is awefully tiny. [04:10] wasabi, perhaps change it to "all in one dvd" ;) [04:12] and all in two CDs. [04:15] the one-CD thing is really non-negotiable [04:16] if OOo2 doesn't fit, we get to (a) trim it down (b) throw out other stuff (c) bin it :-) [04:17] The one CD thing has many benefits. [04:17] It's one of my fav features heh [04:19] mark_al: it really depends on the bug [04:19] mark_al: try in #ubuntu [04:21] cheers [04:21] jbailey: the mount in klibc is giving me headaches [04:21] it doesn't seem to pass all of the options correctly [04:22] mdz, i have uploaded a fixed antlr. it now builds with gcj-4 and friends. Problem is they are all in universe... so now antlr is. [04:22] The entire stack needs to jump over. [04:22] wasabi: gcj-4 is in main [04:23] java-gcj-compat isn't, then. [04:23] that'll be.... [04:23] ecj-bootstrap and java-gcj-compat. [04:25] wasabi: I'll take a look at it once everything is built, so we can see the dependency chains [04:34] how do i regenerate a makefile.in after changing makefile.am? [04:36] Unfrgiven: autoreconf [04:37] mdz: nfsroot is working, yes. I tested both regular boot and nfs boot today. [04:37] jbailey: yeah, got that working, now only mount is causing me trouble [04:37] it mounts, but apparently not all of the options make it to the kernel [04:37] doing a similar mount later, with util-linux mount, does the right thing [04:37] Ugh. =( === wasabi wondering what new super-cool features are on the horizon if ya'll are messing with nfsroot [04:37] jbailey: if I set BUSYBOX=y, will that use busybox instead of uclibc's tools? [04:37] Could easily be a klibc bug then. I haven't explored all of that. [04:37] wasabi: ThinClientIntegration, for one [04:38] mdz: No, the only thing it does is call busybox if you drop to a shell (add break to the kernel command line) [04:38] I didn't want behaviour to change just because you wanted a sane shell for debugging. [04:38] er, s/uclibc/klibc/ of course [04:39] jbailey: anyway, other than that hangup, I've got thin clients booting [04:39] to a shell [04:39] by hacking up /usr/share/initramfs-tools/scripts/nfs [04:39] what I really want is to let it mount an NFS root fs as normal, and then run a hook later which can set up the unionfs [04:39] Cool. Will you email me your nfs stuff? I'll hack in a script. I tried an import earlier and wound up filing a bug on baz. [04:40] jbailey: the hacking up was entirely thin-client specific except for two things [04:40] 1. mkdir /tmp [04:40] 2. . /tmp/net-${DEVICE}.conf [04:40] but I'll send you a copy so that you can get an idea of my needs [04:41] Should the unionfs be set before init is called? [04:41] jbailey: unless we want to go back to using pivot_root, yes [04:41] 'k [04:42] init needs to run within the unionfs root [04:42] Okay. I wasn't sure, since many filesystems mount readonly anyway. [04:42] root filesystems, rather. [04:43] well, the unionfs is actually a separate filesystem, so we need to run inside it [04:44] Okay. [04:44] Will the nfs script you send include the options you need passed in? I can check the klibc nfsmount code at the same time. [04:44] jbailey: what's the trick to running the stuff in /usr/lib/klibc/bin? [04:44] jbailey: yes, it does === bob2___ [~rob@crumbs.ertius.org] has joined #ubuntu-devel [04:45] lamont: debconf still isn't building [04:45] admin/debconf_1.4.50ubuntu1: Dep-Wait by buildd+vernadsky [important:out-of-date] [04:45] I was just going to test that hypothesis on a real system using klibc mount, but it doesn't run [04:45] Dependencies: libqt-perl (>= libqt-perl_3.008-1.3build1) [04:45] lamont: that version looks bogus [04:45] oh, it expects to find a klibc loader in /lib [04:46] mdz: They're not real ELF executables, that's why I didn't stuff them in the path. It needs the klibc library in /lib [04:47] jbailey: is there a trick I can use to get them to run, or do I need to mess with /lib? [04:47] mdz: If you copy the library into /lib they should run. === mdz twitches at the cruft [04:47] mdz: I can include a symlink in the next upload (or just move the library) if that's worthwhile. [04:48] jbailey: is it possible to run them from the build tree? [04:48] I'm sure it's some silly C string handling bug or something [04:49] in particular, this doesn't look right: [04:49] The last time I did it was before they had dynamic linking working, I don't know off hand. [04:49] if (extra->used_size) [04:49] *extra->end = ','; [04:49] strcpy(extra->end, s); [04:49] surely that should increment ->end rather than clobbering it? === camilotelles [~Camilo@200.217.77.77] has joined #ubuntu-devel [04:52] I think this code was written and never tested [04:52] since a typical initramfs setup will never exercise it === thundrcleeze [~thundrcle@c24.183.106.5.mad.wi.charter.com] has joined #ubuntu-devel [04:54] Entirely possible. [04:55] I don't know how much of it was hacked out of the kernel or where it came from originally. [04:55] I know that it's targetted to be shipped with the kernel, though, so that seems likely. === manulito [~joel@v10-23.vildanden.afb.lu.se] has joined #ubuntu-devel [04:56] Hi, im a 25 year old cs-student from sweden, have some short questions about how to join a team, anyone have a few minutes to spare ? [04:57] just ask your question [04:57] (s) [04:57] im planing to join some opensource devel-team after the summer, just wanna know what skills are needed and what i should improve on during the summer [04:58] have some extra time to read/learn, and wanna get some pointers on what skills are important :> [04:58] there is some sketchy stuff in there [04:59] if (newlen >= extra->alloc_size) { [04:59] char *new; [04:59] new = realloc(extra->str, newlen + 1); /* +1 for NUL */ [04:59] if (!new) [04:59] return; [04:59] extra->end += new - extra->str; [04:59] "new - extra->str" where "new" is the return value of realloc and extra->str is some other pointer? [05:00] manulito: depends on what you want to work on [05:00] oh, creative [05:00] extra->str was the original pointer [05:00] manulito: it depends on what sort of projects you'd like to work on. C programming and Python programming are both useful [05:00] people generally work on things they care about [05:00] so it's the delta of the length (newlen - oldlen) [05:00] how do you figure? [05:00] realloc is allowed to return you a pointer someplace far away in memory if it feels like it [05:01] that may work if it never ever moves [05:01] wow [05:01] mdz: what are you looking at/ [05:01] dilinger: klibc [05:02] cute [05:02] doesn't extra->str already have the +1 for \0? [05:02] it makes me want to die [05:02] and why can't extra->end += newlen - extra->alloc_size? [05:03] and yes, if realloc moves it.. extra->end is no longer valid, i would gueess [05:03] dilinger: don't ask me; look at the previous snippet I pasted too [05:03] mdz: hey, I'm just saying what the code is supposed to do, not what it necessarily actually does ;) [05:04] daniels: I bet it does exactly that 99% of the time [05:04] bob2: last 3 years in school have been different types of courses in programming etc, tho i never written anything real, i have no clue were to start, how to take the step from schoolproject -> real project [05:04] i assume the previous snippet just has a typo. strcpy(++extra->end, s); [05:04] and then randomly falls over if it grows beyond some unpredictable length [05:04] the purpose would be to simply add a comma if there's something already there (extra->used_size != 0) [05:04] mdz: The klibc realloc code might be dumb enough to just fail if it can't expand the code segment. [05:05] dilinger: I understand what it intends to do [05:05] but it's buggy [05:05] jbailey: nothing like writing code that depends on such a feature :) [05:05] in more than one place [05:05] Frightening behaviour to depend on, though. [05:05] somebody remind me why we need yet another libc implementation? [05:06] mdz: combination of size (not glibc), arch support (dietlibc) and licence (uclibc) IIRC. [05:07] I'm not entirely convinced by the size argument [05:07] yeah, glibc is huge [05:07] but the stuff we're putting on top of it is *tiny* [05:08] We don't have the very-small requirement, but this stuff is intended to ship with the kernel as a standard set of utilities to run for early userspace. [05:08] mdz: so improving my c knowledge is prehaps best way to start then? [05:08] manulito: yes, in particular you could help me debug klibc's mount option parser :-) [05:08] For an embedded system, this would be beautifully small. === netdur [~adel@adsl-84-125-192-81.adsl.iam.net.ma] has joined #ubuntu-devel [05:09] mdz: that must be fun [05:09] mdz: might help you out in half a year or so :> === mdz [~mdz@ca-studio-bsr1o-251.vnnyca.adelphia.net] has left #ubuntu-devel ["Client] === mdz [~mdz@ca-studio-bsr1o-251.vnnyca.adelphia.net] has joined #ubuntu-devel [05:17] this code could not possibly have ever worked [05:17] anybody saw this? google is giving US$ 4.500,00 for studentes who wants to help open source [05:18] camilotelles: yes, we're one of their open source partners [05:18] http://code.google.com/summerofcode.html [05:18] camilotelles: look at the list of participating organizations [05:18] mdz: I saw this. Maybe it's interisting to put in the bounties page. [05:19] camilotelles: that's backwards; Google links to the bounties page [05:19] mdz: ok. [05:20] i swear, latest breezy update remapped all of my control keys in some way I have not yet figured out [05:21] wasabi: x in breezy is currently 'interesting' [05:22] exciting. [05:22] daniels: speaking of which, what's the plan wrt /usr/include/X11 vs. /usr/X11R6/include? [05:22] there seems to be stuff in both places at the moment [05:22] mdz: right. everything will eventually move to the former. [05:22] by 'eventually', I mean 'asap' [05:23] I had something ftbfs because it didn't -I/usr/X11R6/include; not sure if that's a common failure or not [05:23] it's reasonably common, yeah [05:24]