[00:03] lifeless: practice makes perfect [00:17] \sh_away: just confirmed gaim is not affected by the pidgin exploit [00:18] doko: Why was icedtea bootstrapped with gcj, btw? Because it was formerly bootstrapped with a non-free implementation? === jamesh_ is now known as jamesh [00:39] slangasek: Just for fun, I decided to see if I could figure out why the heck we log_success_msg the usage in the samba init scripts. All I've found out is that it's been that way since warty. [00:40] soren: I think we can chalk it up to lamont's bad influence [00:40] soren: if it's been that way since warty, then it's probably crack. [00:40] lamont: Clearly :) [00:40] slangasek: it was a gentoo guy... [00:41] he didn't work for canonical for very long.... [00:41] There's a changelog entry from a Nathaniel McCallum that says that he prettified the init script. [00:41] lamont: Him? [00:41] oh, really? :) [00:41] just long enough to "fix" all of desktop to do lsb-ish inits. nevermind about that -e flag at the top and checking return values and such... [00:42] hahaha [00:42] yeah, I'm still bitter about having had to drop the -e from the samba init scripts :) [00:42] Is the name supposed to ring a bell? [00:42] you know... change it to say start-stop-daemon .... ; lsb_ $? in a script that says #!/bin/sh -e .... [00:43] * lamont doesn't discuss names [00:43] lamont: changelog does that for you [00:43] lifeless: I will nether refute nor confirm that. :-) [00:44] "never speak ill of the braindead"? [00:45] "what's an soname?" [00:45] Ah. Him. [00:46] Haha! [00:46] "I mean, if you just recompile everything, then you don't need multiple copies of the library" [00:46] * lamont wonders if it would violate the CoC to say that some distros foster gentoo-esque thinking [00:47] * soren goes to bed [00:47] Good night. [00:47] 'night [00:47] Amaranth: yes. the fix is to make the window menu work [00:47] lamont: As long as you don't explicitly define gentoo-esque, I think you're safe. [00:47] Amaranth: imo :) [00:48] good night soren. [00:48] Amaranth: in that all of the functions of the window menu are still entirely applicable to an undecorated window [00:48] ScottK: heh [00:49] ScottK: context can be a real bitch, you know. [00:49] Whatever people choose to infer from that is not your responsibility. === rob1 is now known as rob [01:04] * Hobbsee waves [01:06] * lamont waves Hobbsee [01:06] heya lamont! [01:26] hey uh, who put thinkfinger in from my ppa? [01:26] ? [01:27] pwnguin: it's not in the archive, unless it's in source new or something. [01:30] i just got an email [01:30] wasn't my fault! I swear! ;-) [01:31] pwnguin: it's probably something bothced with LP [01:31] LP likes sending random email [01:32] of course, ti doesn't help that people seem to like assigning bugs to groups with 100+ people in them [01:32] Yup. There it is, sitting in NEW. [01:33] I think you should not be able to assign [01:33] only accept [01:33] well, i filed a sync request [01:33] pitti seconded it [01:33] i thought it was gonna come from debian, not my ppa [01:33] pwnguin: so you got the accepted mail, for the sync. [01:33] arr, not funny [01:34] cjwatson: debian-installer has failed to build with mklibs errors, am hunting now [01:34] pwnguin: http://launchpadlibrarian.net/10626425/thinkfinger_0.3-2_source.changes - isn't that from experimental? [01:34] Hobbsee: im not sure, but the .dsc has my name on it [01:34] pwnguin: because of maintainer mangling [01:34] see the spec. [01:34] ok [01:35] it is from experimental. [01:38] oh, heh [01:38] the next email in line has the log from the sync ;) [01:57] lamont: http://launchpadlibrarian.net/10554595/buildlog_ubuntu-hardy-hppa.libgpod_0.6.0-2_CHROOTWAIT.txt.gz if you've not seen it [01:58] lamont: Specifically, Inconsistency detected by ld.so: dl-fini.c: 180: _dl_fini: Assertion `ns != 0 || i == nloaded' failed! [01:59] FTW. [01:59] Heh [01:59] It's PA-RISC hardware wonderful? :-) [01:59] Er, Isn't [02:00] doko: what did the new gcc-4.2 do to hppa???? [02:01] what is hppa useful for? [02:02] LaserJock: it's useful for keeping lamont happy [02:02] useful indeed [02:02] its useful for finding bugs [02:02] StevenK: on the bright side, libgpod is the first such failure in a while. so dunno... maybe I'll just give it back and see how it does.. [02:02] because its an anal platform [02:02] lifeless: and big endian [02:03] and can be told to throw a fit at unaligned loads/stores [02:03] big endian, weird alignment rules [02:03] if only it had stayed slow [02:03] ;) [02:04] "This special 2-4 week Online program that will earn you a full non-accredited [02:04] degree based on your work experience and past/present knowledge." [02:04] * lamont rotfl [02:05] cool [02:05] I shoulda tried that instead of 9 years of sweat and tears [02:07] Blood. Don't forget the blood. [02:07] LaserJock: its useful as a doorstop [02:08] ScottK: I tried to avoid the blood [02:08] ScottK: although i did crush a NMR tube and cut my hand once [02:09] Ow. [02:10] well, the bad part is it was a $180 1cm, quartz, flat-bottom NRM tube [02:14] * ScottK figured there'd be blood. What with the lasers and all. [02:14] well [02:14] lasers tend to coagulate blood well [02:15] Don't they also blow up sometimes? [02:15] well ... [02:15] it didn't exactly "blow up" [02:15] LaserJock isn't human anymore, he's a grad student [02:15] argh [02:15] argh to you too [02:15] don't make me feel bad for finally taking a break from writing these pracs up. [02:16] heh [02:16] you deserve a break, don't you? [02:16] * Hobbsee wants to get this stuff done! [02:16] * Hobbsee was bad, procrastinating earlier in teh semester. [02:17] * slangasek gives lamont a full non-accredited degree: ⁰ [02:17] This is how you can tell Hobbsee isn't a grad student. We procrastinate *all through* the semester. [02:17] haha [02:19] slangasek: thank you so much. remind me to by you $BEVERAGE next time we meet. [02:19] RAOF: i've got a deadline of friday, and i said i'd have it done by yesterday. [02:20] haha [02:20] so given how long it's taking bzr to pull down the debian-installer repo, I should probably ask sooner rather than later - is there anyone around willing to be my designated core-dev for this evening? :) [02:20] another non-gradstudent remark [02:20] no really. friday is the solid deadline [02:20] saying you'd have *anything* done before a deadline? [02:21] yes, but why would it be done before the deadline? [02:21] the deadline that they were quoting for it, during semester, was the 8th, iirc. [02:21] (someone to merge changes to https://code.launchpad.net/~ubuntu-core-dev/debian-installer/ubuntu, and to upload the results) [02:21] well, i didn't know the hard deadline until a few hours after when i got the phonecall saying "where's your stuff?" [02:21] Hobbsee: ouch [02:21] you're starting to redeem yourself ;) [02:21] * TheMuso feels for Hobbsee, having been through something similar when he was at uni. [02:22] LaserJock: I'd say Hobbsee is quite qualified to be a grad student, don't you think? [02:22] ajmitch: nah, i don't have the patience [02:22] * RAOF used to be able to hit deadlines. [02:22] * ajmitch hits deadlines & keeps on sailing through [02:25] slangasek: if you need a core-dev-bitch, I suppose I could fill in [02:25] but I'm gonna take back that $BEVERAGE [02:26] that's fair :) [02:26] slangasek: but only until 11PM your time. [02:27] ack [02:29] ajmitch: I just forget deadlines [02:29] "oh, there was a meeting today? huh" [03:18] lamont: Heh, libgpod moves from chroot problem to dep wait [03:18] yep. I was just going to say that [03:28] slangasek: so what are you breaking in debian-installer? [03:29] hopefully something encryption-related [03:31] lamont: I'm planning to break the FTBFS [03:31] but haven't had any luck yet [03:31] oh [03:31] it's a thrice-damned mklibs problem [03:35] ah, we may be in luck and it's just Debian bug #443248 surfacing because newt isn't merged yet [03:35] Debian bug 443248 in newt "Please link with -ldl" [Normal,Fixed] http://bugs.debian.org/443248 [03:37] * StevenK tries some of his wife's "digestive helping tea" [03:37] It even looks insipid === choudesh_ is now known as choudesh [03:38] StevenK: ugh that stuff is all placebo and marketing [03:39] ... [03:39] jdong: if it works, who cares? [03:39] If it helps my stomach get over lunch, I don't care. :-) [03:39] * StevenK high fives lamont [03:39] /usr/bin/ld: final link failed: Nonrepresentable section on output [03:40] * StevenK sighs [03:40] StevenK: in China they use domperidone for that purpose.... I find that slightly unsafe though extremely effective :) [03:40] I'm little bemused that the ingredients list was "Please refer to leaflet" [03:41] StevenK: lol that's to separate it from the "FDA says this is total BS" label they are required to affix. [03:42] not in Australia [03:42] Right [03:43] * StevenK sighs at XUL 1.9 [03:43] ah, right you people aren't from the states [03:43] * lamont got a kick out of update-manager's "this is gonna take forever, and you can't stop and restart it, ok?" question [03:43] jdong: slangasek is from portland. that's not in the US> [03:43] Bwahaha [03:43] Now I get undefined reference to the gtkmozembed functions it's using. [03:43] I was referring to StevenK, but thank you all the same [03:44] StevenK: are you racing slangasek to a solution? [03:44] I'm from that island prison that the English keep [03:44] lamont: I suspect slangasek and I are sighing about different things [03:44] StevenK: awesome, with the kangaroos! I learned about that in 4th grade geography! [03:44] ah, ok [03:44] * StevenK kicks jdong in the head [03:44] yes, gtkmozembed is not used in d-i [03:45] hahahaha [03:45] hrm.. I need a fax/modem board with caller id [03:45] Although, help with debugging linking against gtkmozembed gratefully accepted [03:46] slangasek: good. I didn't want to have to hurt StevenK [03:46] StevenK: which arch? [03:46] amd64 [03:46] StevenK: look for the bits built without -fPIC? [03:47] or the system .a files being linked in [03:47] slangasek: Last time I saw -fPIC problems, it told me. Here I'm just getting undefined references [03:48] you said "Nonrepresentable section on output", that's not an undefined reference [03:48] I see both; http://paste.ubuntu.com/2339/ [03:49] oh, then, find where those functions are defined :) [03:49] * slangasek <-- help with debugging [03:49] * StevenK digs up the patch asac gave him for tinymail [03:50] Maybe applying that to modest will help [03:50] modest mussorgsky? [03:50] * lamont finally dist-upgrades his home hppa box from gutsy-stage0 to gutsy [03:51] slangasek: There's a e-mail client called 'modest' [03:52] StevenK: named after mussorgsky? [03:52] Named because it's designed for modest hardware [03:52] slangasek: How can I determine which library defines those symbols? [03:53] StevenK: with nm -D /usr/lib/library-I-suspect.so? [03:54] I wonder what nm is supposed to expand to. [03:54] (Like ls being list) [03:54] name? [03:55] That seems too obvious. :-) [03:56] Number Muncher [03:56] jdong: That tea has seemed to help, so nyah [03:57] lamont: tada, d-i fixed. I suppose you want a full-fledged merge request rather than me just dumping files somewhere? :) [03:57] slangasek: for you? signed source.changes and files [03:58] Which you'll have to resign [03:58] duh [03:58] Heh, sorry [03:59] StevenK: I just want trackability... [03:59] slangasek: Okay, I have the library that contains those symbols [04:00] StevenK: and it'll have some fixup that's not PIC, and you're building a PIC binary [04:00] apply sledge-hammer love to the library [04:00] The library in question is /usr/lib/xulrunner-1.9b1/libxul.so, it has it's own sledge-hammers [04:00] heh [04:02] StevenK: slangasek signing the changes lets me go 'he should know better' when cjwatson slaps me upside the head for uploading some silliness of slangasek's. :) [04:02] slangasek: unless you think I should review your changes.... [04:03] I'm just waiting for slangasek to upload a d-i to Ubuntu that installs Debian [04:03] * StevenK hides [04:04] slangasek: The final gcc command run contains -I, but does not add -lxul or /usr/lib/xulrunner-1.9b1/libxul.so to the command line [04:05] is it guilty of using gcc to link g++ binaries? [04:06] I don't think so [04:06] I wonder if the hppa box will boot without me typing ctl-D this time [04:07] lamont: It won't. [04:07] lamont: http://people.ubuntu.com/~vorlon/newt_0.52.2-11.1ubuntu1_source.changes [04:08] fsck.ext3: Unable to resolve 'UUID=37b0a8ff-e4d7-40af-b7e7-597001b30bf6' [04:08] fsck died with exit status 8 [04:08] ...fail! [04:08] * File system check failed. [04:08] and now I have the error message. [04:12] * StevenK sprinkles some magic inside configure.ac and tries a rebuild [04:13] lamont: do I still have your attention, or are you lost to the wonders of libuuid and udev? :) [04:13] still fetching [04:13] do I need to test this? [04:13] no [04:14] so, once it gets published and queue-mangler gets done, then the buildd will get to it. [04:15] I assume you can do the retry-a-thon on d-i once it's built/published? [04:15] otherwise, whack me. [04:15] TTBOMK I'll need to whack you [04:15] unless there's a dep-wait capability in soyuz [04:15] checkgrp [04:15] admin: 109 106 [04:15] crontab: 106 107 [04:15] lpadmin: 108 104 [04:15] scanner: 104 105 [04:15] ssh: 107 108 [04:15] root@j6700:~# [04:15] I hate it when ldap and /etc/group disagree [04:16] slangasek: Soyuz does automatic depwait stuff [04:16] StevenK: automatic dep-wait isn't useful for this case [04:17] this is a "we know why it failed to build, we know which version fixes it, admin wants to set a manuall dep-wait on the fixed version" [04:17] slangasek: not that I've seen [04:18] I mean, one could do sql-foo if one had access, but that's not exactly "in launchpad" [04:18] :) [04:18] Sigh, my magic had no effect [04:21] -rwxr-sr-x 1 root crontab 78472 2007-10-04 22:26 /usr/bin/ssh-agent [04:21] see, because that's just wrong [04:23] slangasek: you're right. I'll need to whack it [04:23] remind me when, eh? [04:24] lamont: yep, I'm polling vigorously :-) [04:24] queue-builder is a 30 min runtime, and then it sleeps 1800 before it runs again [04:25] so start polling at 20:44 PST, and you might get results [04:25] slangasek: how come your key doesn't have your ubuntu.com addr on it? [04:26] because /me didn't do --recv-keys recently [04:26] doh [04:27] * slangasek grins [04:27] better question, why does dpkg-buildpackage not acknowledge that it has my ubuntu.com addr on it :P [04:30] Sigh. Nothing I do has any affect. [04:31] StevenK: Move the package out of the globe of invulnerability; your magic will be more effective [04:34] Disabling interface: eth0 ... done. [04:34] * Reloading system message bus config... [ OK ] [04:34] * Restarting network connection manager NetworkManager [04:34] neato. [04:34] somehow that do-release-upgrade run over ssh kinda hurt [04:38] must. kill. network-mangler. [04:39] This is news? [04:39] I keep forgetting that it keeps sneaking back onto the machine and requires another application of the sledgehammer. [04:40] * lamont will file a bug against update-manager to die horribly (or DTRT) if networkmangler is going to trash an sshed do-release-upgrade [04:41] on the bright side, it appears that the update-manager run finished. [04:41] I've always had good luck with apt doing upgrades (even though I know it's 'wrong') [04:42] it's the whole remembering to remove the crap that shouldn't be there any more that got me to switch to playing mvo's game [04:43] Still needed even with autoremove in apt? [04:44] dunno [04:46] sudoers Defaults !env_reset is a good idea when running pdebuild and using sudo... [04:46] that was annoying to find... [04:49] * calc uses dselect for package management :) [04:50] calc: 1995 called, they want their package manager back [04:51] Hah [04:52] does aptitude sort decently yet, last time i tried it was a bit pita to change sort order [04:52] and i like to know what new packages are available each time i update [04:52] which is extremely easy to do with dselect [04:52] hit 'o' a few times and it pops them up [04:53] besides the fact it is very easy to fix conflicts, etc in dselect by placing things on hold or back out if the archive is too screwed up [04:54] i'm sure using something like update-manager/aptitude/etc is great if you only run stable releases that don't change hourly [04:55] eg the last time i used aptitude you had to remember a bunch of arcane characters to change the format of the sort order, it didn't have an option to cycle through common ones [04:55] synaptic shows new packages [04:55] and lets you hold things and such [04:55] so it had a more powerful sort mechanism at the expense of being easy to do [04:55] Amaranth: ah synaptic may be a good replacement for dselect (for me) then [04:56] of course i am also an old fart that has been using dselect over 9 years now so why change... ;-) [04:57] by the time i learned what dselect was i'd already been using synaptic for a year :P [04:58] I used dselect once when I installed hamm [05:15] newt building as of a minute ago [05:15] * slangasek nods [05:20] I'm going to forego putting publisher on auto though, I'm too tired to properly attend it [05:20] heh [05:20] so we did freeze for the CD burn? [05:22] no [05:22] aren't we not freezing but playing broken package roulette? :) [05:22] lamont: see, I'm so tired I said "auto" instead of "manual" [05:22] \o/ [05:33] * lamont builds bind9 9.4.2-1 for debian [05:34] hardy amd64 Successfully built (ACCEPTED) [05:35] * lamont twiddles thumbs [05:35] Uploaded By: Steve Langasek [05:35] launchpad for the win [05:36] dear launchpad, is there any way to identify the signer of the source.changes file? [05:36] ECHAN [05:36] lamont: that'd be nice to have [05:36] lamont: Grab the _source.changes file from the librarian and run who-uploads on it [05:37] StevenK: feh. boring === cjwatson_ is now known as cjwatson [05:37] lamont: It shouldn't misrepresent Changed-By like that, though [05:37] wow. sig survived a cut/paste from the web site [05:37] but how else can we blame slangasek ? [05:38] Changed By: Steve Langasek ; Uploaded By: LaMont Jones [05:38] ? [05:39] it should say that, yes. [05:39] perhaps it misrepresents changed-by because he signed the .changes I sent him instead of re-rolling it? [05:39] Uploaded By: Steve Langasek; Sponsored By: LaMont Jones [05:39] * StevenK sighs at his rsync skills not sticking around, but disappearing into the ether five minutes after he uses them [05:39] Changed-By is meant to be the person in the changelog. [05:39] slangasek: I _SAID_ I was trusting you... :-) [05:39] Not the sponsor. [05:40] Fujitsu: slangasek did the work, lamont uploaded it. My line is correct [05:40] lamont: well I hope you didn't expect me to build the changes file using your name :) [05:40] StevenK: Right. What is the Changed-By it gave? [05:40] I daresay, slangasek [05:40] But I've not looked [05:41] Fujitsu: https://launchpad.net/ubuntu/hardy/+source/newt/0.52.2-11.1ubuntu1 currently shows an Uploaded By: me, which is not entirely accurate [05:41] Changed-By -> Uploaded By in LP speak. [05:42] slangasek: it's your upload... I just sponsored it... :-) [05:42] fair enough [05:43] slangasek: I gave back d-i on the theory that it shouldn't noticibly hurt if I did it a few minutes early [05:44] is it going to noticeably help, though? I don't know what the timings look like for publishing [05:45] the publisher is running cron.germinate now, so the archive itself should be in place [05:46] oh, except for mirrors, hmm [05:46] but buildds fetch directly off ftpmaster.internal (i.e. drescher) [05:46] cjwatson: what mirrors. [05:46] :-) [05:46] so I think it should work [05:46] right :) [05:46] as long as Packages is right [05:47] so, the fact that newt is still listed in 'accepted'...? [05:47] slangasek: means it needs another publisher run [05:48] libnewt-pic | 0.52.2-10ubuntu2 | hardy | amd64, hppa, i386, ia64, lpia, powerpc, sparc [05:48] saith madison on drescher [05:48] and d-i rescored to 10000 [05:48] lamont: try again in an hour [05:48] Heh is it really called madison? [05:48] of course it is [05:48] Hah, way cool [05:49] she's been doing that job for a number of years [05:49] well. it's madison-lite [05:49] post fen-fen then [05:49] cjwatson: as opposed to daring to put things on manual at this point and shove it through? [05:49] I just have it aliased to 'm' so I forget what it's really called [05:49] slangasek: works for me if you're sufficiently awake [05:49] cjwatson: Ah, and it just points directly at the Packages files that the publisher generates [05:50] StevenK: right [05:50] if we're looking at a full hour cycle more, I can probably be that awake [05:50] there's a sort of madison-a-like in the LP web output now mind you [05:50] slangasek: As in, the coffee pot is just boiling? [05:50] slangasek: depends how much we think that the result is going to work [05:50] * StevenK waits for hildon-input-method, so he can do the next part of the dance. [05:51] probably a good time for me to sponsor evand's base-installer merge ... [05:51] Ugh. I had no idea it was that early in London. [05:51] cjwatson: well, getting it in at least lets us do livecds [05:51] and should give us alternate CDs to inspect [05:52] to an extent, yeah. ubiquity will probably have to be uploaded before you can install live CDs [05:52] and I'm planning to get to that point before falling off the network for the night [05:52] oh, erm, ok [05:52] but being able to run them would be a mighty good start [05:53] so what you're saying is that I failed to release-manage my manager in a timely manner... :) [05:53] ubiquity -> evand [05:53] :-) [05:53] ok :) [05:55] * StevenK tries to remember the flags to get rsync to tell him just the files it's transferring [05:56] cjwatson: does it make sense to add slangasek to launchpad-buildd? [05:56] that would at least let him rescore things [05:59] although that requires tech board or infinity [05:59] or a duck [06:00] StevenK: -avk [06:00] er, -avn [06:00] fat fingers === cjwatson_ is now known as cjwatson [06:00] cjwatson: network challenges? [06:01] * lamont throws bind9 9.4.2-1 at sid [06:01] lamont: bloody router [06:01] lamont: -n doesn't transfer though [06:01] quit bleeding on it then. [06:01] Muahaha [06:02] -n says just say what you're gonna do [06:02] Which spits out every file [06:02] Not all of them have changed [06:02] cjwatson: since you went away... does it make sense to add slangasek to launchpad-buildd, so he can rescore things? (which would require tech-board, infinity, or a duck.) [06:02] StevenK: no. it spits out every file it would change [06:03] fine by me [06:03] either by changing contents, or by changing times [06:03] pitti is in launchpad-buildd-admins for basically the same reason [06:03] yeah - that group [06:03] I'll let you use your daylight hours to get TB/infinity/duck to do that for him then [06:04] StevenK: maybe you want -I in there to, so it'll ignore timestamp[s [06:05] lamont: Mmm. I just wish I could remember rsync's runes [06:05] like, say, that guy. :0) [06:05] StevenK: there aren't _THAT_ many... [06:06] and for everything else, there's "man rsync" :-P [06:06] lamont: I know. I don't use it often enough. [06:06] don't ask me to expand -a :-) [06:07] cjwatson: and if someone wants to make me an admin, I can be very good about saying 'no' :) [06:07] ah, evidently no need to publish on manual [06:07] * lamont wants to never ever ever have a duck [06:07] since accepted seems to have been emptied for me in the meantime [06:08] -a should expand to -dpr, or thereabouts [06:08] and more [06:08] -rlptgoD [06:08] à la cp -a [06:08] -a expands to "what I want, for starters" [06:08] right [06:08] -a is "yes, I'm using Unix already" [06:09] -a isn't what I want in this case, though, since I want to ignore symlinks. [06:09] yeah [06:09] piffle. clearly you're not using Un*x [06:09] Clearly, I don't want to copy the build logs to rookery. [06:10] slangasek: so d-i didn't ftbfs on ia64 [06:10] yet. [06:10]