[00:37] broder: bet you're happy about UDS :-) [00:38] i was initially disappointed about not getting to travel, but i'm over it now and pretty stoked [00:38] tour guide hat at the ready [00:38] yep === cdbs is now known as bilal [02:22] Be careful where you tour in Oakland. You may not come back. [02:29] Heh [02:42] psh. it's cool - the hotel is right by the bart station, so the temptation to go into oakland is minimal [02:44] That would be true BART station or not. === bkerensa is now known as bkerensa|AFK [04:31] broder: If you're still around, would you please sync unbound from unstable for me? === Aaron_ is now known as TiMiDo [05:21] ScottK: I can sync unbound for you if you like [05:21] micahg: Yes. Please. syncpackage isn't working for me. [05:21] I'm about to have a mess of security debdiffs too. [05:22] ScottK: well, I'll add this sync to that bug you opened so it has everything [05:22] micahg: Great. It's CVE-2011-4528 [05:22] Unbound before 1.4.13p2 attempts to free unallocated memory during processing of duplicate CNAME records in a signed zone, which allows remote DNS servers to cause a denial of service (daemon crash) via a crafted response. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-4528) [05:22] Urg. [05:22] Bug 907983 [05:22] Launchpad bug 907983 in unbound (Ubuntu Precise) "Multiple security issues with unbound [DSA 2370-1]" [Medium,New] https://launchpad.net/bugs/907983 [05:28] ScottK: done [05:28] Thanks. [05:28] sorry, I haven't updated ubuntu-dev-tools to the version where it says in the bug who it's sponsored for [05:28] but you're in the precise-changes mail [05:29] * micahg knows you don't care about the paperwork :) [05:30] No. [05:30] But if you want to make it up to me you can upload the fix for lucid-security. [05:31] (debdiff in the bug) [05:31] last time I tried that, I killed several hours and still didn't get it done, I"ll try to get someone to look at it in the morning though [05:32] Thanks. [05:43] micahg: If you have time, I'd appreciate it if you could look at oneiric for unbound. There's a configure error where it can't find the python headers and I've no idea why (I think I'm just too tired) [05:44] ScottK: ok, will see if I have time [05:44] Thanks. [06:03] ScottK: looks like a missing build-dep, works fine on a live system [06:03] Weird. [06:03] It build-depends on python-all-dev which was the first obvious choice. [06:04] So that's not it. [06:07] libpython2.6/2.7 are also pulled in. [06:08] * micahg is getting the chroot build log [06:08] Missing -lssl [06:08] ah [06:09] What about that? [06:09] * micahg is getting the chroot to not purge to check [06:09] Dunno why that would work on a live system though. [06:11] BTW, the release version won't build either [06:11] I guess that's obvious considering the patch [06:14] wow, a whole bunch of stuff is broke [06:15] * micahg has no idea why this would work on a live system [06:18] It built at one point. [06:21] in July when we had the autosync [06:23] not sure, it seems to think my installed python env is fine, but the chroot env is not [06:24] ah, it doesn't seem to be multiarch ready, but still that shouldn't work on the live system either unless I have some old hack installed [06:26] it also won't build twice in a row which is annoying, but that's for another discussion (happening on debian-devel) [06:27] The next version (the one you just sync'ed over_ is multi-arched. [06:28] right [06:28] * ScottK is going to go to bed. [06:28] Good night. [06:29] good night === almaisan-away is now known as al-maisan [08:04] good morning [08:10] morning dholbach [08:11] hi ajmitch === al-maisan is now known as almaisan-away [09:56] hi there [09:58] what can I do to get my packages to ubuntu? I already have a working PPA... [10:20] ondras: https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages [10:21] tumbleweed: ah, thanks! [10:22] tumbleweed: the bug should be filed even if the software is already packaged? [10:22] tumbleweed: with [needs packaging] prefix? [10:36] ondras: did you miss the bit about going through debian? [10:37] that really is the beast approach [10:38] yeah, I already sent them a RFS email [10:38] but according to #debian-mentors, this can take months or years [10:38] :/ [10:41] it's normally a lot faster than that [10:41] what are you packaging? [10:43] tumbleweed: http://code.google.com/p/v8cgi/ [10:44] (basically a wrapper around libv8 + several modules; somewhat similar to node.js but with different architecture/usage) === Quintasan_ is now known as Quintasan [10:45] tumbleweed, hum, I don't know what is the normal duration, but at least the package that I have submitted has taken ~2month to pass the debian ftpmaster step (and I was lucky to personnally know the debian sponsor which speedup a little bit the process) [10:51] ondras: ah, that sounds like a non-trivial package [10:51] jporsini: ftpmaster review is a lot faster than it used to be: http://ftp-master.debian.org/stat.html [10:53] tumbleweed, yep, I was currently on this page, and you are right, it appears to be now very fast, definitively a good news, so forget my previous sentence:) [10:53] just FTR, average time in Debian NEW during 2011 has been 3 days and 5 hours [10:54] tumbleweed: it is a multiple binary package, yes [10:55] tumbleweed: but it is a standalone software with basically no dependencies [11:13] DktrKranz: and we appreciate it! [11:35] * Laney bakes DktrKranz a cake http://farm1.static.flickr.com/23/37895746_90c5c30ca9.jpg * [11:35] * disclaimer: I may not have baked that [11:50] Looks like cheese cake. [11:52] it does, and has put me right in the mood for some [11:53] * Laney eyes the build queues [11:53] * Laney goes to decorate the tree instead [11:57] Laney: does this imply that you would have a undecorated tree if the build queues were empty? [11:58] * geser wonders if it's possible to deduce how full the build queues were from how much Laney decorated his tree? [12:17] geser: that is something we cannot know === yofel_ is now known as yofel [13:27] ScottK: I'm uploading your unbound fixes now, just as soon as I fix the oneiric ftbfs. I've added two more older cve fixes to the lucid package also. [13:27] mdeslaur: OK. I wasn't aware I'd missed any. Thanks. [13:28] mdeslaur: Please let me know what the solution is on oneiric once you sort it out. It's got me confused. [13:31] ScottK: I'm still poking at it...I suspect this is the fix: http://anonscm.debian.org/gitweb/?p=users/edmonds/unbound.git;a=commitdiff;h=650df562a19e196e9acbc4898559f2fa61a8d90e [13:32] hrm, maybe not [13:33] micahg said he got it to build on a non-minimal system, so it may just be a missing dependency. [13:33] * ScottK has NFC what though. [13:34] huh [13:36] [01:03:13] ScottK: looks like a missing build-dep, works fine on a live system [13:36] [01:03:32] Weird. [13:36] Based on the failure though, I'm not sure how that's possible. [13:39] is there a build log available somewhere? [13:41] No, but just try to rebuild the oneiric unbound package. [13:41] It dies pretty quickly in configure. [13:41] Actually, i may have one [13:43] I do. [13:44] geser: http://kitterman.com/kubuntu/unbound.oneiric [13:52] I think it's multiarch related, and I'm backporting a small change from the precise package [13:55] hrm, that didn't work [13:55] * mdeslaur scratches head [14:24] does the oneiric version build in a precise chroot? and does the precise version build in a oneiric chroot? (I don't have access to my pbuilders right now to test it myself) [14:36] The precise version does not build on the oneiric chroot. [14:37] The oneiric version builds on precise. [14:38] mdeslaur: That (to me) points to a bug in another package. [14:43] Could it be python itself? As this check tries linking with only python, on the other side (if it's really python) I wonder why no other package stumbled over it [14:43] That's what I was thinking, but no idea. It seems distinctly odd. [14:47] ScottK: yeah, I tried the precise package on oneiric, and it failed as well [14:47] But the other way around works (oneiric package on precise) [14:48] yep, so it's something with oneiric [14:48] I'm still poking at it [14:48] Better you than me. This is way over my head. [15:01] well, I'm stuck...don't know what to try next [15:03] Ask doko since it might be Python related? === almaisan-away is now known as al-maisan [15:22] Reading -devel, I don't fell bad I didn't figure that out. [15:22] fell/feel [15:30] ScottK, geser: we need to build it with binutils from oneiric-updates [15:30] ScottK: heh, I wouldn't have figured it out :P [15:30] Great news though. [15:31] It also explains why it built on micahg's system. I'm sure he has -updates enabled. [15:34] ScottK: yeah...thanks for suggesting I ask him [15:35] It's ironic though since I suggested him for a reason entirely unrelated to the actual answer. [15:35] I had tried rebuilding with -updates enabled in my schroot, but binutils wasn't updated in it === al-maisan is now known as almaisan-away [20:03] tumbleweed: You're famous - http://askubuntu.com/badges/40/tumbleweed [20:04] :) [20:04] now I need to ask a question so I can earn it [20:09] broder: maybe backportpackage should use -nc [20:10] (just ran across one of those package that needs all its build deps to clean) [20:11] makes sense if there are no source changes [20:12] * micahg had one of those last night also [20:13] * tumbleweed does that for syncpackage too [20:13] oh and sponsor-patch [20:28] tumbleweed: makes sense to me [20:32] ScottK: you're using askubuntu now as well? [20:33] i answered some questions there a bit back [20:33] * micahg thinks ScottK mistook it for as kubuntu :) [20:33] heh [20:33] as kubuntu [20:34] want points? answer http://askubuntu.com/questions/90070/how-does-the-shutdown-button-works-without-asking-for-password [20:34] that one's easy [20:35] exactly [20:35] * tumbleweed doesn't know why anyone bothers with askubuntu. jorge answers everything [20:36] not *quite* everything [20:36] just 95% [20:36] * ajmitch did at least get a free tshirt from them :) === Resistance is now known as EvilResistance [22:16] hmm, I'm a bit out of date on this: is requestsync completly deprecated in favor of syncpackage? [22:16] (given that I have the necessary upload rights) [22:18] yes [22:18] for doing your own syncs [22:18] for sponsoring syncs, you need to be sure to have a fairly recent version of syncpackage that supports it [22:19] (either from precise or ppa:udt-developers/daily) [22:19] ok, thanks :) [22:21] broder: is there an easy way to determine what build-deps or required packages are defined in the debian/control file in a source package between versions (i.e. precise vs. oneiric versions) [22:21] without me having to download the source packages and then diffing them [22:21] s/them/the control file/ [22:22] packages.ubuntu.com shows it, or UDD [22:24] packages.ubuntu.com does show the dependent packages for the program to run, but i dont see any refs about the build-deps... what's UDD? [22:25] http://packages.ubuntu.com/source/precise/libproxy [22:25] UDD is the ultimate debian database, see udd.debian.org [22:26] ah source that's what i forgot to look at :P [22:26] or they are in the Sources file in a repository [22:26] which you can query with e.g. grep-dctrl [22:27] is there a way to get debian source via apt-get source in ubuntu? [22:28] it's in the Sources files [22:28] ah, Laney said that [22:28] nixmaniack: we have pull-debian-source to make it easy [22:28] but yes, you can have deb-src lines pointing at Ubuntu [22:28] err at Debian [22:29] tumbleweed, thanks for the help :) [22:29] apt-get doesn't make that particularly user-friendly, though, you have to specify which release you are downloading from, whenever you download it [22:29] also see chdist in devscripts [22:29] but for most uses, pull-debian-source [22:30] which default mirror it uses? [22:33] the primary archive. But you can configure it (see ubuntu-dev-tools (5)) [22:34] tumbleweed, okay, got it. [23:39] um... i'm getting a 500 error when trying to upload a backported package to a PPA for my use... is this a bug, or did it not upload? https://pastebin.com/aPwzfQhU [23:40] (using backportpackage btw)