[01:17] made it to level 6 :) [01:33] fyi, if anybody's working on the ctf, i recommend adding -stripe to the beginning of any google searches you do [01:33] there are, e.g., gists on github with solution writeups that have already been indexed === jtechidna is now known as JontheEchidna === lifeless_ is now known as lifeless [09:48] If I want to get an icon-size fix into Precise at this stage, would it be easier to sync over a Debian update (with very minor packaging updates), or do the patch as an isolated patch for Ubuntu? [09:55] minor packaging updates => sync [09:55] (assuming they won't impact othe rpackages) === Quintasan_ is now known as Quintasan === and`_ is now known as and` [17:43] hello [17:44] Hi all! [17:44] im a c embedded programmer [17:44] could i get some help on getting into the ubuntu project? [17:51] qwebirc91098: sure [17:53] qwebirc91098: do you have any particular questions? [18:07] hm yes [18:07] first, can you point me to a specific bug/feature that i can work on? [18:08] never worked on a an opensource project before [18:08] would glad to get some overview [18:08] https://wiki.ubuntu.com/MOTU/GettingStarted has some hints [18:09] qwebirc91098, first change your nick iwth /nick command :). Then read channel topic === qwebirc91098 is now known as SemaphoreVsMutex [18:10] we're just coming up on beta 1, https://wiki.ubuntu.com/PrecisePangolin/ReleaseSchedule so the things of most interest are fixing release-critical bugs, not bringing in new packages or new versions of things [18:10] thanks for the link.. i read some of the bugs in the bug section [18:10] was hard for me to follow since it was all fixed or being fixed [18:12] harvest should help you to find things to look at [18:15] ooh people around. [18:16] i am investigating bug 898463 [18:16] Launchpad bug 898463 in software-center (Ubuntu) "GParted is in USC "Themes & Tweaks" section rather than "System"" [Undecided,Confirmed] https://launchpad.net/bugs/898463 [18:16] i note that there is no fault in gparted or app-install-data [18:17] and i am not sure where the problem is in usc. any pointers? === bulldog98_ is now known as bulldog98 [20:26] tumbleweed: did you get any chance to take a look at tesseract 3.0? [20:26] ockham: no, but I suppose I could now [20:27] tumbleweed: cool! tell me if there are any menial tasks i can help with [20:28] pipedream: ohi [20:29] ohi [20:29] tumbleweed: https://lists.ubuntu.com/archives/ubuntu-motu/2012-February/007188.html [20:30] pipedream: yar, just saw that [20:30] did you see http://wiki.debian.org/DebianScience/Sage ? [20:30] No [20:30] * tumbleweed doesn't know too much about sage, but as I understand it, it's a big bundle of other stuff [20:31] yes [20:31] It's a python front-end to like 50 packages [20:31] so, when we are shipping different versions of the "other stuff" to sage, things get problematic [20:31] So I don't want to ship a sage based on the debian vcersions [20:31] sage makes all it's own versions [20:31] which means you we have to carry essentially two copies of all of those packages [20:31] sage has thigns called spkg's (tarballs) which are the upstream exact versions whic hmake sage compile so cleanly [20:32] Yes, the sage package will be huge [20:32] but my goal is to aid Sage adoption, and besides using bandwidth/space, I don't really tsee the problem with this [20:32] Sage-spkg'es do not overwrite system versions [20:32] you call them from sage (or optionally install direct access to /usr/local/bin/) [20:33] the most commonly-stated downside of duplication is that security issues have to be fixed in multiple places [20:33] So how about I just make a PPA first, and worry about all that later [20:33] yeah, why not [20:33] I also suggest talking to the debian science people [20:33] Then debian/ubuntu won't be shipping insecure stuff [20:34] tumbleweed: apt-add-repository ppa:aims/sagemath [20:34] (there is nothing in there ;) [20:34] :) [20:35] the build failure in your mail didn't actually include gcc's error [20:35] does sage's build system hide gcc output? [20:35] nope [20:35] I forgot to attach the log file [20:35] pastebin? [20:35] it's a 1.2M text file [20:35] only the bit around the error is important [20:35] http://users.aims.ac.za/~jan/maxima-5.23.2.p3.log.lzma [20:36] ah, or that [20:36] the line that failed is a pretty long line [20:36] I figure that is details [20:36] I need a mentor as I will run into such problems again and again [20:36] and I can't pinpoint them to the package or to my (lack of) packaging skills [20:36] pipedream: I'm more than happy to help out [20:37] looks like a quoting issue [20:37] maxima, btw, built cleanly when NOT run from debuild [20:37] we already know each other, so why not :P [20:37] tumbleweed: I am at an ICTP workshop, and this is my project for the week [20:37] ah, I see [20:37] whatever time you have to help, super thanks [20:37] but I guess this will go on after that [20:37] and also as sage thakes 5 hours to build the iterations are a bit long [20:38] So I will keep asking on sage-devel, ubuntu-motu, and #ubuntu-motu for a few days [20:38] do you use ccache? that will cut down time on repeated builds of C/c++ parts [20:38] then I am on two weeks holiday and will pick it up again after [20:38] * pipedream googles ccache [20:38] sage itself will also pick up from where it left off [20:38] where it b0rked [20:39] jtaylor: what quoting issue do you see? [20:39] I just wish I could see gcc's outptu [20:39] that maxima.log I pasted has it? [20:39] /usr/bin/ld: unrecognized option '-Bsymbolic-functions -Wl' [20:39] oh, I missed that [20:39] at the end of the massive line, duh [20:39] a failuire like that also occurs in scons packages quite often [20:39] yes, that is a quoting issue [20:40] (yes, but only via debuild?) [20:40] debuild exports certain buildflags a normal build would not [20:40] including -Bsymbolic-function and security related stuff [20:40] so here is where I have a lot to learn ;) [20:41] jtaylor seems to know everything about these types of issues, you're in good hands :P [20:43] pipedream, look topic links.... maybe you read then something useful :) [20:44] valdur55: I am reading a lot at developer.ubuntu.com, debian guides, and launchpad [20:45] pipedream, ok! good to know! [20:48] ScottK: if I were to kick the ball rolling on tesseract (I'm just starting some local tests), are you up for a pile of NEW review? (there are 67 tesseract source packages) [21:00] tumbleweed: I'm getting some reply on sage-devel mailing list. [21:00] I guess it is not appropriate to keep this going on both sage-devel and ubuntu-moto, and yet that is exactly the cross section where people can help solve this problem... [21:01] I'm busy downloading the source so I can have a look [21:01] anything interesting in your debian/rules? [21:02] uhm, no [21:02] I literally read and followed http://developer.ubuntu.com/packaging/html/packaging-new-software.html [21:02] from the top down [21:02] cool [21:03] tumbleweed: Are they sync's from Debian? If so, no problem. [21:03] yes [21:03] tumbleweed: Go for it. [21:03] ok, as soon as I've tested it locally [21:17] pipedream: re "googling and not finding much" [21:17] the flags are coming from dpkg-buildflags [21:17] heh [21:18] which is absent in my package so I get defaults? [21:18] I can also try to pre-empt the difficult part by doing an easier project first: [21:18] debuild runs it [21:19] (in Debian, this doesn't happen any more, as these days, dh runs it, at compat levels >=9. But we kept this enabled in Ubuntu for precise) [21:19] a package, say, sagemath-upstream-binary, which gets sage's prebuilt binar and make a postinst script that untars it, moves it, fixes permissions [21:19] err s/debuild/dpkg-buildpackage/ [21:20] ick [21:20] building from source is definitly preferable :) [21:20] yes, but just to practice my packaging a bit, and to have a result for people to use while I work on it [21:21] fair enoguh [21:21] (it'll also build, much faster :P ) [21:21] Yes, and I'll actually have something to show at the end of the workshop ;) [21:22] I cxould probably do that in one day if I learn how to make a virtual package from scratch and add a postinst, but that seems to rely on me already knowing how to build source packages, bit of a catch 22 [21:23] I suggest not doing anything fancy in a postinst [21:23] rather take a linux binary tarball as your source, and build debs containing everything where you want them [21:24] the only reason people do crazy downwloading and unpacking in postinsts, is that the software they are installing isn't redistributable [21:25] pipedream: look at the Ubuntu nvidia-cg-toolkit package for an example of what I'm suggesting [21:26] ok [21:30] sorr, that's not a simple example, there is lots of other stuff going on there too [21:30] Yes, how about a doc package that builds nothing? [21:31] what is a proper debian location for sage? [21:31] /usr/bin/sage and /usr/share/sage/* ? [21:32] /usr/lib/sage, if you want it all in one place. /usr/share is for arch-independant data only [21:32] and bin, yes [21:32] ok, thanks [21:32] pipedream: sage used to be packaged in debian. it ended up getting removed because the packager got busy running a startup, but have you looked that up? [21:32] yes, Tim Abbott/ksplice . He was however debianizing, depending on all the components' debian versions [21:33] I want an all-in-one sage [21:33] sage versions of all the components [21:33] oh, i see [21:33] which probably wouldn't be acceptable in Debian, but it's closer to what sage upstream wants, and debs of that would be useful (in your PPA, of course) [21:33] (just curious - tim is a friend of mine from school) [21:34] that project is too big for me now [21:34] maybe one day [21:40] 3 jan@osprey:~/sagemath-upstream-binary$ls usr/lib/ [21:40] sage-4.8-linux-64bit-ubuntu_10.04.3_lts-x86_64-Linux [21:40] 0 jan@osprey:~/sagemath-upstream-binary$bzr dh-make sagemath-upstream-binary 4.8-10.04.3-0ubuntu1 [21:40] bzr: ERROR: command 'dh-make' requires argument TARBALL [21:41] ^^ so this package would have no upstream tarball [21:41] pipedream: rename the tarball you downloaded to $PKG_$VER.orig.tar.gz [21:42] (on 12.04 it renames for one, I think, but OK, I had removed the tarball as it is uncessary for this method) [21:42] you still need an orig tarball [21:42] (also it is lzma, I guess I need to decompress and use gz?) [21:42] no, .lzma is fine [21:42] (although tell the upstream that they should use .xz instead :P ) [21:43] bzr: ERROR: Unable to import library "lzma": No module named lzma [21:43] install python-lzma? [21:44] right [21:46] ockham: tesseract 3 doesn't break the world (that I can see) [21:46] * tumbleweed starts pressing buttons [21:46] tumbleweed: yeah! [21:46] go for it! [21:47] * tumbleweed quickly adds a -y option to syncpackage [21:49] http://pastebin.ubuntu.com/857071/ [21:49] ^^ sagemath is a bit big? [21:50] lol [21:50] I do suggest filing a bug about that [21:51] Go the bzr-less route for now, or do a merge-mode bzr branch (only containing the debian directory) [21:52] network-manager coredumped while trying to file a bug (I'm running 12.04 right now, next to 11.10 on this laptop) [21:53] cyphermox: ^ [21:53] And I have a sage compile running, so I'll reboot to 11.10 later [21:55] unreportablereason: networkmanager version has changed since then [22:02] seocnd time round it worked (nm didn't crash) [22:02] same bug as this one maybe: https://bugs.launchpad.net/ubuntu/+source/bzr/+bug/531612 [22:02] Launchpad bug 531612 in bzr (Ubuntu) "failure to extract content > 2^31 bytes (OverflowError: signed integer is greater than maximum)" [Low,Confirmed] [22:03] your traceback looked bzr-builddeb specific [22:03] is sage bigger than 4gb? :O [22:07] https://bugs.launchpad.net/ubuntu/+source/bzr/+bug/941131 [22:07] Error: launchpad bug 941131 not found [22:07] 3G [22:07] "This report is private" ? [22:08] ^ First time I've seen that [22:08] crash bugs potentially contain private data [22:08] ok [22:08] the re-tracer removes the dumps and attaches stack traces [22:09] ScottK: all yours: https://launchpad.net/ubuntu/precise/+queue :) [22:09] So, I don't know the bzr-less route or merge-mode bzr. Reading http://doc.bazaar.canonical.com/bzr.dev/en/mini-tutorial/index.html [22:09] removals filed too? [22:09] also, damn you re: ctf [22:09] Laney: :) [22:09] level05 now [22:10] 5 is easy [22:10] I don't think I'll start on ctf, it'll kill any time I had [22:10] i pasted the curl line it gave and it didn't come back, so i decided to give up for now [22:10] tumbleweed: that's a lot of syncs :) [22:10] yup [22:10] level 5 is obfuscated, but easy once you realize what you have to do [22:11] i liked 3 [22:12] i still don't know what specifically i'm supposed to do for 6 [22:12] got more useful things to do, like make a haskell graph go green [22:12] heh [22:13] * ajmitch should get back to the fun rcbugs list [22:13] so many to look at on there [22:13] broder: I found a rather neat solution to that, after lots of dead ends [22:22] Thanks. I'm off until Monday morning now, I think. [22:23] those sure are a lot of tesseract packages :) [22:23] tumbleweed's trying to win on the upload activity graphs :) [22:24] they mostly just contain one file [22:24] * tumbleweed blames the debian maintainer [22:25] * ajmitch still has a long way to go to get back on that list of uploaders for precise [22:25] where is that list again? [22:25] There's a list? [22:26] http://people.ubuntuwire.org/~stefanor/ubuntu-activity/ [22:26] Ah. Didn't know about that. Thanks. [22:27] it's intentionally not listed anywhere :) [22:27] still interesting to look at [22:27] tumbleweed: don't worry, I won't tell anyone about it ;) [22:27] :P [22:28] * Laney tweets it [22:28] as long as nobody sends me angry e-mail about comparing canonical contributions to community members, do whatever you want :) [22:28] and spoil all our fun? [22:34] tumbleweed: I'm assuming you're tracking if all the tesseract builds succeed. I'm just worrying about Newing the stuff that does. [22:35] a --subscribe-me-to-bugs-and-build-failures on syncpackage would be nice - I currently don't know if the person syncing gets the build failure emails for syncs [22:35] ScottK: yeah, keeping an eye on them (and they all built locally) [22:36] ajmitch: If you get the accepted mail, you'll get the build failures, I'm pretty sure. [22:36] ajmitch: at least tde debian maintainers aren't getting emails any more [22:36] bug 862251 for the sync requestor [22:36] Launchpad bug 862251 in Launchpad itself "Sync requester doesn't receive build failure emails" [High,Triaged] https://launchpad.net/bugs/862251 [22:38] tumbleweed: thanks, hopefully I haven't missed any failures on archs that I don't have [22:38] not that I've got many syncs :) [22:39] fortunetly there's a column on launchpad.net/people/+me/+synchronised-packages [22:40] listing failures [22:40] great, one armhf failure [22:43] oh, even test-rebuild failures show up there, neat [22:50] tumbleweed: sikuli depwait [22:51] yeah, it looks like tesseract-dev isn't published yet [22:51] although LP said it was, before I retried it [22:55] LP says it's published at the start of the publisher run, but the package isn't actually available until shortly after it finishes. [23:18] Laney: there aren't plans to update ghc in precise, are there? [23:18] yes [23:19] ok, was just looking at the rc bugs list, found some packages like gtk2hs-buildtools that will need synced if it's updated [23:20] it will all be synced [23:20] * Laney rubs palms [23:20] sounds like madness [23:25] tumbleweed: thanks for syncing tesseract! incidentally, this has just been posted to LP: https://bugs.launchpad.net/simple-scan/+bug/483391/comments/21 [23:25] Launchpad bug 483391 in Simple Scan "Extract text using optical character recognition (OCR)" [Wishlist,Triaged] [23:26] hmm, getting proper an OCR option into Simple Scan sounds tempting...