| === sergiuse1s is now known as sergiusens | ||
| sarnold | tkamppeter_: I'm sorry to NAK openjpeg with so little time before featurefreeze, but that code needs serious attention before it should be exposed to untrusted data. | 04:37 |
|---|---|---|
| maclin | cjwatson, hi, I have a problem about the ubuntukylin daily iso. The user cannot login neither in live mode nor after installation. | 06:05 |
| maclin | we have proposed the Bug #1218172. I guess it's about the gnome-session and gnome-settings-daemon packages. Can you help to confirm it? | 06:08 |
| ubot2` | Launchpad bug 1218172 in UbuntuKylin "lightdm cannot login (系统无法登录) on ubuntukylin daily iso(0822+)" [Critical,Confirmed] https://launchpad.net/bugs/1218172 | 06:08 |
| infinity | maclin: Probably a better question for #ubuntu-desktop than here. | 06:19 |
| maclin | infinity, thanks. we do the test in ubuntu daily iso and the problem does not exist. So I guess this problem is about the image building process? | 06:23 |
| infinity | maclin: Hard to say without someone who would be better at debugging it (like the #ubuntu-desktop people) | 06:25 |
| maclin | I find that the gnome-session and gnome-settings-daemon get lost in UbuntuKylin daily ISO | 06:25 |
| infinity | maclin: Hrm. Curious. | 06:26 |
| infinity | Where do the ubuntukylin seeds live? | 06:26 |
| maclin | infinity, we use the default-setting | 06:29 |
| infinity | maclin: So, you've got bigger problems than just a few missing packages. No compiz, no unity, no ubuntu-desktop. This is probably our fault somewhere. :P | 06:36 |
| infinity | maclin: Let me respin a new ISO and see if this was transient.. | 06:38 |
| mlankhorst | hmz | 06:48 |
| mlankhorst | infinity: yikes llvm-3.3 was kicked back to universe because mesa ftbfs because of libjsoncpp | 06:48 |
| infinity | mlankhorst: Looks like libjsoncpp needs an MIR... | 06:50 |
| infinity | mlankhorst: lcov as well. | 06:50 |
| infinity | http://people.canonical.com/~ubuntu-archive/component-mismatches-proposed.svg | 06:50 |
| mlankhorst | hm fun | 06:51 |
| infinity | mlankhorst: Happy filing. | 06:51 |
| infinity | At least your life doesn't suck as hard as the people who thought location-service needed to depend on 7000 ruby gems. | 06:52 |
| mlankhorst | fortunately lcov only depends on debhelper | 06:54 |
| infinity | And libgd-gd2-perl, but that's already in main. | 06:55 |
| RAOF | infinity: location-service is ruby? Sounds like muchos fun! | 06:56 |
| infinity | RAOF: It build-deps on some ruby doc generation thing, it looks like. | 06:56 |
| infinity | RAOF: Though, why this was chosen for a Canonical project (I assume this is a Canonical project?) is beyond me. | 06:56 |
| infinity | Given the "libubuntu-" names, I'm going to go with probably yes. :P | 06:56 |
| RAOF | :) | 06:57 |
| mlankhorst | infinity: meh, lcov loooks like an optional dependency, only used when codecoverage is added to DEB_BUILD_OPTIONS :P | 06:58 |
| infinity | mlankhorst: Alright, so we could patch that our or MIR it. Whichever. | 06:58 |
| mlankhorst | but shrug I'll mir it | 06:58 |
| infinity | mlankhorst: I'm assuming we're less lucky on the json thing, though. | 06:58 |
| infinity | mlankhorst: MIRs for build, debug, and analysis tools are generally non-controversial, compared to runtime stuff. | 06:59 |
| mlankhorst | ah k | 06:59 |
| mlankhorst | I added one for lcov https://bugs.launchpad.net/ubuntu/+source/lcov/+bug/1218209 | 07:15 |
| ubot2` | Launchpad bug 1218209 in lcov (Ubuntu) "[MIR] lcov" [Undecided,New] | 07:15 |
| mlankhorst | still working on libjsoncpp | 07:15 |
| mlankhorst | but meh from the bug reports in lcov it looks like it might be better to patch out codecoverage support.. | 07:16 |
| infinity | Hrm? I see one bug in Ubuntu, 3 in Debian... | 07:17 |
| infinity | Certainly nothing terrible. | 07:17 |
| mlankhorst | yeah about gcc-4.7 support, makes me wonder if things even work :P | 07:17 |
| infinity | Hrm, fair enough. | 07:19 |
| mlankhorst | I'll give it a shot, but if not could lcov be dropped as build-dep? it would probably fail if doing a manual build with codecoverage added to DEB_BUILD_OPTIONS though, but then again that would already be | 07:19 |
| infinity | Patching that out for now seems reasonable. | 07:19 |
| infinity | I'll just patch it out now. | 07:19 |
| mlankhorst | ok :-) | 07:20 |
| mlankhorst | shall i mark the lcov bug as invalid then? | 07:24 |
| infinity | I'm on it. | 07:24 |
| mlankhorst | k | 07:24 |
| infinity | mlankhorst: I think I'll just leave it "new", so it magically pops up again if we re-add the build-dep after lcov is fixed. :P | 07:26 |
| infinity | mlankhorst: Hrm. So, llvm-toolchain-3.2 uses a bundled jsoncpp. Ick. | 07:35 |
| infinity | mlankhorst: But we could revert 3.3 to doing the same thing. Or fix 3.2 to use the system library. Only doing one or the other seems silly. | 07:37 |
| infinity | Ahh, and indeed, if I merged 3.2 with Debian, it has the same "use the system version" change. | 07:39 |
| infinity | mlankhorst: So, yeah, that's probably the right thing to do, so carry on with the MIR. Mention that it affects both 3.3 and 3.2 (or will, after we merge it :P) | 07:40 |
| mlankhorst | https://bugs.launchpad.net/ubuntu/+source/libjsoncpp/+bug/1218220 | 07:42 |
| ubot2` | Launchpad bug 1218220 in libjsoncpp (Ubuntu) "[MIR] libjsoncpp" [Undecided,New] | 07:42 |
| mlankhorst | infinity: but llvm 3.2 will no longer be in main after mesa builds against 3.3 :P | 07:45 |
| infinity | Is mesa really the only thing that keeps llvm in main? | 07:46 |
| mlankhorst | i think so | 07:46 |
| infinity | Hrm. | 07:47 |
| infinity | Given how stagnant this is upstream, maybe I should just revert to the static copy for now, then. | 07:47 |
| infinity | The libjson0 package also has no symbols file, or other such fanciness. | 07:48 |
| infinity | Which, of course, a static version wouldn't need. | 07:48 |
| * infinity ponders. | 07:48 | |
| mlankhorst | ah right, I missed that | 07:48 |
| infinity | Okay, so, when you mean to remove libjsoncpp0, don't scare the crap out of yourself and type "apt-get purge libjson0" | 07:49 |
| mlankhorst | haha | 07:50 |
| mlankhorst | interesting, it has no issues trying to remove upstart here.. | 07:51 |
| infinity | Yeah, maybe I'll just re-enable the static copy for now. It's a few thousand lines of public domain cargo-cult, might actually be safer than a shared lib with lousy versioning. | 07:51 |
| mlankhorst | hm probably | 07:51 |
| infinity | Wish I'd come to this conclusion before I made you file bugs... And before I uploaded. :P | 07:52 |
| mlankhorst | lol | 07:52 |
| mlankhorst | hey sometimes showing people they're wrong is best done by doing exactly what they want you to do | 07:52 |
| infinity | Alright, uploaded AGAIN. :P | 07:56 |
| mlankhorst | infinity: will it promote the build to main because llvm-toolchain is still being kicked back to universe? | 07:57 |
| infinity | mlankhorst: llvm-toolchain-3.3 is still in main. component-mismatches wants me to move it, that doesn't mean I have done so. | 07:58 |
| mlankhorst | ah | 07:58 |
| infinity | mlankhorst: So, once it's built, retries on the mesa builds should work. | 07:58 |
| mlankhorst | yeah | 07:59 |
| infinity | Now, I totally didn't test build this, cause I'm a cowboy like that, so watch it fail... | 07:59 |
| mlankhorst | if you think you're rough, I didn't look twice before crossing the road today | 08:00 |
| infinity | Don't forget to look up for Cesnas making emergency landings. | 08:01 |
| infinity | Cessnas* | 08:01 |
| mlankhorst | :D | 08:01 |
| infinity | Okay, the Internet is kinda awesome. | 08:02 |
| infinity | I had no idea Cessna actually listed prices online. If there was a shop with a "buy now" button, I'd be set. | 08:02 |
| pkern | The Internet is full of cats. | 08:03 |
| infinity | Only 26M for a top of the line Citation, a bargain. | 08:03 |
| mlankhorst | infinity: hahaha | 08:03 |
| mlankhorst | but it only takes 3 hours to build on armhf, I'll get some coffee :) | 08:15 |
| infinity | mlankhorst: 39 minutes on PPC though, so at least I get a test build soon. :P | 08:17 |
| Laney | can we haz emails about component-mismatches-proposed? | 08:17 |
| infinity | Hrm, yeah, we may have failed to advertise its existence. | 08:18 |
| infinity | Then again, most people aren't aware of the non-proposed version either, from what I can tell. | 08:18 |
| Laney | I don't mean advertisement | 08:18 |
| Laney | I mean "New component-mismatches" | 08:18 |
| infinity | Oh, you mean emails to -release on changes. | 08:18 |
| Laney | that thing | 08:18 |
| infinity | That could be noisier than expected. Not sure. | 08:18 |
| infinity | Perhaps still a good idea. | 08:19 |
| Laney | Probably if we have both going | 08:19 |
| Laney | but still, it's remarkably easy to miss, at least for me | 08:19 |
| sil2100 | infinity: hello! Can I poke you about moving unity-scope-mediascanner from the NEW queue? I see the sync request is there and it would be cool to get it out to the universe (tm) | 08:22 |
| infinity | sil2100: W: unity-scope-mediascanner source: field-name-typo-in-dep5-copyright file -> files (line 6) | 08:31 |
| infinity | sil2100: Care to fix your copyright file in bzr? | 08:31 |
| sil2100 | Oh noes | 08:31 |
| sil2100 | Doing that | 08:31 |
| infinity | Otherwise, looks fine. | 08:32 |
| sil2100 | What a typo... thanks for accepting! | 08:34 |
| cjwatson | maclin: gnome-session is removed as part of the line "apt-get purge --auto-remove -y ibus ibus-gtk ibus-gtk3 python-ibus ibus-pinyin-db-android libopencc1" in ubuntukylin-default-settings/hooks/chroot | 08:50 |
| cjwatson | maclin: So perhaps you need to work out some more subtle way to do that | 08:50 |
| maclin | cjwatson, thanks, we will check the problem of dependency | 08:55 |
| Laney | We have gnome-settings-daemon depending on ibus so removing that is going to be problematic | 08:56 |
| maclin | Laney, the dependency between ibus and gnome-settings-daemon is added recently? | 09:00 |
| Laney | maclin: Some time this cycle; let me check when | 09:00 |
| seb128 | this month | 09:00 |
| seb128 | that's quite recent yes | 09:00 |
| seb128 | but that was in the work for some months, we emailed desktop/devel list about that mid-cycle | 09:00 |
| Laney | Wed, 21 Aug 2013 11:22:42 -0400 | 09:01 |
| Laney | But indeed it was no surprise | 09:01 |
| maclin | yes, the problem appears since 22th this month | 09:01 |
| jbicha | maclin: you guys are subscribed to ubuntu-devel, right? | 09:05 |
| seb128 | Laney, maclin: it would be much easier for everyone if UbuntuKylin was using the same input method framework than Ubuntu, not sure why we can't reconciliate both on ibus :/ | 09:07 |
| Laney | Yes, but I can't pretend to understand the issues there | 09:08 |
| Laney | Don't we have some new guys to work on this stuff now? :-) | 09:08 |
| maclin | seb128,we have to choose fcitx as default instead of ibus. | 09:09 |
| seb128 | maclin, you don't "have" to, you decide to do that ;-) | 09:09 |
| seb128 | maclin, you could add the input methods engines you need to ibus instead | 09:09 |
| maclin | seb128,yes,it's suitable for us:) | 09:12 |
| Laney | I thought you could use im-config to configure this stuff anyway | 09:12 |
| Laney | or do they conflict? | 09:12 |
| maclin | seb128, Laney, if we don't remove ibus, the default input method will be ibus | 09:13 |
| Laney | maclin: If you don't do anything else, that is true - I'm saying that I thought this was the problem that im-config solves | 09:14 |
| maclin | Laney, does im-config change the default option in default-settings? | 09:17 |
| Laney | maclin: I'm not sure how you set a system default with it | 09:18 |
| Laney | You could talk to happyaron though | 09:18 |
| infinity | im-config invoked from the root account updates the system configuration file | 09:19 |
| infinity | ^-- From the manpage. | 09:20 |
| infinity | So invoking im-config during image build should set it system-wide. | 09:20 |
| cjwatson | man --nj IYF | 09:20 |
| cjwatson | (for pasting) | 09:20 |
| infinity | But I like all the extra spaces! | 09:20 |
| Laney | It made it scan in an exciting way | 09:21 |
| cjwatson | To the tune of the Flintstones? | 09:21 |
| Laney | maclin: So yeah. I'd look into that if I were you. Yay for fewer forced removals and unpleasant surprises. | 09:21 |
| * infinity is not having a yabba dabba do time. | 09:22 | |
| maclin | jbicha, just see your message, yes, we did | 09:23 |
| maclin | Laney, thanks for your suggestion :-) | 09:24 |
| infinity | mlankhorst: mesa built on powerpc, I call it a success. Screw the other arches. | 09:25 |
| Laney | maclin: Bear in mind beta 1 is quite soon ;-) | 09:25 |
| mlankhorst | infinity: hah I just wanted to look if I could find llvm in the archive before retrying mesa builds:P | 09:26 |
| mlankhorst | other !armhf are finishing up too | 09:27 |
| jbicha | maclin: what I mean is that we announced a call for testing several weeks ago and didn't hear from Kylin that there could be a problem until today | 09:28 |
| cjwatson | (If you've ever read the poem "Paradise Lost", read its first one and a half lines in light of my previous comment; you'll never read it the same way again) | 09:29 |
| infinity | The meter's a bit off. | 09:30 |
| infinity | Or my copy has an extra word yours doesn't. | 09:30 |
| cjwatson | You have to read "Of" on the up-beat, yes. | 09:30 |
| infinity | It's certainly catchy. | 09:31 |
| maclin | jbicha, sorry to say that we miss the testing call...but we will catch up... | 09:33 |
| infinity | mlankhorst: I'm going to try to get some sleep. I leave the armhf mesa retry to you. | 09:43 |
| mlankhorst | sure | 09:43 |
| xnox | please hint upstart to go past mysql-5.5 adt failure ( 5.5.32-0ubuntu3 ) mysql never passed. So it's not a regression that upstart is causing. | 09:44 |
| * xnox did ping #-server about fixing mysql adt. | 09:44 | |
| sil2100 | It seems I'm not really knowledgable about this, but can anyone tell me why libcolumbus cannot leave -proposed? http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#libcolumbus doesn't list any particular problems | 10:37 |
| xnox | sil2100: please look in both reports listed at http://people.canonical.com/~ubuntu-archive/proposed-migration/ | 10:37 |
| xnox | sil2100: trying: libcolumbus | 10:38 |
| xnox | skipped: libcolumbus (16 <- 53) | 10:38 |
| xnox | got: 88+0: i-88 | 10:38 |
| xnox | * i386: unity-lens-applications | 10:38 |
| xnox | sil2100: it causes unity-lens-applications to be uninstallable. | 10:38 |
| xnox | (on i386) | 10:38 |
| sil2100 | xnox: thanks for clearing that up! Didn't really know how to interpret this output | 10:39 |
| xnox | sil2100: yeah, it's criptic, but one gets used to it. Usually search for package name in question, as low down the list as possible. As it could be tried a few times, in combination with other packages. In the end it will list what breaks / is uninstable. From that point, one can open e.g. saucy+saucy-proposed chroot to see for one self why combinations of a&b fail to install together. | 10:40 |
| sil2100 | xnox: makes sense now, thanks ;) | 10:41 |
| xnox | sil2100: in this case it's easy, unity-lens-applications still depends on libcolumbus0-0 instead of libcolumbus1 | 10:42 |
| cjwatson | sil2100: See also https://wiki.ubuntu.com/ProposedMigration | 10:46 |
| mlankhorst | yay, armhf mesa build started | 11:18 |
| xnox | can mysql-5.5 adt failure be hinted through for upstart? a new mysql which fixes it has been upload, but it will take at least 3 hours to build & it would be great to have upstart though to release pocket before FF & Beta1 migration block. | 13:40 |
| xnox | Laney: ^ | 13:40 |
| Laney | xnox: I'd rather see if it works properly | 13:44 |
| Laney | if necessary I'll unblock upstart | 13:44 |
| rbasak | I'm not convinced mysql adt failures were really blocking anything | 13:44 |
| rbasak | Since they've never passed | 13:44 |
| Laney | that doesn't matter | 13:45 |
| cjwatson | autopkgtests in proposed-migration aren't a ratchet | 13:45 |
| cjwatson | They must pass or they block | 13:45 |
| rbasak | eg. I got 5.5.32-0ubuntu3 migrated despite the failures | 13:45 |
| rbasak | Or did someone ack that manually? | 13:45 |
| cjwatson | That would have been hinted manually then, or else there was a failure in the tools | 13:45 |
| Laney | people asked p-m to skip them | 13:45 |
| rbasak | Ah | 13:45 |
| cjwatson | cjwatson:force-badtest mysql-5.5/5.5.32-0ubuntu1 | 13:45 |
| cjwatson | (older version, but) | 13:45 |
| Laney | Good to see things being fixed rather than ignored | 13:46 |
| xnox | rbasak: failed adt mysql is blocking upstart migration to -release pocket. | 13:46 |
| rbasak | I ignored it because I didn't think it was blocking anything, and therefore not urgent. | 13:48 |
| rbasak | I'm annoyed that a failing dep8 test was written and uploaded and never passed. | 13:48 |
| rbasak | Rather than helping, it feels that uploading tests that fail from the beginning just create extra work. | 13:49 |
| jodh | https://staging.jenkins.qa.ubuntu.com anyone ? :) | 13:50 |
| cjwatson | rbasak: That is certainly true. | 13:51 |
| Laney | rbasak: did you see the review of failures that pitti posted to ubuntu-devel some time ago? | 13:52 |
| Laney | there's a couple of maas ones that fail too | 13:52 |
| rbasak | Laney: I did. But none of those are related to any work I've done, and I didn't see them as a priority to pick up over the work I'm already doing. | 13:53 |
| rbasak | When I touch a package I generally try and fix anything outstanding with it. mysql was an exception because I was reverting yet another problem that didn't get tested before upload. | 13:54 |
| === Ursinha is now known as Ursinha-afk | ||
| === Ursinha-afk is now known as Ursinha | ||
| xnox | rbasak: sure. but failing mysql-5.5 test will block migrating _all_ packages from -proposed to -release which are listed in Depends:*. Thus a failing test adt test, always blocks migrations. Unless it's a package with no Depends. | 13:58 |
| rbasak | xnox: right. I understand that now. | 13:58 |
| xnox | ok. | 13:58 |
| Laney | halp | 14:46 |
| Laney | I can't reproduce the problems p-m has with gst-plugins-{good,bad}1.0 | 14:46 |
| Laney | maybe they want hinting? there are mutually updated conflicts/replaces | 14:52 |
| cjwatson | I haven't looked, but in general mutual conflicts won't get autohinted | 14:53 |
| cjwatson | so you probably want to "easy" the relevant set of packages | 14:53 |
| Laney | Worth a punt | 14:53 |
| cjwatson | autohinting basically works off Depends as I understand it | 14:53 |
| Laney | cjwatson: yeah that worked | 15:16 |
| cjwatson | cool | 15:17 |
| === doko_ is now known as doko | ||
| smartboyhw | skaet, that list should be correct, although I'm thinking that some people might be good to be included too.... | 15:50 |
| skaet | smartboyhw, if there are other names that should be included, please respond to the email with them. :-) Best if we have backups identified, etc. | 15:52 |
| smartboyhw | skaet, um, I might get killed for over-interfering if I post them... | 15:53 |
| smartboyhw | skaet, I can private message you, and you can ask the flavour leads | 15:53 |
| smartboyhw | :) | 15:53 |
| skaet | smartboyhw, what ever works. :-) | 15:54 |
| === exekias_ is now known as exekias | ||
| === alecu is now known as alecu_lunch | ||
| === alecu_lunch is now known as alecu | ||
| bdmurray | Could an SRU team member have a look at my ubuntu-release-upgrader uploads? | 20:53 |
| slangasek | bdmurray: looking | 21:49 |
| bdmurray | slangasek: thanks | 21:50 |
| === marrusl is now known as marrusl_afk | ||
| === marrusl_afk is now known as marrusl | ||
| phillw | Riddell: does kubuntu have ppc tester(s) yet, if not - would you like me include kubuntu in the call for PPC testers on Monday? | 22:21 |
| === Ursinha is now known as Ursinha-afk | ||
| === Ursinha-afk is now known as Ursinha | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
