| slangasek | so, some octave challenges left I see | 03:11 |
|---|---|---|
| slangasek | giving back a number of packages on armhf now that octave is available there | 03:51 |
| darkxst | infinity, any ideas why gnome-control-center is stuck in proposed? update_output, seems to pin it on gnome-user-share from what I can tell, but that was updated for bluez5 transition | 10:31 |
| infinity | darkxst: Looks like vorlon hinted it and it's in. | 11:52 |
| darkxst | infinity, ok cool, that was a messy transition! | 11:57 |
| darkxst | guess its due to general lack of deps for dbus stuff | 12:00 |
| infinity | darkxst: That was a tiny transition. The messy one is still in progress. :P | 12:00 |
| darkxst | infinity, messy in the sense that most of it migrated at will | 12:01 |
| darkxst | in bits and pieces | 12:01 |
| infinity | darkxst: Yeah, true that. It would have been less bad if it weren't tied up in all the other things ongoing right now, since we could have done it in a block and it would have just worked. | 12:02 |
| infinity | But, such is life. | 12:02 |
| infinity | Hopefully we won't break C++ ABI for another half decade or more. :/ | 12:02 |
| darkxst | hope so, thinking there will be a lot of FFe requests this cycle! | 12:02 |
| darkxst | and really the only things in the GNOME stack to use c++ are spidermonkey and webkit | 12:05 |
| slangasek | infinity, darkxst: yes, the autohinter was insufficiently clever in this case; I don't know why but probably related to gnome-user-share /not/ being listed in p-m as having a dependency on the new gnome-bluetooth | 12:41 |
| darkxst | slangasek, I guess that is life, dealing with antique packages, all the bluetooth stuff had moved into gnome-bluetooth itself for 3.16, but we are stuck on 3.0 due to UI changes | 12:47 |
| cyphermox | slangasek: confirming the image boots fine now, so syslinux issue is resolved. | 14:09 |
| infinity | cyphermox: \o/ | 14:33 |
| slangasek | cyphermox: good-o, thanks | 14:47 |
| Laney | Can someone clean up the NBS libcpprest2.4 in wily-proposed? | 15:29 |
| infinity | Laney: Yep. | 16:35 |
| infinity | Laney: Except, what NBS is that? | 16:36 |
| infinity | Oh, only on one arch, so it looked like a build failure at first. | 16:36 |
| infinity | Which is why I didn't get it in my last pass. | 16:37 |
| Laney | Cheers duck | 16:41 |
| xnox | i am confused why boost-defaults is not migrating. | 17:28 |
| xnox | Trying easy from autohinter: boost-defaults/1.58.0.0ubuntu1 libftdi/0.20-3ubuntu1 libftdi1/1.2-4 | 17:28 |
| xnox | leading: boost-defaults,libftdi,libftdi1 | 17:28 |
| xnox | somehow fails, yet in -proposed chroot they do install... | 17:28 |
| xnox | also | 17:28 |
| xnox | https://launchpad.net/ubuntu/+source/libftdi | 17:28 |
| xnox | & | 17:28 |
| xnox | https://launchpad.net/ubuntu/+source/libftdi1 | 17:29 |
| xnox | look conflicting to me.... | 17:29 |
| xnox | slangasek: infinity: why was libftdi forked from libftdi1 package, and why do we have both? | 17:29 |
| xnox | i guess i'm missing something... | 17:30 |
| infinity | xnox: Confusingly, they generate two different versions of the library. | 17:36 |
| infinity | xnox: But the failure to migrate is because of the packages they make uninstallable. | 17:37 |
| infinity | * amd64: ubuntu-sdk, ubuntu-sdk-libs-dev | 17:37 |
| infinity | Which, I assume, still depend on the old unnamed sovers. | 17:37 |
| infinity | I can fix that up after I get out of this video game. :P | 17:37 |
| infinity | (Yay, weekends) | 17:37 |
| xnox | infinity: yes, but they generate the same name packages, imho that's wrong. | 17:37 |
| xnox | (binaries that is) | 17:38 |
| infinity | xnox: No they don't. | 17:38 |
| infinity | foo1 and foo1-2 | 17:38 |
| infinity | (Now foo1v5 and foo1-2v2) | 17:38 |
| infinity | s/v2/v5/ | 17:38 |
| xnox | infinity: am i wrong?! -> http://paste.ubuntu.com/12089667/ | 17:38 |
| xnox | as in doing it wrong... | 17:39 |
| infinity | apt might be lying to you. | 17:39 |
| infinity | That's certainly not what I see in LP. | 17:39 |
| * xnox fetches source packages, ok. | 17:40 | |
| xnox | yeap they are different. | 17:40 |
| infinity | Try an --only-source on those command lines. apt might be trying clever reverse mapping. | 17:40 |
| infinity | Now, let's see how these sdk packages are generated and what I can do about it. | 17:42 |
| xnox | they are a seed. | 17:43 |
| infinity | Yeah. | 17:43 |
| infinity | Fiddling. | 17:43 |
| xnox | anyway, i'm in chroot, now install all of ftdi, will try to install boost and finally see how/why sdk baffles. | 17:43 |
| infinity | xnox: Keep in mind that britney assumes removal of NBS. | 17:43 |
| infinity | So those apt-get install lines of yours need an explicit libfoo1- libfoo1-2- | 17:44 |
| infinity | Etc. | 17:44 |
| xnox | ok, checking nbs now. | 17:45 |
| infinity | Hrm, this seems to work: | 17:47 |
| infinity | apt-get install ubuntu-sdk ubuntu-sdk-libs-dev libftdipp1- libftdipp1-2- | 17:47 |
| * infinity scratches his head. | 17:47 | |
| infinity | I wonder if something else needed for this isn't actually migrating yet. | 17:47 |
| xnox | sigh http://paste.ubuntu.com/12089732/ | 17:48 |
| xnox | no, that's C.... | 17:49 |
| infinity | YEah, that's fine. | 17:49 |
| xnox | could britney be confused about the dual/clashing name libftdi1 thing? | 17:50 |
| infinity | Probably not. | 17:50 |
| infinity | And ubuntu-sdk doesn't depend on any of those. | 17:50 |
| infinity | Not directly, anyway. | 17:50 |
| infinity | Did boost-defaults drop any packages? | 17:51 |
| * xnox checks | 17:51 | |
| xnox | i didn't prepare it, doko did. | 17:51 |
| infinity | FWIW, there are no package name clashes, you were just victim to apt's binary mapping. | 17:53 |
| infinity | Which britney doesn't use. | 17:53 |
| xnox | ack. | 17:53 |
| xnox | boost build looks weird and has dbgsyms packages, and previously didn't. | 17:54 |
| infinity | More realistically, this just needs a couple more packages on the hint to make it work, but given that ubuntu-sdk+ubuntu-sdk-libs-dev pulls in 1103 binaries (!), it's not trivial to unpack. | 17:55 |
| xnox | all packages are the same in boost-defaults. | 17:55 |
| xnox | (55 vs 58) | 17:55 |
| infinity | At the very least, that hint needs libjsoncpp with it, looks like. | 17:59 |
| infinity | And jsoncpp is probably tied up in a larger batch. | 17:59 |
| xnox | hm. | 18:03 |
| infinity | ubuntu-touch-meta depending directly on libraries makes it critical path for a lot of hint combinations. I wonder if just removing it entirely for a week would upset anyone. :P | 18:03 |
| xnox | ok, i'll go poke easer things for now. | 18:03 |
| xnox | infinity: i'm sure it has nothing important there. it's not like wily is shipping anywhere. | 18:04 |
| xnox | infinity: the more crap migrates, the better from my point of view =) | 18:04 |
| xnox | has anything migrated so far? | 18:04 |
| infinity | slangasek: Thoughts? | 18:04 |
| infinity | xnox: Some things, here and there. vorlon's been doing a good job of unwinding and by-hand hints. | 18:05 |
| infinity | slangasek: I wasn't being facetious about removing ubuntu-touch-meta. The part where it depends directly on libraries means that it forces all the libraries it depends on to migrate together, which is much harder than doing small transitions. | 18:05 |
| infinity | slangasek: Would anyone lose their mind if we just blew it out of the water for a few days? | 18:06 |
| xnox | demote it to -proposed. | 18:07 |
| xnox | that will also uncover all the random jenkins that use release pocket without proposed =) | 18:07 |
| infinity | xnox: Well, it's already in proposed. But deleting from release is what I meant, yes. | 18:07 |
| xnox | right. | 18:07 |
| infinity | It means image builds will explode. | 18:07 |
| infinity | But by their own admission, the touch guys aren't too focused on wily right now. | 18:08 |
| infinity | I don't really want to get yelled at, though, so I'll let slangasek vote. | 18:08 |
| xnox | infinity: i wish i was an archive admin, lol. | 18:23 |
| xnox | "fixed" octave-communications, was ready to be rebuild for dep-wait. | 18:30 |
| xnox | looking into octave-nan | 18:30 |
| * xnox giggles at the new changelog. | 18:33 | |
| xnox | https://launchpad.net/ubuntu/+source/octave-nan/+changelog | 18:37 |
| xnox | fixing ubuntu-core-meta for libapt-pkg. | 18:52 |
| xnox | how often britney runs? | 18:53 |
| tumbleweed | every publisher run, according to Laney | 18:55 |
| tumbleweed | also, why aren't yuo here? | 18:55 |
| tumbleweed | that applies to both of you | 18:55 |
| infinity | tumbleweed: Debconf conflicted with Plumbers. | 18:58 |
| infinity | xnox: Define fixing. | 18:58 |
| infinity | xnox: If ubuntu-core-meta is refering to libapt-pkg, that's a bug. | 18:58 |
| tumbleweed | infinity: pity. next time | 18:58 |
| xnox | infinity: it's refering to it, seed is updated, but not the package. cause germinate doesn't look like it looks at -proposed, still. | 18:59 |
| infinity | xnox: What seed is this? | 18:59 |
| infinity | xnox: Yeah, see, your fix is wrong. :P | 19:00 |
| xnox | tumbleweed: i have sprained ankle. | 19:00 |
| xnox | infinity: ubuntu-core.wily seed. | 19:00 |
| xnox | last modified by slangasek... | 19:00 |
| infinity | Well, IMO. libapt-* shouldn't be hardcoded in core-libs at all. It's a BS lie, since no one's committing to making that a stable ABI. | 19:01 |
| infinity | Oh well. I'll argue with people about that some other day. | 19:02 |
| xnox | ... | 19:02 |
| infinity | ...? | 19:02 |
| xnox | germinate should be fixed to look into -proposed. | 19:03 |
| xnox | hm. | 19:21 |
| xnox | edubuntu-desktop, kubuntu-desktop, kubuntu-full uninstallable =/ | 19:21 |
| xnox | do we need to transition insighttoolkit4 to v5?! | 19:31 |
| slangasek | xnox, infinity: yes I was bandwidth-constrained so didn't get ubuntu-core-touch uploaded yet. I was prepping it now, have I been superseded? | 19:31 |
| xnox | slangasek: si, senior. | 19:32 |
| slangasek | infinity: ubuntu-touch-meta> yes it is harder but we're almost there; can you give it another day to see where we get? alternatively if we get the big chunk in now and then resolve ubuntu-touch-meta by Monday it's probably about the same | 19:32 |
| xnox | i believe insighttoolkit4 needs v5 treatment. | 19:32 |
| slangasek | xnox: what's this about a sprained ankle? are you with a sprained ankle here or with a sprained ankle somewhere else? | 19:34 |
| xnox | slangasek: i'm in LHR. | 19:34 |
| xnox | this week. | 19:34 |
| xnox | not coming to FRA | 19:34 |
| infinity | slangasek: Does the big chunk not require touch-meta? That was sort of my point. | 19:36 |
| infinity | slangasek: The big chunk might be smaller chunks if there wasn't this one crazy package that depends on a ton of libraries. | 19:36 |
| xnox | hm, fixed all of octave, but they are not migrating =( | 19:36 |
| xnox | hm. failing to see what else to fix. can we hint everything to go together? | 19:42 |
| xnox | error: unknown type name 'BOOST_PP_IIF_BOOST_PP_BOOL_BOOST_PP_COMPL_BOOST_PP_NOT_EQUAL_CHECK_BOOST_PP_NOT_EQUAL_' | 20:32 |
| xnox | oh my | 20:32 |
| xnox | i think clang++ just hangs. | 20:48 |
| slangasek | xnox: spending a week in Heathrow! What crime could possibly fit that punishment? | 21:02 |
| slangasek | xnox: too bad, anyway | 21:02 |
| slangasek | infinity: ubuntu-touch-meta itself is easier to transition if it's first pulled out; but it's not the only metapackage in the transition | 21:03 |
| xnox | slangasek: nah, at home in greenwich. they actually cannot keep me in heathrow,lol. | 21:08 |
| === hyperair is now known as Guest10717 | ||
| xnox | =( gnuradio takes forever to compile. | 22:08 |
| === hypera1r is now known as Guest60557 | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!