15:07:45 #startmeeting kde-sig -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2011-12-20 15:07:45 Meeting started Tue Dec 20 15:07:45 2011 UTC. The chair is rdieter. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:07:45 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:07:51 #meetingname kde-sig 15:07:51 The meeting name has been set to 'kde-sig' 15:08:07 #topic roll call 15:08:15 hi ho, who's present today? 15:08:23 * rnovacek is here 15:09:04 than: ping 15:09:23 rnovacek: might be the quiet and quick meeting (yay) 15:09:24 Present. 15:09:30 present 15:09:35 #info rnovacek than Kevin_Kofler rdieter present 15:09:39 #chair rnovacek than Kevin_Kofler 15:09:39 Current chairs: Kevin_Kofler rdieter rnovacek than 15:10:34 #topic agenda 15:10:44 I added qt-4.8.0/kde-4.7.4 updates to agenda initially 15:10:50 anything else we should discuss today? 15:11:28 rdieter: perhaps kde splits state 15:12:03 btw: 4.8rc1 should be tagged today 15:13:54 rnovacek: yay 15:14:00 well, let's get started 15:14:14 #topic qt-4.8.0/kde-4.7.4 updates 15:14:53 in short, the qt/kde stuff in updates-testing is ready for stable, I think 15:15:08 qt still has the nagging issue of what to do with libicu support, though 15:15:46 nice, +8 karma 15:15:55 currently, qt tries to dlopen libicu at runtime, and gives some qWarning output on failure 15:16:12 rdieter: i think it's time to push kde-4.7.4 in stable 15:16:37 #agreed kde-4.7.4 to be queue'd for stable updates today 15:17:54 wrt qt, before queuing for stable, options include: 1 do nothing, push as-is, 2. revert adding icu support, 3. silence the qWarning, 4. add a hard dep on libicu, 5. something else better? 15:18:24 opinions? 15:18:30 IMHO, 4. is cleanest, but I'm worried about the impact on future (F17+) live image sizes. 15:18:44 So 3. might be the best solution after all. 15:18:50 right, libicu is non-trivial size-wise 15:19:09 I'm for 3. and adding to @kde-desktop 15:20:22 ok, I'm going to be a little naughty here, suggest we do both 1 and 3. 1 push what we have (0.29.rc1) to stable, 3. build qt-4.8.0 (final) silencing the qWarning and send that to updates-testing asap 15:20:41 3 is ok with me 15:20:52 rnovacek: and +1 for comps 15:21:07 rdieter: I think that's OK. 15:21:36 rdieter: I agree 15:22:05 I can do the qt patching, I'm familiar with looking at the qlocale stuff already 15:22:36 OK, great. 15:22:56 #action rdieter to queue qt-4.8.0-0.29.rc1 for stable updates, prep new qt-4.8.0 (final) update with silenced libicu qWarning 15:23:18 anything else wrt kde-4.7.4 or qt-4.8.0 updates? 15:24:29 #topic kde-4.8 pkg split status update 15:24:32 ok, moving on... 15:24:58 https://bugzilla.redhat.com/showdependencytree.cgi?id=kde-4.8&hide_resolved=1 15:25:17 according to this tracker, we've only a handful of items remaining 15:25:34 I can do them tomorrow 15:25:37 most of kdeutils is done and in rawhide, minus kde-printer-applet 15:25:54 jovie review is done, just not closed (or imported?) yet 15:26:33 I suppose there's a handful of bindings-related ones to do yet too, which I may or may not get to. 15:26:41 What about all the other accessibility stuff? 15:26:46 Is that already in? 15:27:10 Kevin_Kofler: I'll have to double-check honestly, but I *think* so (minus jovie) 15:27:26 and I'll work on a kdeaccessiblity metapackage for upgrade path 15:27:33 Re the bindings, that stuff should have been done 6 months ago. :-/ 15:27:37 similar to how we've handled pkg spilts in the past 15:27:56 I KNEW it would end up that way when you decided to just go with the split packaging without having everything packaged. 15:27:58 Kevin_Kofler: you're welcome to work on it. :) 15:28:16 We should have completed the monolithic multi-tarball SRPM instead, I even offered to help with that. 15:28:19 I'd argue if someone wants it and is interested in it, they can help work on it. 15:28:23 It'd just have meant some CMake patchery. 15:28:50 you're willing to work on monolithic hackery, but not split packaging? 15:29:07 I think it would have been the better long-term solution. 15:29:14 *shrug*, we've already had that discussion 15:29:23 One small patch to maintain instead of a bazillion of specfiles to update at every single release. 15:29:43 Kevin_Kofler: I've scripted most of it now. it's not too bad 15:29:44 But now that basically everything's split, of course it doesn't make sense to unsplit only kdebindings. 15:30:00 (also considering that it was all subpackaged anyway) 15:30:08 while we're at it, I guess we've hinted at renaming a few modules to match upstream, like 15:30:16 kde-baseapps, kde-runtime, kde-workspace 15:30:22 Right. 15:30:36 anything else beyond that? 15:31:04 * rdieter checks 15:31:16 Maybe I can get some split kdebindings first, but I want to fight with a parted2 compat package first, to unbreak kde-partitionmanager and qtparted. 15:31:31 oh, we're purposely using kde-printer-applet vs (upstream) printer-applet 15:32:00 The problem there being that the parted packaging is black magic, using git rather than %patch to apply patches, among other weird things. 15:32:05 Kevin_Kofler: ok, I can do the bindings, the parted stuff seems better suited to you 15:32:27 There are also tons of patches, even on 3.0, worse on 2.x. 15:32:32 It's going to take a while to sort out. 15:32:38 the only other modules that don't match is: oxygen-icon-theme vs oxygen-icons 15:32:41 (That's why I haven't managed to do it yet.) 15:32:53 and PyKDE4 vs pykde4 15:33:14 the icons would should probably keep as-is, imo 15:33:28 and I'm loathe to re-review just for the case-change 15:34:04 PyKDE4 is at least similar to PyQt4 15:35:06 OTOH, uppercase sucks… 15:35:12 (in package names) 15:35:21 anyone able/willing to work on kde-* rename reviews? I can do it, but I've already got 3-4 things to do today 15:35:26 Maybe we should use pyqt4 too. 15:35:29 imo we should rename oxygen-icon-theme to oxygen-icons to match upstreaem too 15:35:46 than: The problem is that all the other icon themes use *-icon-theme. 15:35:56 It's a Fedora naming convention, why does Oxygen need to be different? 15:36:39 I think we should prefer Fedora conventions before KDE conventions 15:36:50 so stick with -icon-theme 15:36:52 I have mixed feelings wrt oxygen 15:37:14 Kevin_Kofler: it's good question, i prefer upstream naming 15:38:13 regardless, we can and should add a Provides: oxygen-icons , to handle the case if anyone expects to use the upstream name, at least. 15:39:24 Another good question would be why the upstream tarball is named just "-icons". Aren't the other popular icon themes also using -icon-theme naming upstream? 15:39:44 (I'm not convinced we can get them to rename though…) 15:40:00 honestly, I don't think all that much thought went into the project/git module names kde uses, esp how it would impact packaging 15:40:03 Kevin_Kofler: we should rename pyKDE4 to python-kde4 if we prefer fedora convention 15:40:06 so... :) 15:40:53 than: there's a packaging guideline exception for python pkgs with a py* prefix 15:42:07 * than is checking... 15:42:40 than: https://fedoraproject.org/wiki/Packaging/NamingGuidelines#Addon_Packages_.28python_modules.29 15:42:49 * rdieter doesn't care strongly, less exceptions for pkg name != tarball name makes my scripting updates all that easier. :) 15:42:55 "If the upstream source has "py" (or "Py") in its name, you can use that name for the package. So, for example, pygtk is acceptable. " 15:43:49 ok, then it's fine with pykde4 :) 15:44:43 ok, let's put it another way, anyone object to using pykde4 ? (provided someone is willing to work on the new renamed pkg) 15:45:10 No objection. 15:45:16 * rdieter has no objection, but someone else needs to do the work. :) 15:45:51 rdieter: no objection, i can take care of it 15:45:58 okie dokie 15:46:05 no objection from me either 15:46:08 #action than to work on pykde4 pkg rename 15:46:32 #action rdieter to work on kde-baseapps, kde-runtime, kde-workspace renames 15:46:37 and how about PyQt4? 15:46:39 how about PyQt4? 15:46:55 imo we should rename it too 15:47:08 no opinion, in that case, upstream is CamelCase too 15:47:46 it would be better to at least add Provides: pyqt4 15:47:52 camelcased pkg names are a bit silly these days 15:47:56 rnovacek: sure 15:48:17 than: if you want to do pyqt4 while you're at it, that's fine with me too 15:48:22 I guess the recommendation is to keep the SillyCamelCaseName when upstream uses it. 15:48:44 I don't like those SillyCamelCaseNames though… :-/ 15:50:22 ok, we keep the upsstream name and just add provides: pyqt4 15:50:41 Kevin_Kofler: +1, i hate SillyCamelCaseName too 15:50:58 ok, I think that covers all the kde-4.8 split packaging 15:51:08 #topic open discussion 15:51:13 anything else for today? 15:52:45 rdieter: Why did you queue only the F16 Konversation to stable and not the F15 one too? 15:52:59 Kevin_Kofler: bodhi rules for f15 are different 15:53:00 I have a question. Where I can get kde tarballs after tagging and before release? 15:53:05 ie, it wouldn't let me 15:53:09 WTF… 15:53:28 rdieter: ftpmaster.kde.org 15:53:29 rnovacek: i'll show you after meeting 15:53:32 rnovacek: ftpmaster.kde.org 15:53:48 rdieter: ok 15:54:10 You need your SSH public key registered for a user account there though, we can add it to ftpfedora. 15:54:18 That server is intentionally NOT public. 15:54:33 It's packager-only SFTP/SCP access. 15:55:30 ok, let's wrap things up, thanks everyone! 15:55:32 #endmeeting