=== andyrock is now known as andyrock|zzz === chaoticuk_ is now known as chaoticuk [05:48] thomi, hey, I think I have a fix for all the ibus ap tests! [05:48] oooo! [05:48] Sorry I wasn't online - my INternet crapped ut [05:48] *out [05:49] bschaefer: so, what was the problem, and what's the fix? [05:49] so the addCleanup for the self.deactivate gets executed kinda oddly, executing it explicitly seems to fix it [05:49] :-/ [05:49] and no worries, got back like an hour ago and now im doing some testing to make sure it works [05:50] because what happens is sometimes deactivate gets called AFTER the engines gets reset, so it gets back to false [05:50] sometimes [05:50] I should have a branch up soon ish [05:50] hmmmmm [05:51] that shouldn't happen, but if it helps I'm happy [05:52] hmm one sec i need to relog [05:53] hey, how do i find what unity-2d version i am currently using? [05:54] thomi, ok im push a branch right now, [05:54] would you be able to run them? [05:55] bschaefer: sure [05:55] sweet, I just want to double check [05:55] jokerdino: apt-cache policy unity-2d [05:55] jokerdino: or, if you prefer: dpkg -l unity-2d [05:55] https://code.launchpad.net/~brandontschaefer/unity/ibus-ap-fixes [05:55] thanks thomi. [05:56] thomi, also another thing we can change is the im context getting reset so often, as that really doesn't ever happen that often [05:56] only have it reset when we switch from pinyin to anthy [05:56] bschaefer: I actually looked into that [05:56] it's tricker than I thought [05:56] hmm really? I was going to start looking into that [05:57] we could have an self.active_engine for the IBus class [05:57] and then check what engine is in the current list of active engines (should only be 1) [05:57] bschaefer: yeah, and where are you going to call it from? Remember that after the tests are finished you need to set ibus back to how it was before the tests started [05:58] can't you edit the activate_input_engine_or_skip function [05:58] that's the hard part - we need to restore the system to the same state we found it in [05:58] to check [05:58] o [05:58] yeah [05:58] bschaefer: :) [05:58] there may be a way... hmmmm [05:58] * thomi checks something [05:59] got it [05:59] awesome! What did you do? [05:59] bschaefer: don't call set_engine in 'setUp', call it in 'setUpClass' [05:59] that second method only gets called once for each test class. [05:59] haha nice [06:00] that is a good way :) [06:00] there's also tearDownClass [06:00] bschaefer: you want to add that to your branch as well? [06:00] yeah I can [06:02] is the function actually called setUpClass? or just setUp for each pinyin,anthy and hangul? [06:05] no, it's really called setUpClass [06:05] setUp gets called before each every test method [06:05] setUpClass gets called once before running any tests in a particular class [06:05] tearDown and tearDownClass follow a similar pattern, except obviously they happen after tests are run, not before [06:05] o nice [06:06] I don't think there's any equivilent of 'addCleanup' for the whole class.... that seems like an oversight to me. [06:06] TypeError: unbound method setUpClass() must be called with IBusTestsPinyin instance as first argument (got nothing instead) [06:07] huh? [06:07] I have the def setUpClass [06:07] I have the def setUpClass(self): [06:07] and it gives me that error :( [06:07] * thomi tries quickly [06:08] ahh, it's a classmethod [06:08] so you need to put @classmethod [06:08] on the line before it [06:08] and also, instead of 'self', use 'cls' [06:08] hmm that is weird [06:08] like this: [06:08] @classmethod [06:08] def setUpClass(cls): [06:08] I have it working now :) [06:08] cool [06:09] The only reason I know python is because of this ap test haha [06:09] so I havn't heard of classmethod before haha [06:09] it's like a static method, except you get the class object passed to the function [06:09] there's also @staticmethod BTW, and that doesn't get passed anything [06:10] hmm so if it is static how am i going to call another funciton? [06:10] function* [06:10] with out self [06:10] ie. to call activate_input_engine_or_skip("Pinyin") [06:10] bschaefer: you can't, unless it's also static or classmethod [06:10] bschaefer: you may have to make that a module-level function [06:11] * bschaefer looks that up [06:11] or make it a classmethod as well - I'm not sure if it stores anything? [06:12] brb, gotta get some dinner going [06:12] alright, Ill mess around with this :) [06:12] fun learning [06:27] bschaefer: yep :) [06:31] hmm it's having problems using skipTest, also the _old_engines is also use hmm [06:32] you were right when this was not as straight forward as it seems haha, I should be able to figure something out! back to digging [07:02] thomi, im reading this: http://docs.python.org/dev/library/unittest.html#class-and-module-fixtures. [07:02] and im trying to figure out how to correctly [07:02] skip a test [07:03] * thomi looks [07:03] in a methodclass, im trying to figure out how to raise the exception as a SkipTest [07:03] instead of an assert fail [07:05] bschaefer: so you want to skip a test from the setUpClass method? [07:05] yeah [07:05] well [07:06] activate_input_engine_or_skip [07:06] would be nice, but I can have activate_input_engine_or_skip return true and false [07:06] to have it done in the setUpClass method [07:07] thomi, let me pastebin what I have so far, as it is working but I need to figure out how to skip it correctly [07:08] bschaefer: I don't think you want skipTest [07:08] * thomi looks at how setUpTest is called [07:08] thomi, http://paste.ubuntu.com/937925/ [07:09] I cut out the other hangul anthy test for readablity [07:09] so im still able to save the _old_engines around correctly and reset them [07:10] sure, as a class-level attribute. you just need a way to skip the tests [07:10] yeah [07:10] hmmm [07:11] it says if the exception is a SkipTest then it will skip it in the setUpClass but im not sure how to raise a SkipTest expectation [07:11] oh ok [07:11] ipython [07:11] exception* [07:12] yeah [07:12] in a classmethod, I see how to do it regularly [07:12] ok, let me do some digging [07:13] alright, Ill do the same :) [07:13] got it: [07:13] raise TestCase.skipException("reason to skip") [07:14] sweet, let me try that :) [07:14] I also get a weird ibus deamon error though doing stuff this way :( [07:15] (ibus-daemon:5488): GLib-GIO-WARNING **: unknown control message type 1:2 [07:15] hmmm, dunno what that's all about [07:16] im not sure what is causing it and a quick google search doesn't give much but I can look into that once I have this ibus test working [07:16] but it's only a WARNING, right? We can probably ignore it :P [07:16] it doesn't seem to fix anything [07:16] i mean mess anything up [07:16] idk why I said fix haha [07:18] thomi, NameError: global name 'TestCase' is not defined, ill start looking for the import [07:18] but you might now it off the top of your head [07:18] bschaefer: sorry, you should be able to use AutopilotTestCase [07:18] I forgot that we subclass it :) [07:19] so raise AutopilotTestCase.TestCase.skipException("reason to skip")? [07:19] no worries :) [07:19] hmm that doesn't work [07:19] yup [07:20] thomi, AttributeError: type object 'AutopilotTestCase' has no attribute 'TestCase' [07:20] raise AutopilotTestCase.skipException("Reason to skip") [07:20] oo [07:20] duh [07:21] I went to __init__.py and just saw the import TestCase haha [07:22] yeah - all our test classes eventually derive from TestCase [07:23] yeah, I was like ugg what am I doing wrong!! Ok let me run this from the start and see if it goes well :) [07:30] thomi, sweet they all passed for me, let me check the diff and ill push the changes [07:30] if you had a chance to run them [07:31] bschaefer: can you ping me when there's a MP up? [07:31] yeah! [07:44] thomi, https://code.launchpad.net/~brandontschaefer/unity/ibus-ap-fixes/+merge/102803 [07:44] * thomi looks [07:45] opps the diff is still updating [07:46] there is a 2325 coming in...ugg this diff is insanely slow... [07:55] thomi, diff is done updating, and I just got done running the ap test again and all of them passed :) [07:56] cool [07:57] bschaefer: why is tearDownCLass in the diff twice? [07:57] opps [07:57] that should be setUpClass [07:58] bschaefer: I thought so. How did it work without that? [07:58] well it wasn't None [07:58] It might have messed up my ibus active list [07:58] but that is all that code really does [07:59] thomi, yeah it messed up my active engine list :(, but push the changes [07:59] ok [07:59] pushed* [08:02] bschaefer: I have a suggestion [08:03] yeah? [08:03] have only one setUpClass method, in the IBus test case base class, and have it use 'cls.engine_name'. [08:03] Then, in each of the derived IBus classes, you can do 'engine_name' = 'pinyin' [08:04] you mean under setUp [08:04] wait.. [08:04] how will it do that for each different engine though? Wont it only run once? [08:04] or does it run Once per subclass [08:05] umm, let me show you :) [08:06] ok [08:11] thomi, hmm that is odd my setClassUp in the base class doesn't want to work [08:12] bschaefer: grab this: lp:~thomir/unity/ibus-ap-fixes [08:12] I had it set up slightly different before to test the ibus engine active list were get put back and thought this way would check for no engines [08:12] * bschaefer grabing [08:12] sorry, it was too hard to explain, that branch shows what I mnean [08:12] THat should work [08:13] alright! Looking [08:14] thomi, that is why I was asking if setUpClass for the base class get called for each sub class [08:14] bschaefer: no, I don't think it does [08:14] I'm not sure though [08:14] so the base class will only set the engine once? [08:15] Yes, I'm pretty sure :) [08:17] thomi, hmm it works, so when it switches from IBusTestPinyin to IBusTestsHangul it calls setUpClass each time [08:18] yes, because it's a new class (it just has the same base class(es)) [08:18] that is what I was trying to ask [08:18] I think I was asking it wrong though haha [08:18] and now that makes sense.. [08:18] hahaa ugg im getting tired [08:19] do you want to propose your branch? === chaoticuk_ is now known as chaoticuk [08:19] thomi, ^ [08:19] sure [08:19] it'll be easier haha [08:20] I'll override your MP [08:20] alright sounds good :) [08:20] sweet, hopefully that will stop all the ibus failing :) [08:20] bschaefer: https://code.launchpad.net/~thomir/unity/ibus-ap-fixes/+merge/102808 [08:20] we may as well both approve it... [08:21] thomi, yeah haha; looks good though :) [08:23] thomi, cool, well im off to bed. Have a good weekend! [08:23] you too [08:23] thanks for sticking around so late on your Friday :P === rye_ is now known as rye === andyrock|zzz is now known as andyrock === MacSlow is now known as MacSlow|lunch [11:46] AlanBell, can you try this branch? https://code.launchpad.net/~andyrock/unity/fix-949448 [11:47] it should unmute the screen-reader [11:57] andyrock: how do I do that? [12:00] AlanBell, i'll try to create a ppa === greyback is now known as greyback|lunch [12:05] AlanBell, are you on Precise right? [12:08] Hello there I would first like to take the time to say thanks for reading this and all that you all have done. Here is my question. Will the unity doc ever be able to be moved to say the right hand part of the screen or the bottom ect. I know that this would not happen over night but just wondering if there is a shot at all in the next 5 to 10 years? Thanks agaon [12:12] andyrock: yes, I have done a bzr branch of your branch [12:12] andyrock: do I have to build all of unity? [12:13] ppa would be great [12:14] AlanBell, I am looking at his files and that would be a hard build [12:15] so many depends and postinst scripts would be crazy also [12:16] they would all have to be in different packages that is alot work [12:17] AlanBell, check to see what is installed and read the INSTALL file [12:18] make sure depedencys are not gettig mashed togeather so to say [12:32] AlanBell, not sure but it should work https://code.launchpad.net/~andyrock/+archive/unity-a11y === _salem is now known as salem_ [12:47] thanks andyrock, installing it now [12:48] bobweaver: the launcher goes on the left by design, except for right to left languages where it does go on the right [12:48] bobweaver: it would be great if it was on the left on the left screen and on the right on the right screen for dual monitor setups, but I don't think that will happen [12:48] Yes I [12:49] that would be awesome ! [12:49] andyrock: looks like it hasn't finished building in the ppa yet [12:49] AlanBell, is there a place where I can find bluprints or like work tables [12:50] I looked on LP but there is nothing that I can see [12:50] http://unity.ubuntu.com/getinvolved/ [12:50] https://wiki.ubuntu.com/Unity [12:50] Thanks === MacSlow|lunch is now known as MacSlow === greyback|lunch is now known as greyback [12:58] AlanBell, try now [13:01] already installing it [13:16] right, lets see what happens with a unity --replace . . . [13:19] hmm, not much different, lets try restarting [13:19] AlanBell, it's not perfect but now the screen reader should not be mute [13:27] andyrock: yup it is speaking a bit \o/ [13:28] it doesn't read the application names in the dash [13:28] or lens names [13:29] AlanBell, regarding the application names in the dash... [13:29] it works intermittently [13:30] HUD is about the same as it is in unity2d, it speaks that you are in the HUD pane, but doesn't speak the options that it is presenting [13:30] maybe gord will leap into action to fix that :) [13:31] Hud is a massive improvement it is awesome I use all the time [13:31] generally this seems nearly as good as it was just before beta 1 [13:31] AlanBell, yeah I just fixed the regressions... [13:31] bobweaver: I am guessing you have functional eyes [13:32] @_@ [13:32] now we just need a couple of more months before the release to get the smaller bugs fixed :) [13:33] a couple more months would be nice. [13:33] AlanBell: you've got 6.5 months, how much more do you want? [13:33] i wonder who is the master enbugger of unity. [13:34] I just hope we will get a full cycle with accessibility actually not totally broken next time [13:38] but this is really great that unity3d is nearly back to the same level as 2d, thanks very much andyrock [13:39] AlanBell, I'll try to fix the other bugs too... [13:39] andyrock: will this go into the unity team PPA for wider testing? [13:39] AlanBell, I'll push a MP sooner [13:39] *soon [13:40] then we should fix the dash reader... [13:44] mhall119: what was the bzr link for hello-unity please I lost the post with it in [13:45] lp:hello-unity :) [13:45] davmor2: are you going to break it? [13:45] mhall119: oh nice I thought it might of been in your working ppa obviously not :) [13:45] mhall119: possibly [13:46] it is in my PPA too: https://launchpad.net/~mhall119/+archive/ppa [13:46] hello eveyrone! I just filled bug #986191, I was wondering if it was known or workaroundeable [13:46] Launchpad bug 986191 in unity (Ubuntu) "Dash is always empty" [Undecided,New] https://launchpad.net/bugs/986191 [13:47] nessita: all the lenses are empty? [13:47] or just the home lens? [13:47] mhall119: yes. Music and video I expect to be empty, I uninstalled those lens. I also confirmed that the app and files lens is installed and up to date [13:47] mhall119: apps lens is definitely empty as well [13:48] how did you uninstall the music and video lenses? [13:48] greyback: ^^ any ideas? Looks to be unity-2d [13:48] mhall119: I can't remember if they ever worked for me. This happens the same in both my laptop and home computer (both unity-2d) [13:48] AlanBell: sudo apt-get purge unity-lens-gwibber unity-lens-video unity-lens-music [13:48] nessita: did you restart unity or log out/in? [13:49] I still see icons for the music and video lenses in your screenshot [13:49] mhall119: several times, every day. I can try again if you consider is worth it [13:49] weird, it works for me. [13:49] nessita: if you do a search, do you see options? [13:49] nessita: it's just odd, they shouldn't appear in the dash after you uninstalled them... [13:49] mhall119: perhaps that's because I haven't re-logged in after uninstalling those [13:49] greyback: nopes [13:49] oh, nessita try logging out and back in then, see if they go away [13:49] nessita: yeah, that's not right [13:50] mhall119: sure. But that is not related to not having anything in the rest of the lens, right? [13:50] please try as mhall119 asks, then can investigate further if necessary [13:50] greyback: of course. Will be back then! [13:50] nessita: fingers Xed :) [13:50] nessita: I don't know, but it's best to eliminate any other possibilities [13:51] greyback: does unity-2d respond to changes in /usr/share/unity/lenses/ ? I thought I heard that it did [13:51] I know unity-3d requires a restart [13:52] mhall119: We need restart too. Really when it comes to lenses, we just use the data that Unity-Core supplies, which is shared between both 2d and 3d [13:53] hello again! [13:53] wb nessita [13:53] so, confirmed music and video icons are gone, and dash is still empty [13:53] what's the verdict? [13:54] search returns nothings [13:54] can share a screenshot [13:55] nessita: run "ps -ef |grep unity-lens-applications" and see if there's a process running [13:55] screenshot of search attemp is: http://ubuntuone.com/6Kyf0uK9QIPJ1NBAUU1pQ2 [13:55] the app "deluge" is certainly installed, as you can see the blue drop icon in the launcher [13:56] mhall119: 160:nessita 1926 1 0 10:52 ? 00:00:00 /usr/lib/unity-lens-applications/unity-applications-daemon [13:56] ignore the first 160:, is the line number [13:56] (I have an alias for grep to always use -n) [13:56] nessita: kill it and run it manually from the command line, see if it gives any error messages [13:57] mhr3: ^^ any ideas what might be going on here? [13:57] sure [13:57] mhall119: terminal shows several of: [13:57] (process:2339): LibZeitgeist-CRITICAL **: Failed to create proxy for Zeitgeist daemon: Error calling StartServiceByName for org.gnome.zeitgeist.Engine: GDBus.Error:org.freedesktop.DBus.Error.Spawn.ChildExited: Process /usr/bin/zeitgeist-daemon exited with status 1 [13:58] and dash is till empty [13:58] is still* [13:58] nessita, try running zeitgeist-daemon --log-level=debug [13:59] nessita, btw i was looking at that bug like 10minutes ago :) [13:59] http://pastebin.ubuntu.com/938303/ [13:59] mhr3: the requested output ^ :-) [14:00] are you upgrading from lucid? [14:00] mhr3: nopes, this is a clean precise install made by alpha1 and updated every day [14:00] is worth noting that I've never had results in the dash since installation [14:00] but just now I have the time to report it and debug it [14:01] nessita, you must have had old home partition? [14:02] mhr3: well, yes, home partition is not new, but was created on March last year, I installed pre-released natty [14:02] with a clean home back then (computer was brand new) [14:02] hmm, natty had so old zg? [14:03] nessita, and can you run the apps lens from terminal? [14:03] mhr3: no idea about zg version, but as a user I would expect that migration path would do its deed? [14:03] mhr3: sure, what command? [14:04] nessita, it would if you had oneiric at some point :) [14:04] pkill -f unity-applications-daemon [14:04] /usr/lib/unity-lens-applications/unity-applications-daemon [14:05] mhr3: http://pastebin.ubuntu.com/938307/ [14:06] nessita, that's all if you try to search? [14:06] ah, did not try to search, will do [14:07] mhr3: nothing new appears in the terminal [14:07] give it 30seconds [14:07] :) [14:07] although... nah nvm [14:07] 30 seconds for output? [14:07] still nothing [14:07] and nothing :-) [14:08] nessita, anyway, `rm ~/.local/share/zeitgeist/activity.sqlite` will fix it [14:09] so, from an end user POV, is there any way to detect that ZG is not working and do some other (perhaps less efficient) search? [14:09] all this time has been very frustating for me to open the dash, type "firef" and get nothing, knowing that firefox is right there [14:10] nessita, yes, we try in many cases, obviously we missed this one where you didn't follow the official upgrade path [14:11] i mean zg tries to recover [14:11] itself [14:12] ok, I see. In ubuntuone, to upgrade user metadata, we detect the current version, and apply upgraders from the user version up to current... so it really does not matter what ubuntu version the user had [14:12] perhaps something similar can be considered? [14:12] so we go from 0 to 1, then to 2, etc etc [14:13] mhr3: so, if I remove the ZG sqlite, I will loose the 23M of info I have there? [14:13] nessita, yes, we had that as well, we dropped the super old schema upgrades when porting to vala, cause who'd care about the lucid version where zg wasn't even installed by default right... [14:13] nessita, yep [14:13] mhr3: I can confirm I never had lucid on the home in this computer (nor maverick) [14:14] nessita, who knew such old schema was in natty as well :( [14:14] i'd push for porting those as well if i did [14:15] well, I would guess lots of people will move from natty to precise without going thru oneiric... perhaps some workaround can be added for them? [14:16] mhall119: oh dear broked it :D can I not run the ./hello-unity (that has +x permission) from the dir on my system that the bzr branch created? [14:16] nessita, i'm looking why the search doesn't work, it should [14:16] only the recent apps are grabbed from zg [14:16] mhr3: can I help debug further? [14:16] nessita, if you want to build ula? [14:16] mhr3: what's ula? :-) [14:16] lp:unity-lens-applications [14:17] mhr3: if you really need me to, I will be happy to do that next Monday from my home computer (I'm traveling now so I only have my laptop handy) [14:18] mhr3: my home computer is in the same situation (same versions, same symptoms) [14:18] nessita, dont worry then, i can easily lower the schema ver here [14:18] mhr3: ok then, let me know if you need anything else [14:18] I will around (or you can easily find me in the company channel) [14:19] sure [14:19] davmor2: I think you need to copy the .desktop file into /usr/share/applications/ first [14:19] but otherwise it should be runnable from the bzr checkout [14:20] mhall119: fair enough I'll have a play with it over the weekend then :) [14:20] davmor2: have fun :) [14:21] mhall119: will do lets see how bad I break stuff :D erm I mean test obviously [14:23] davmor2: I'm going to invoke the "weekend project" excuse for any quality shortcomings [14:24] mhall119: haha [14:24] but, since most of the code is exposed in the UI, I did make an extra effort to do it right === salem_ is now known as _salem === _salem is now known as salem_ [15:30] andyrock: added the video lens.. http://i.stack.imgur.com/alAPh.jpg (bug #937334) [15:30] Launchpad bug 937334 in Ayatana Design "Unity shortcut overlay needs to include shortcut for video lens" [Low,Triaged] https://launchpad.net/bugs/937334 [15:35] is it my luck that he left the room soon after? :)) [15:41] hey guys. is the unity dash blur effect implemented in glsl? if so, where abouts in the source code could i find it? [16:08] why if I'm using precise and Rhythmbox is the default application for mp3, searching music with the Unity Dash the music is opened with Banshee? [16:09] is it a known bug or I'm doing something wrong? [16:11] and the music lens is not working anymore I think.... clicking on a song it does't open anything more... [16:11] how do I restart the lenses without restarting the whole session? [16:17] Andy80: i wonder if banshee is set to your default music player? [16:18] oh, he's gone [16:59] The new Unity contributor pages are public: http://unity.ubuntu.com/getinvolved/ === racarr_ is now known as racarr === fenris is now known as Guest89753 === Guest89753 is now known as ejat === kenvandine_ is now known as kenvandine [21:31] hmm, there seems to be an issue with the dash blur with the current unity in precise [21:32] the background is incorrectly shifted downwards by some amount, seems to be the height of the panel [21:33] wait, to be more accurate, it's the *lense* background that is wrong, the *dash* background is correct [21:33] it looks pretty broken at the boundary between those, though [21:39] ah, it's https://bugs.launchpad.net/unity/+bug/977922 [21:39] Ubuntu bug 977922 in unity "[regression] Blurred view of the current workspace is shifted down when unity dash is in fullscreen mode" [Medium,Fix committed] [21:39] would be great to see an updated package in precise... [21:40] zgreg, as you can see, the bug is "Fix committed" [21:40] yes, I know [21:41] that does not mean we'll see the fix in precise soon [21:41] zgreg, the update will come as a 0-day SRU, an update available on release day. [21:42] so it's not possible to merge any bugfixes at this state in the release process? [21:43] zgreg, critical bugs only. [21:43] ok === salem_ is now known as _salem === racarr_ is now known as racarr|train