[00:51] When installing a new scope, how do I tell dbus that there is a new scope without having to reboot? (in packaging) [00:57] I thought it was the scope that made the connection? [00:58] forget that. I'm tired. :) [00:59] so the issue is that it installs just fine, but I have to reboot the machine in order for stuff to show up when searching [00:59] indicating that dbus doesn't know it exists [00:59] certainly not reboot? A logout and back in must be sufficient? [01:00] nope [01:00] well, I did a quick test, it might have been searching [01:00] sounds strange. It runs on the session bus as the local user? [01:01] um, sure? [01:02] IDK, it's my first time really playing with dbus stuff [01:02] I'm just creating a scope for the videos lens for MythTV [01:03] which works great, although I had to monkey around with some stuff since it doesn't seem unity supports episodic content too well [01:15] no, I'm not sure. [01:16] tgm4883, what do you mean Unity doesn't support episodic content? [01:17] so pushing things into the videos lens, there isn''t a good way to display episodic content. Basically I had to stuff NameSeasonEpisode into the name field [01:17] jo-erlend: I think he means heirarchical data [01:17] right. [01:17] it works fine for Unity, since it's searching, but will need fixed when UTV rolls out [01:17] Unity results are very flast [01:17] flat [01:18] yes [01:18] the Video lens could easily support Season categories though? [01:19] jo-erlend: you'd have to be able to drill down [01:19] heh, categories are another pet peeve of mine [01:19] what does that mean? [01:19] I'm not sure they would work so well for episodic content [01:19] you mean have a category for "season 1"? [01:19] pick "Show A" and get "Season 1, Season 2, Season 3". Pick "Season 1" and get "Episode Z, Episode Y, Episode X, etc" [01:20] like a directory tree [01:20] you can't do that with flat results [01:20] that would be the usual way of looking at episodic content [01:20] it would show all episodes for Season 1, regardless of series. [01:21] yea that would be kinda strange I think [01:21] that's a hard way to find an episode [01:21] sure. It would work though. Not saying it'd be optimal. :) [01:21] but while we're at the subject.. Is the unity-tv thing available for install yet? [01:22] it wasn't a functional thing [01:22] just a mockup app [01:22] it's completely made up [01:22] it didn't even use Dash, it was just made to look like it [01:22] like europe ;) [01:23] oh. [01:24] Categories aren't great right now as you have to look in the lens source to see whats available [01:24] Damn. I wish Canonical would be better at communicating things like that better. I've made a fool of myself showing this to people. I never got the impression it wasn't real. And now I have to wonder if Ubuntu for Android is also just a mockup? [01:25] If the scope could create categories, that would be better IMO [01:25] tgm4883: the lens author could always provide documentation [01:25] tgm4883: it would be messy though [01:25] mhall119, who is the author of the videos lens? [01:25] mhall119, yea it would [01:25] tgm4883: davidcalle I believe [01:26] yea looks like he is the maintainer [01:26] the good thing about the current categories is that they're always the same, no matter what scopes you install [01:26] installing a scope should just give more results, not change the user experience [01:27] mhall119, I'll agree with that [01:27] a better solution is to fix the categories, there are only 4 right? [01:27] online, local, remote, ??? [01:28] I'd like a recorded/dvr category [01:28] I see "recent", "my videos" and "online" as categories [01:28] I think "recorded" might be a good category [01:28] or "downloaded" [01:29] same thing, different ways of doing it [01:29] yea [01:29] downloaded and recorded [01:29] two new categories [01:30] mhall119, do the categories only exist for videos, or do they also have to work for music as well [01:37] tgm4883: they are per-lens [01:37] so the music lens has it's own set of categories [01:37] ok that makes sense [01:37] then no reason not to add these I say [01:37] why would downloaded and recorded be separate categories? [01:38] the both mean the same thing "Something that was available from somewhere else, and is now available locally" [01:38] perhaps [01:38] but the MythTV stuff isn't local to the machine [01:38] it's still on the backend [01:38] as would any other DVR backend [01:38] local to the network then [01:39] ok, but still not downloaded [01:39] you don't have to fetch it from the internet or broadcast [01:39] downloaded to the backend [01:39] I would consider the backend "local" [01:39] tgm4883: maybe the category should be "Shows" [01:40] that way it can be local or remote [01:40] I think downloaded and recorded are fundamentally different enough to warrant different catagories [01:40] tgm4883: are they different enough from a user's intent? [01:40] I believe so [01:40] actually, I'm not sure you even need downloaded [01:40] do you approach it thinking "Do I want to watch something recorded, or something downloaded?" [01:40] I want to watch a show :) [01:41] then make the category "Shows" [01:41] doesn't matter where they live [01:41] hmm [01:42] perhaps Local is the better spot for it [01:42] I think "Shows" and "Movies" would be good categories [01:42] I agree with that [01:42] Shows and Movies [01:46] mhall119, should I be filing bugs against unity for the videos scope, or is there a better place? [01:50] against the video lens [01:51] https://launchpad.net/unity-lens-videos [01:58] !bug 984507 [01:58] Launchpad bug 984507 in unity-lens-video (Ubuntu) "unity-lens-video needs additional categories" [Undecided,New] https://launchpad.net/bugs/984507 [01:58] thanks mhall119 [01:58] np [04:47] thomi, ping, or more or less to confirm you new changes to the ibus branch [04:47] Ran 32 tests in 270.267s [04:47] OK [04:47] :) [04:48] bschaefer: hey [04:48] I was looking for you [04:48] sorry, was in class [04:48] for 3 hours, now im back haha [04:48] no worries - could I get you to review some branches? [04:48] sure! [04:49] cheers :) [04:51] any branch in particular or all the ap ones you pushed? [04:52] bschaefer: if you have time - all the AP ones [04:53] lots of them were approved, then I pushed more revisions [04:53] alright! Yeah I do, I spent this morning confirming the mem leaks I was looking at the last couple days gone [04:53] so reviewing for a little would be a nice change of pace haha [04:54] Sweet, I need to head out for a bit. WIll be back later though [04:54] talk to you later :) [04:54] alright cool! Ill be running this ap test after I review the code, so it might take a litte haha [04:54] yup [05:41] tgm4883, ping [05:55] davidcalle, pong [05:56] tgm4883, nice scope :) [05:56] thanks [05:56] tgm4883, I've just answered your bug. I think I could push something for you in 12.04.1. [05:57] that would be awesome [06:00] bschaefer: I'm back [06:00] only one branch left - well done! https://code.launchpad.net/~thomir/unity/ibus-test-to-wait_for-feature/+merge/102236 [06:03] thomi, I need to review the hud one, but it crashed, then I saw your branch to fix that haha [06:03] and was about to look at it [06:04] and no problem! These ap test are look nice! [06:04] :) [06:04] but forgot to re approve the ibus test, done [06:04] cheers! [06:05] thomi, wait are you working on the ap hud crash atm? [06:06] thomi, cause I see one that isn't proposed and one branch that is suspended [06:06] bschaefer: I'm not sure what you mean by the hud crash [06:06] thomi: nice work on all the AP stuff! :) [06:06] https://code.launchpad.net/~thomir/unity/ap-hud-crashing [06:06] I didn't get the chance enough to tell you that ;) [06:06] didrocks: thanks - any chance we can get the merge bot fixed? It's filling me with rage :( [06:06] thomi: the pre-dependency issue? [06:06] it doesn't handle prerequisite branches well at all [06:06] yeah [06:07] thomi: do you have a test case? you are the only one triggering it. I have some tests for prerequisite and when both branches are approved, I don't reproduce the issue [06:07] did yeah I know what' shappening [06:07] thomi: so, really interested in knowing what exactly triggers it :) [06:07] thomi: if you can decypher any pattern, I would be interested [06:08] thomi, because I got this crash while running the hud ap test [06:08] http://paste.ubuntu.com/935041/ [06:08] The merge bot doesn't merge a branch while it's prerequisite branch isn't merged (so far so good), but once the prerequisite branch is merged, the MP dissapears, and the merge bot gets stuck because it can't find a MP for the listed prerequisite branch [06:09] bschaefer: running unity trunk? There was a problem, but gord fixed it [06:09] thomi: ah, so you mean, master branch merged, and dependency branch just approved? [06:09] thomi, oo, cool must have not pull that version when I pull your branch [06:09] thomi: if you are interested, we can have a look together at the pre-sprint, wdyt? [06:09] didrocks: that'd be cool. I can probably come up with a test case before then [06:10] thomi: because I tested that case IIRC and didn't trigger the issue either. There is clearly something wrong, we just need to find what :) [06:10] thomi: thanks ;) [06:10] no, thank *you* :) [06:22] thomi, hmm a hud ap test is failing but when I check it manually it passes [06:22] bschaefer: which one? [06:22] i mean manually like I try it and it works [06:22] def test_hud_to_dash_disabled_alt_f1(self) [06:22] AssertionError: True != dbus.Boolean(False, variant_level=1) [06:22] .... that's interesting, how does it fail? [06:22] let me run it on its own [06:23] because the code looks fine... [06:23] timing issue possibly? Maybe we need a wait_for somewhere in an emulator [06:24] oo the dash doesn't close! [06:25] http://paste.ubuntu.com/935053/ [06:25] the whole stack trace [06:25] ahhh [06:26] that's my fauly [06:26] *fault [06:26] you can't use the launcher emulator in that test [06:26] you need to use the keybindings [06:26] since we don't expect it to work [06:26] oo [06:26] can you change it please? [06:26] can I push it on your branch? [06:27] * thomi is eating dinner and watching pycon talks :) [06:27] o haha, yeah, let me look those keybindings up! [06:27] bschaefer: nah, just make a new MP [06:28] ok [06:35] thomi, dammit I just found a bug, un related to the ap tests [06:35] well not dammit [06:35] haha good, but Im surprised it hasn't been found. If you go from hud then to dash then exit you don't get window focus back [06:37] bschaefer: I think there's tests for that [06:38] hmm, really? That one wasn't failing for me... [06:38] thomi: on the "no commit specified", please blame thumper who asked for this check ;) [06:38] the 2 that were failing were because the launcher thing [06:38] thomi, Ill check if there is an ap test for it [06:38] after I push this branch [06:46] thomi, https://code.launchpad.net/~unity-team/unity/hud-tests-to-wait/+merge/102440 (I forgot the _for after wait) [06:46] the diff should be done soon [06:46] * thomi looks [06:47] it saved 2 more lines of code [06:47] http://bazaar.launchpad.net/~unity-team/unity/hud-tests-to-wait/revision/2249 [06:47] easier to see [06:48] bschaefer: so that branch superceeds mine? [06:48] yeah [06:48] thats what you wanted isn't it? [06:48] or do you want to me do just make a side branch? [06:49] yeah, I think I messed that one up haha. Let me make a new branch...and Ill delete that one [06:49] bschaefer: not hat's fine, [06:49] I'm asking if you replaced my MP (you probably should) [06:49] no, I didn [06:49] didn't [06:50] I just put it against unity, because I've never had to merge it to a different back [06:50] ok, if you just click 'resubmit' on my MP for that branch and list your own [06:51] ...as the source branch then we avoid conflicts [06:51] yeah [06:51] I soon as you said that I realized what other open was for in the mp process [06:51] haha [06:53] thomi, wait, it already is the source, and the target branch should be pointed at your branch? [06:53] no... target should be lp:unity [06:53] actually, I'll just delet my MP - much simpler :) [06:53] ok then I was just confused from something, it should be ready already [06:54] cool. [06:55] bschaefer: I just approved it... seems a bit cheeky as it's still mostly my code, but you approved mine, so we're safe :) [06:55] thomi, I was thinking it would be cool if I could just merge a change into your branch you made, so you would approve of any of someone elses changes [06:55] yup : [06:55] bschaefer: we could have done that. [06:55] :) [06:56] o well to late haha, that is what I thought you were asking haha [06:56] but when I tried to put your branch as a target it couldn't find it :( [06:57] ahh well [06:57] well it's all fixed, and I should make dinner (as my I had a very very late lunch) [06:57] as I* [06:58] thomi, do you need any thing else reviewed? [06:58] bschaefer: no thanks, I'm just trying to get everything merged now :) [06:58] nice, don't forget commit messages! [06:58] Ill also check if that ap test there for the hud to dash then window focus [06:59] else Ill report that bug... [06:59] and most likely work on it tomorrow haha. Have a good night! [07:01] cheers, you too! [07:36] thomi: in most of the case, the rejection is because in those case is because the parent branch got rejected (no commit message) and so not approved, rejecting the child one then [07:37] didrocks: yeah I saw those too. [07:37] didrocks: let's look at it in SF [07:37] thomi: but I know there is a wrong case as well, that's the one we need to debug ;) [07:37] in this case, the rejection are "normal" :) [07:38] didrocks: for some of them, yeah. I forgot a bunch of commit messages [07:39] for a few others however, there was a commit message, but it got the other problem. [07:50] davidcalle: this http://pastebin.com/DCN8TC9G gives this error: [07:50] File "/usr/lib/unity-lens-buss/buss", line 82 [07:50] results_list.reverse () [07:50] ^ [07:50] SyntaxError: invalid syntax [07:53] malin, because your line 71 is not correctly closed, it misses a bracket : results_list.append(test(full_url,patjunk) [07:54] malin, and don't reverse the model, it will fail. (l 83) [07:54] ah [07:54] bschaefer: you have conflicts on your hud branch... [07:54] so If I reverse that, it wil reverse the text instead? [07:54] * thomi -> EOD. Talk to you tomorrow maybe [07:55] malin, I don't understand what you mean. [07:56] I could ask another way. What exactely will fail, and how :) [07:56] I added a : to the end of results_ists.append(test(full_url,patjunk): [07:56] so it looks like that, but still it gives syntax error [07:57] ah, okey. I removed the reversing of the model. Indeed that dosen't sounds right :) [07:58] malin, no, the syntax issue you have is because of a missing ")" You need to do "results_lists.append(test(full_url,patjunk))" [07:58] of corse :S [07:58] how could I forget, when I belived I did :p [08:02] davidcalle: http://pastebin.com/NvMyUJKJ [08:05] malin, model = search.props.results_model [08:05] ah, so that one shouldn't be changed? :) [08:06] malin, never ever :) [08:06] oki :) [08:09] in model.append(and things in here) should I change test(full_url,patjunk) with results_list ? [08:09] looks like it don't like it [08:15] malin, if it's a list you have to iterate over it: http://paste.ubuntu.com/935120/ [08:16] ah, so I actually need a for-loop :) I see [08:16] but where do the result comes from, I can't remember having such a variabel [08:16] malin, yeah, because each model.append is a single dash result. [08:17] result is a variable the loop creates for each element in results_list [08:17] yeah, I see. it's made after the for :) [08:17] then I understand [08:17] for number in [1, 2, 3, 4, 5]: [08:17] print number [08:18] yeah, and then it wil print all numbers from one to five [08:18] malin, yep [08:18] in java it would be something like: for (int i = 0; i <= 5; i ++) { [08:18] system.out.println(i) [08:18] } [08:18] ish [08:18] malin, that's why I love Python :) [08:19] I can understand that...... [08:19] I learn java at school [08:20] is it possible to do like this in python? [08:20] for number in number <= 500: [08:20] to get all numbers from zero to 500? [08:20] malin, yes [08:20] nice [08:21] in fact it semce easier, but java is faster [08:21] *looks easier [08:21] i = 0 [08:21] yeah [08:21] while (i < 500): [08:21] i = i + 1 [08:22] print i [08:22] Not exactly the same spirit as java, but yes, it's very easy to read and understand. [08:23] the lense works again, but the results isn't present in reverse order :) [08:24] malin, what do you see when you print results_list ? [08:25] hello, may I poke somebody with compiz knowledge about bug #946388, and especially my last comment with a vala test code which exhibits the same behavior? [08:25] Launchpad bug 946388 in remmina (Ubuntu) "Remmina full-screen mode does not work under Unity" [Undecided,Confirmed] https://launchpad.net/bugs/946388 [08:25] * davidcalle needs coffee and cigarette, back in 10 min [08:26] davidcalle: first it print the first result again [08:27] davidcalle, for x in xrange(500) ;) [08:27] davidcalle: it prints like this: http://paste.ubuntu.com/935133/ [08:33] mhr3, sure, but it was to show the basic structure of loops :) [08:34] rye, duflu here may be able to help :) [08:34] * duflu looks at the bug [08:34] Hi rye [08:35] malin, it's the result of only one search? [08:35] duflu: hello! [08:36] rye: I'm still getting up to speed on the bug, and multitasking :) [08:37] davidcalle: it's the result after two search [08:37] It print the first result twice [08:37] malin: i think you may want to print the whole array first to see whether there are indeed 2 duplicate items [08:38] I do print the arra as far as I know [08:38] I make the array like this: [08:38] results_list = [] [08:38] rye: I encountered a very similar fullscreen bug in geeqie, which is now fixed. Unfortunately the fix was to fix the app in that case. [08:39] then I add results to the array like this: [08:39] results_list.append(test(full_url,patjunk)) [08:39] and I reverse it: [08:39] results_list.reverse () [08:39] and print it: [08:39] print results_lis [08:39] t [08:39] duflu: I don't like the following comment in src/window.cpp of compiz - /* Don't allow maximization or fullscreen of windows which are too big to fit the screen */ - suggesting that if a window + decorations does not fit, it won't be switched to fullscreen mode, even though it is already in fullscreen mode [08:40] rye: Agreed. [08:40] malin, ok. Then replace lines 70 and 71, with results_list = test(full_url,patjunk).split('.Buss ') [08:40] The aforementioned geeqie fix was *before* I joined the compiz team [08:40] * duflu looks [08:41] duflu: i wonder if it works if i ask the wm to disable the decorations... [08:42] malin, it will directly make a list from your result string returned by test. Creating a new item in the list each times it finds ".Buss " [08:42] rye: What is the expected behaviour for oversized fullscreen windows? How does Gnome etc deal with it? Just show part of the window? [08:43] duflu: it is not oversized, it is exactly the size of the screen [08:43] davidcalle: shit... it made an answer for each bus, how did taht happend :D [08:43] duflu: however in gnome the request_size_change does not make the window to stop being fullscreen [08:44] sometimes two different bus routes goes from the same place to the same stop, but different routes [08:44] rye: OK, the bug in window.cpp looks simple [08:44] and the results was splited up in two D: [08:44] :D [08:44] I'll just verify your test case and then it's triaged.... [08:44] duflu: and no, if i disable decorations the window is simply displayed the same broken way but without the decorations [08:45] davidcalle: that looked great. But still, it puts last answer in the end and not in the start [08:45] rye: Yes, it looks likely to be bad dimension calculations in window.cpp [08:47] malin, even with results_list.reverse () ? [08:47] davidcalle: yes, but maybe I have it on wrong line ? [08:48] anyway. Now I really understand the .split thing [08:48] that was very smart [08:48] You should have it between the line where you create the list and the for loop [08:49] ah [08:49] but it is [08:51] malin, hmm. Try to print the list before reverse, and after. Is the list the same ? === chaoticuk_ is now known as chaoticuk [08:54] rye: Thanks. The bug is triaged. Looks like a relatively simple fix. I hope we get it done in the next maintenance update. [08:54] hm : [08:55] it looks the same [08:56] hm, it reverse it another way [08:56] it takes all after split first and the thing before the split first [08:56] It dosen't reverse the array, but the content of the array [08:57] malin, ok, then don't use reverse and use results_list = results_list[::-1] [08:57] malin, it will reverse the array. [09:00] aha [09:04] hm, now it just printed the previous results once more, before it gave the next result. It still puts last answer in the last and not the top [09:04] maybe it is possible to use a stack in stead? [09:05] malin, can you pastebin your code? [09:05] yeah [09:06] http://pastebin.com/4kYiTgrY [09:07] malin, uncomment model.clear () [09:09] aha [09:10] malin, I thought your issue was about several busses results from a single search and that you wanted this list reversed :) [09:10] malin, but in fact your lens wasn't cleared of results from previous searches. [09:12] what I want is: several results, but the last search-result at top, not bottom. if there is a long list with previous results, thats okey [09:13] hi guys, i'm having trouble with a java application in unity, it opens two pixel wide, a vertical ribbon that's impossible to resize [09:13] malin, in the Dash, you can't reorder results once they have been displayed. Results from next searches will only be added after the existing ones. [09:14] it works in unity-2d [09:14] davidcalle: okey, so it is impossible to get it at the top? I see [09:14] but then I it just is the way it is then :) [09:15] malin, yeah, to have a result in the first position, you have to clear the Dash from existing results. [09:16] I see, but In fact I think that's the best solution. There is no good reasons to store an old result, as the result is not valid for long time, as the day goes on [09:17] so I think I will let it be like it is now. Splitting up the results in two, so the different buses appears in different results is smart and I didn't know about it :) === malin is now known as Malinkb [09:18] malin, this way, most of the time, you won't need to click on the result to get the full text. === Malinkb is now known as malin [09:19] davidcalle: indeed. I never thought about it actually because I didn't know it was possible, but I now understand how it work to split the resutlts [09:20] something.split('text where the new result should start and the first stop ') [09:20] that's so genius :) [09:22] malin, hehe :) [09:22] yeah :) [09:23] maybe I should make it not write: Tidene angir tidligste passering osv også [09:23] ah, that's norwegian :p [09:27] malin, yes, just need to result.replace ('the tidene string', ''), in the for loop, before appending to the model [09:27] ok [09:27] or I can clean it up in the patjunk-thing too? [09:28] malin, yes you can [09:28] but when you say in the for-loop [09:28] you mean in the for result in results_list [09:28] ? [09:28] but on the line before model.append [09:29] hi guys, i'm having trouble with a java application in unity, it opens two pixel wide, a vertical ribbon that's impossible to resize, it works in unity-2d, anybody has infos about that? [09:30] malin, yeah, to do it for each result. But it also works if you do it in patjunk. [09:31] if I do it in patjunk, will it be done in every result then? [09:32] I think next time I make a lense, I will go on to use the for-loop and result.replace ('the string to change', 'string to change to') :) [09:33]