[07:32] <elvisd> didrocks, hi
[07:32] <elvisd> didrocks: Alan P. asked me to report to you a unity bug
[07:33] <elvisd> didrocks: mentioned here http://goo.gl/byrd6
[07:33] <didrocks> elvisd: hey, you are on the unity-team/ppa right?
[07:33] <elvisd> didrocks: wait i check
[07:34] <elvisd> yes, i am
[07:34] <didrocks> elvisd: can you try adding temporarly the unity-team staging ppa? (unity-team/staging)
[07:34] <didrocks> then upgrade
[07:34] <didrocks> (and then remove the entry in your source.list, it's not a safe ppa generally ;))
[07:35] <elvisd> didrocks: should i remove the unity-team/ppa first?
[07:35] <didrocks> elvisd: no, you can keep it, that's fine
[07:40] <elvisd> didrocks: ok. give some minutes, should finish ajob first
[07:40] <didrocks> sure :)
[07:57] <elvisd> didrocks: ok. ppa added and dist-upgrade too. rebooting now. i'll contact you after.
[07:59] <davidcalle> mhr3, ping
[07:59] <didrocks> elvisd: thanks!
[08:05] <elvisd> didrocks, i rebooted, but after login i had the same issue
[08:06] <elvisd> didrocks, no launcher, no panel, no right click
[08:06] <didrocks> elvisd: ok, we have maybe a fix, do you have time?
[08:06] <elvisd> yes
[08:06] <didrocks> great ;)
[08:06]  * didrocks pushes a new compiz to the ppa shortly
[08:07] <mhr3> davidcalle, hey
[08:09] <elvisd> didrocks, what i can report too, is that right click on desktop doesn't work and the panel hasn't the shadow, it's so flat. :)
[08:09] <elvisd> didrocks, and desktop icons aren't shown
[08:12] <didrocks> elvisd: yeah, everything is linked
[08:14] <elvisd> didrocks, ok.
[08:18] <davidcalle> mhr3, when zeitgeist gives me the n latest video activities, does it checks for the file existence? Same question for GAJ, does it displays deleted files?
[08:18] <mhr3> davidcalle, no, the files might be already deleted
[08:19] <mhr3> davidcalle, we need the bug where video lens displays old results fixed
[08:19] <davidcalle> Ok, that's what I thought.
[08:22] <davidcalle> mhr3, I will push the fix tonight/tomorrow for you to test it in the PPA.
[08:22] <mhr3> davidcalle, show me the code, i'll tell you if that fixes it :)
[08:23] <davidcalle> Ok :) I think I'm just going to check the status of the search before appending my batch of results to the model.
[08:23] <mhr3> you need to spin the mainloop in between though
[08:24] <mhr3> (i guess everything you're doing is synchronous)
[08:25] <davidcalle> mhr3, no actually, the thumbnailer and locate are async.
[08:26] <mhr3> davidcalle, but the rest isn't
[08:27] <davidcalle> mhr3, right
[08:27] <mhr3> davidcalle, where did you find the zg code you're using?
[08:27] <mhr3> even that is sync, and we strongly discourage its usage
[08:29] <davidcalle> mhr3, a buch of places I believe : rainct code samples, rb plugin, etc
[08:29] <davidcalle> mhr3, oh, I wasn't aware of it
[08:29]  * mhr3 wants to eradicate all sources where the sync variant is used :)
[08:30] <mhr3> davidcalle, this is how the api is supposed to be used> http://paste.ubuntu.com/896135/
[08:32] <mhr3> well.. minus the awful magic numbers for the constants :)
[08:33] <davidcalle> mhr3, hehe. Ok, I will switch to async zg.
[08:34] <mhr3> davidcalle, also doing that will force you to spin the mainloop, so the cancellable will have correct state when you'll be about to append results to the model
[08:34] <davidcalle> Indeed
[08:34] <mhr3> and you'll be able to do simple if cancellable.is_cancelled(): return
[08:35] <davidcalle> mhr3, that's what I was thinking about, but without caring about the loop, thanks :)
[08:36] <mhr3> davidcalle, btw should you also wait for the updatedb call to finish?
[08:36] <mhr3> shouldn't?
[08:37] <mhr3> or does locate block in case updatedb is running?
[08:38] <davidcalle> mhr3, nope. I don't want to wait, in case there is an existing updatedb, I still want to trigger the first results of the search on it.
[08:39] <didrocks> elvisd: around?
[08:39] <elvisd> didrocks:  oh yes
[08:39] <didrocks> elvisd: if you upgrade from the staging ppa, you should get compiz 1:0.9.7.2-0ubuntu1~ppa5
[08:39] <didrocks> elvisd: can you try with that?
[08:41] <mhr3> davidcalle, but then you might get results which aren't there, or not get results which are there now
[08:41] <elvisd> didrocks:  dist-upgrade says: The following packages have been kept back: compiz
[08:41] <didrocks> elvisd: apt-cache policy compiz?
[08:42] <elvisd> didrocks:  http://pastebin.ubuntu.com/896143/
[08:43] <didrocks> elvisd: looks good, maybe you had some archive skew
[08:43] <didrocks> elvisd: try apt-get update
[08:43] <didrocks> if dist-upgrade still says keeping back compiz
[08:43] <didrocks> try apt-get install compiz
[08:46] <elvisd> didrocks:  http://pastebin.ubuntu.com/896148/
[08:46] <didrocks> elvisd: ok, every binary is not published, funny ppa :)
[08:46] <didrocks> elvisd: let's wait 5 minutes
[08:46] <didrocks> should be fine
[08:46] <davidcalle> mhr3, I GIo check several time during the process for the existence of the file.
[08:47] <mhr3> davidcalle, ok that fixes the first part, not the second
[08:48] <davidcalle> And results which are there will arrive at some point during the search. I've tested updatedb with more than 1000 video files in nested folders and it's stays fast.
[08:48] <davidcalle> mhr3^
[08:48] <davidcalle> it*
[08:49] <didrocks> elvisd: if you are braced, and if apt-get upgrade and dist-upgrade don't work right now (not publishe)
[08:49] <mhr3> davidcalle, on your machine :)
[08:50] <didrocks> elvisd: you can download the binary directly:
[08:50] <didrocks> https://launchpad.net/~unity-team/+archive/staging/+build/3311346
[08:50] <didrocks> under "Built files"
[08:50] <didrocks> you need compiz-core_0.9.7.2-0ubuntu1~ppa5_amd64.deb , compiz-gnome_0.9.7.2-0ubuntu1~ppa5_amd64.deb, compiz-plugins-default_0.9.7.2-0ubuntu1~ppa5_amd64.deb , compiz-plugins_0.9.7.2-0ubuntu1~ppa5_amd64.deb , libdecoration0_0.9.7.2-0ubuntu1~ppa5_amd64.deb
[08:50] <davidcalle> mhr3, I mean, it deals with very big files, is waiting for updatedb to finish worth the wait for everything to be displayed on the first letter in the search field?
[08:50] <didrocks> and then install them with sudo dpkg -i <all deb files>
[08:52] <mhr3> davidcalle, well from what you're saying it seems like it's implicitly waiting for it anyway (it the locate call doesn't return until updatedb finished)
[08:53] <mhr3> davidcalle, plus you need to drop kernel disk caches everytime you test these kind of things, to see how it *really* behaves
[08:53] <elvisd> didrocks:  Ok, downloaded and installed. I'll reboot in minutes
[08:54] <didrocks> elvisd: thanks ;)
[08:54] <elvisd> didrocks:  thanks to you! :)
[08:55] <davidcalle> mhr3, locate will still return results on the existing db. If updatedb takes time, it means there are a lot of files and the lens only displays 100 results max in any case. So, to get a specific result on a large video db, you need to enter several letters. It gives updatedb the time to run.
[08:56] <davidcalle> mhr3, updatedb doesn't block locate. It writes results in a temp file, then copies them to the final db when it's done.
[08:57] <mhr3> davidcalle, or just that the disk is busy
[08:58] <davidcalle> mhr3, indeed, but in this case, any updatedb will take time. The files lens will, for example.
[08:58] <mhr3> if it doesn't block then it's creating races where search for one letter won't return any results at one point, and if you try 10 seconds later it will return something
[09:02] <davidcalle> mhr3, ok, how do you suggest this to be solved? Updatedb is the fastest thing I could find, and I don't want locate to depend on it : big video db would have to wait for every single search.
[09:03] <mhr3> davidcalle, well the first problem i see is that you run updatedb on every single search
[09:04] <mhr3> but any io is going to be slow, you're unlikely to find anything faster
[09:04] <elvisd> didrocks:  Funziona! ("It works!" in italian)
[09:04] <didrocks> \o/
[09:04] <didrocks> elvisd: thanks a lot! :-)
[09:04] <didrocks> I mean, really, *a lot*
[09:05] <mhr3> davidcalle, a "fix" could be to use only zeitgeist, but then there may be files which aren't logged, so... :/
[09:05] <elvisd> didrocks:  as said, thanks to you! Can i leave the staging ppa until released?
[09:05] <davidcalle> mhr3, I could trigger updatedb only when the lens is opened, but I definitely want to cover the use case where the users imports a large video db, and while it's doing it, searches for new videos in the lens.
[09:05] <didrocks> elvisd: sure, just remove it from your source.list
[09:05] <didrocks> elvisd: but keep the packages
[09:05] <didrocks> (no ppa-purge)
[09:06] <elvisd> ok
[09:06] <didrocks> elvisd: we will hopefully release tomorrow and then, you will get the released one :)
[09:06] <mhr3> davidcalle, yea, that is probably a step in right direction (to limit the invocations of updatedb)
[09:06] <elvisd> didrocks:  ok. have a nice day!
[09:07] <didrocks> elvisd: thanks, you made my day until now! Have a nice day as well :)
[09:08] <elvisd> didrocks:  and you my day. 1 day and half without unity and i'm lost!
[09:08] <davidcalle> mhr3, I could correlate, somehow the size of the video folder, the number of items it contains and the size of the db file to see if an updatedb is needed + updatedb on lens opening.
[09:08] <mhr3> davidcalle, that sounds too magical
[09:08] <davidcalle> mhr3, I agree.
[09:08] <didrocks> elvisd: heh, we got really easily used to it, right? :)
[09:09] <elvisd> didrocks: absolutely right!
[09:09] <mhr3> davidcalle, i wouldn't really mind something simple, like only run updatedb when the lens is shown + a limit of one run per minute or so?
[09:10] <elvisd> didrocks: and it's the same for "converted" family and friends (the grand-father too!)
[09:10] <didrocks> elvisd: waow, nice to know! Thanks for the cheering, we work hard to make it the best we can :)
[09:10] <mhr3> davidcalle, how long does the updatedb take approximately on a huge video folder?
[09:11] <davidcalle> mhr3, yeah, but there really is a magic feeling to the lens when you are importing a bunch of videos and they just appear as you search and you can see the thumbnails created if front of you. :) But ok, less magic for less IO...
[09:11] <mhr3> davidcalle, 2 seconds, 30seconds? 5minutes?
[09:11] <elvisd> didrocks: and we'll achieve it! cheers from switzerland, to you and the whole team!
[09:11] <didrocks> thanks again elvisd ;)
[09:11] <davidcalle> mhr3, I currently have more than 500 videos and it's less than one second.
[09:12] <mhr3> davidcalle, in that case there's no harm in waiting for it to finish
[09:12] <mhr3> as long as a spinner shown the user will be happy :)
[09:12] <davidcalle> mhr3, there is harm, big db will have to support this wait each time the open the lens.
[09:12] <davidcalle> they*
[09:13] <mhr3> davidcalle, but if nothing changes the updatedb will be much faster, no?
[09:16] <davidcalle> mhr3, not sure. But I'm going to restrict updatedb on lens opening, it feels like the right thing to do. I'll ask the novacut guys to test the PPA on their outrageous amount of raw video footage.
[09:17] <mhr3> davidcalle, heh, ok, anyway the bug we talked about earlier is more important that this atm
[09:18] <davidcalle> mhr3, ok. Got to go (one hour late for work ... :-)
[09:24] <popey> elvisd / didrocks nice one!
[09:24] <elvisd> popey: Oh yes! Thanks to you too
[09:25] <popey> I did nothing :D
[09:25] <elvisd> popey: you routed me.
[09:25] <popey> "It's not what you know, it's who you know" :D
[10:10] <apw> didrocks, am i expecting the new unity features in already today ?
[10:11] <didrocks> apw: well, you should, we are still fighting since yesterday to some remaining issues
[10:11] <didrocks> apw: so, can't ensure we will release today, we are doing our best and not sleeping a lot…
[10:11] <apw> nasty /me hands didrocks some coffee
[10:12]  * didrocks takes it with pleasure :)
[10:12] <didrocks> speaking of coffee
[10:12] <didrocks> time to take a real cup again ;)
[10:12] <didrocks> apw: if you have time, please update to the staging ppa
[10:12] <didrocks> and reboot
[10:13] <didrocks> just to ensure that the workaround we introduced for some special cases didn't regress for you
[10:51] <greyback> didrocks: hi, a recent Unity change means unity2d has one MR to add: https://code.launchpad.net/~gerboland/unity-2d/averageBgColor-use-our-key/+merge/98838
[10:51] <greyback> didrocks: is this possible now, or shall we be patient?
[10:52] <didrocks> greyback: unblock now please :)
[10:52] <didrocks> it's fine ;)
[10:52] <greyback> didrocks: I've UNBLOCK in commit message & description
[10:52] <didrocks> (I saw the discussion on the ML)
[10:52] <didrocks> well ML -> launchpad emails
[10:55] <didrocks> greyback: there are some unity branches merging, it's probably why it's waiting
[10:55] <greyback> didrocks: ah of course, my bad. Is there any way for me to see that?
[10:56] <greyback> as in a status page of the merger?
[10:56] <didrocks> greyback: no, unfortunatly, you don't have shell access, isn't it?
[10:56] <didrocks> greyback: I can remove the dep between unity and unity-2d
[10:56] <didrocks> greyback: it's just when there is an ABI break, you have to wait manually
[10:56] <greyback> didrocks: yeah, leave it for now.
[10:56] <greyback> bad time to start changing things
[10:57] <greyback> didrocks: ok, thanks for your help. I'll leave you to your other business :)
[10:58] <didrocks> greyback: yeah, I confirm there is a unity merge going on
[10:59] <greyback> didrocks: ok, will wait. Many thanks
[11:56] <didrocks> greyback: on https://code.launchpad.net/~gerboland/unity-2d/update-pots/+merge/98392
[11:56] <didrocks> you set the unblock on the commit message
[11:56] <didrocks> not the description
[11:57] <greyback> didrocks: oops, fixing
[11:57] <didrocks> greyback: the other branch is failing
[11:58] <greyback> didrocks: /lib/x86_64-linux-gnu/libssl.so.1.0.0: version `OPENSSL_1.0.1' not found (required by /usr/lib/x86_64-linux-gnu/libcurl.so.4)
[11:58] <didrocks> looking :)
[12:00] <didrocks> greyback: that's weird
[12:00] <didrocks> greyback: is it a new requirement,
[12:00] <didrocks> or maybe curl doesn't have the right dep
[12:01] <didrocks> greyback: for now, one the pot file merged
[12:01] <didrocks> I can disable test
[12:01] <didrocks> for this run
[12:01] <greyback> didrocks: not our requirement. I'm guessing curl changed
[12:01] <didrocks> not really have the time to look at it more
[12:01] <greyback> of course
[12:01] <greyback> thanks
[12:01] <didrocks> yeah, maybe a dep missing :)
[12:01] <didrocks> greyback: ok, let's wait for the pot change merged
[12:01] <greyback> didrocks: ok, but I expect same error :)
[12:02] <didrocks> greyback: I mean, I remove both testing ;)
[12:02] <mhall119> good morning
[12:02] <greyback> didrocks: ah ok :)
[12:57] <tgardner> didrocks, you mentioned earlier this week a feature that would allow a user to only have one launcher bar on a dual monitor setup. do you know when hat will get released? having my mouse captured every time I transition from one monitor to the other is driving me mental.
[13:04] <didrocks> tgardner: I'm releasing now
[13:05] <tgardner> didrocks, cool, I'll watch for the upload
[14:18] <ccooke> huh. That's an interesting bug. software updater is requiring a partial upgrade, then balking because "Can't guess meta-package", "Your system does not contain a ubuntu-desktop, kubuntu-desktop, xubuntu-desktop or edubuntu-desktop package and it was not possible to detect which version of Ubuntu you are running. Please install one of the packages above first using synaptic or apt-get before proceeding.
[14:18] <ccooke> "
[14:18] <ccooke> Probably transitory apt error
[14:19] <ccooke> (ubuntu-desktop is installed, naturally)
[15:22] <zgreg> dbarth: why did you mark #927441 as fixed?
[15:22] <mfisch> davidcalle: mhall119 says you're writing a lens writing guide?
[15:24] <davidcalle> mfisch, I have two different things going on : a simple lens tuto and more general lenses guidelines for third-party devs.
[15:24] <mfisch> davidcalle: when one or both are done, where can I find them?
[15:25] <davidcalle> The tuto will appear on the Ubuntu appdev blog, the guidelines will probably be on the Ubuntu wiki.
[15:26] <dbarth> zgreg: hi
[15:26] <dbarth> let me check this one
[15:27] <zgreg> this is not fixed yet, I proposed a branch with a simple fix, but it's pending
[15:27] <dbarth> zgreg: committed, the branch has been approved, so it should be in nux's autobuild by now
[15:27] <dbarth> but not release yet
[15:28] <zgreg> well, it's not merged yet, is it?
[15:28] <zgreg> if the approval is good enough that's rather confusing, but ok :)
[15:49] <elopio> greyback: sorry, but ping again :)
[16:01] <greyback> elopio: hey
[16:02] <elopio> hey greyback. I hope not to be bothering you. But did you try your changes to launch testability on u1?
[16:03] <elopio> yesterday after fixing the machine I got caught by other tasks.
[16:03] <elopio> and now I'm trying it and I have the same ApplicationNotAvailable :(
[16:03] <greyback> elopio: my change caused the script not to fail with ApplicationNotAvailable
[16:03] <elopio> the log says TasNativeUtils::pidOfActiveWindow Resolved -1
[16:03] <greyback> elopio: hmm
[16:04] <elopio> greyback: http://paste.ubuntu.com/896606/
[16:04] <elopio> it's even more weird if it works for you.
[16:05] <greyback> elopio: damn it failed on me too. I swear it was okay yesterday
[16:08] <elopio> greyback: I believe you :)
[16:10] <elopio> um, I'll continue playing with the hash to see if I find something...
[16:11] <greyback> elopio: I had the app open in visualizer, that made the script succeed. Now why is that...
[16:14] <elopio> greyback: yes, confirmed that.
[16:14] <elopio> this will let me work. But is there a td channel where I can report this?
[16:19] <greyback> elopio: I'll keep at it. I'm not aware of any channel for TD unfortunately
[16:22] <elopio> greyback: I found the bug tracker. I'll add a ticket, please let me know if you find something else.
[16:22] <elopio> thanks a lot, again.
[16:22] <greyback> elopio: np
[16:33] <angeloc> hi guys, big trouble with a java app in unity
[16:34] <angeloc> in unity 2d, starting the app makes the window 2 pixel wide, shooting with right click and maximizing the window, make it fully visible
[16:34] <angeloc> in unity 3d, maximizing the windows, the windows still remains two pixels wide
[16:42] <angeloc> nobody?
[16:48] <elopio> greyback: I think it also has to do with the -testability argument.
[16:49] <elopio> without the visualizer, if I start the control panel from the command line with ubuntuone-control-panel-qt -testability
[16:49] <elopio> then the script works too.
[16:50] <elopio> so, starting it from the command line and try_attach=>true seems to be a good workaround for me.
[16:53] <greyback> elopio: good find, it must be bug that start_command does ot append the arguments
[16:53] <greyback> elopio: so this works: http://pastebin.ubuntu.com/896661/
[16:53] <elopio> greyback: ahh, that's even better!
[16:53] <greyback> the :arguments switch is probably useless too
[16:54] <elopio> I can start now with the funny part :)
[16:54] <greyback> but in the background, all it's doing is starting the process separately then connecting, which isn't perfect
[16:54] <angeloc> hi guys, big trouble with a java app in unity
[16:54] <angeloc> in unity 2d, starting the app makes the window 2 pixel wide, shooting with right click and maximizing the window, make it fully visible
[16:54] <angeloc> in unity 3d, maximizing the windows, the windows still remains two pixels wide
[16:56] <greyback> elopio: I must strongly point you to https://projects.developer.nokia.com/Testabilitydriver/wiki/FeatureTDriverVerify
[16:56] <elopio> good. I'll read it.
[16:57] <elopio> pura vida greyback
[16:57] <greyback> elopio: don't use assert_* for checking properties you read straight out of the application. The verify_* methods are better, as they allow for timing discrepancies.
[16:57] <greyback> elopio: using 'sleep' should only be at last resort :)
[16:58] <greyback> elopio: I'm here if you've any questions
[16:58] <elopio> greyback: I'll have more questions, that's for sure :D
[16:59] <greyback> elopio: well you know where I am
[16:59] <greyback> :)
[20:03] <will> hey guys. i'm looking at the BamfLauncherIcon.cpp file in the Unity plugin and I'm seeing Focus(arg); and SPread(true,0,false); etc.
[20:03] <will> are these compiz functions?
[20:03] <will> where could i find descriptions of what they do?
[20:10] <will> er, actually. scratch that, I've found what I was looking for. Sorry about that!