=== yofel_ is now known as yofel [05:37] good morning === bschaefer_ is now known as bschaefer [08:01] didrocks: https://launchpad.net/zeitgeist-extensions/trunk/0.0.9 is out. Never mind 0.0.8... [08:02] kamstrup: excellent, thanks :) [08:12] njpatel: I'll be looking into the libunity gir issue [08:13] kamstrup, awesome :D [08:13] the pythonics will be happy [08:13] (and me, I need to write a grooveshark lens for app developer week!) [08:14] s/lens/scope [08:14] njpatel: lol [08:14] njpatel: any bug report? [08:14] kamstrup, none that I can find, but davidcalle has been venturing into making a python lens/scope [08:14] kamstrup, really I just need someone to test to see what's broken, if anything [08:15] kamstrup, main issue is that it's had no testing... [08:15] njpatel: ok, i'll try to forward port my old python lens and see where it takes me [08:15] shweet [08:15] * davidcalle hugs kamstrup [08:16] davidcalle: :-) [08:21] kamstrup: zeitgeist-daemon is craching a lot since the update [08:22] kamstrup: the fts one [08:23] I have been doing some hacking about so that I could use the application lens [08:24] http://people.ubuntu.com/~alanbell/appmenu.ogv [08:24] I feel the lens suffers from extreme scope creep to the point that it fails the primary purpose of launching applications I want to run [08:25] kamstrup: bug #839382 [08:25] so I ripped out the zeitgeist unpredictable stuff at the top and the advert for random stuff from software centre at the bottom and switched the filters to categories [08:25] Error: Launchpad bug 839382 could not be found [08:25] now I can find and run applications again \o/ [08:27] didrocks, i will also look into it [08:27] seif: adding you to it [08:27] would there be a way of making the main applications lens work like this (configuration option somewhere to turn off the cruft)? [08:28] seif: done, you should have access [08:31] wow [08:31] this is messed up [08:32] kamstrup: seif: so, you have the symbol from the cjkied fts, the system one is running now and you should get the rest of the stacktrace in a few min (minus the fts part as the symbol won't match) [08:36] didrocks: odd, I am dogfooding it here without issues... [08:36] kamstrup: I just installed the new version, killed the daemon, killed the datahub, and the unity-files-lens [08:37] kamstrup: then go to the file lens, it restarts the family [08:37] make a search [08:37] a get this crash, reliably (and not result at all, of course) [08:37] kamstrup: do you think I missed something? [08:38] didrocks: don't know... I just killed zg here, and let the rest of the services sort themselves out :-) [08:39] kamstrup: I'm downgrading fts to confirm [08:39] kamstrup: no crash after downgrading and killing zg [08:40] didrocks: ?! 32 or 64 bit? [08:40] kamstrup: 32 [08:40] didrocks: m2 [08:40] kamstrup: weren't you talking about a quiet Friday? :p [08:40] didrocks: so, it's a segfault in libxapian in fact, and not just some Python screwup [08:41] kamstrup: indeed, seems you trigger a bug in the xapian xjk patch, maybe? [08:41] cjk* [08:41] kamstrup: I think your database is upgraded with the zg-fts change? [08:41] didrocks: can you run the test.sh script from the fts package? [08:41] kamstrup: ok, should I reinstall it? [08:41] didrocks: you need to install Xvfb [08:42] or it takes from trunk? [08:42] didrocks: you can run it seprately, it's sets up its own bus and everything [08:42] kamstrup: let's have a try [08:42] didrocks: from tarball or trunk - it's the same [08:43] kamstrup: $ ./fts/test.sh [08:43] Done - tests passed [08:43] ?! [08:43] odd [08:43] so, probably happening when reindexing... [08:44] kamstrup: I don't see how it takes the local fts extension in your script [08:44] didrocks: magic [08:44] export PYTHONPATH="$(dirname $0)" [08:44] ok :) [08:44] kamstrup: I fear magic :) [08:45] as you should [08:45] :-) [08:45] heh :) [08:45] kamstrup: let me reinstall 0.0.9 and try, if it was the reindexing [08:45] didrocks: can you kill zeitgeist-daemon and start it manually in a term to watch the output? [08:46] kamstrup: sure [08:46] didrocks: or wait - we have a ~/.cache/zeitgeist/daemon.log in fact :-) [08:47] kamstrup: I see a lot of "Indexing" [08:47] didrocks: that indicates that it reindexes your db [08:47] [DEBUG - zeitgeist.fts] Committing FTS index [08:47] ok, let's try a search [08:47] working now [08:48] hehe, I feared as muc [08:48] h [08:48] weird, the reindexing sounds fast, but maybe you continued where it failed? [08:48] no, it should be transactional [08:48] ok, so maybe as it was started by the lens [08:48] s/should be/is/ [08:48] and the lens was started while opening the file one [08:48] I didn't wait enough [08:48] hum, would be there a way to avoid that? block answering? [08:49] or returning dummy empty data? [08:49] didrocks: yeah, the reindexing can take a little time depending on log size and cpu [08:49] it will be better than nothing [08:49] you restarted insulting my computer! :-) [08:49] yes, it's slow! ;-) [08:49] didrocks: lol [08:50] kamstrup: so maybe sending back empty data? no result for a while is better than a crash? [08:50] didrocks: I still don't understand why we crash... [08:51] didrocks: in theory it should just work [08:51] kamstrup, i think i know one of the crashes [08:51] well its not ur bug exactly [08:51] seif: "one of"? we have more than one? [08:51] lol [08:52] https://bugs.launchpad.net/zeitgeist/+bug/807950/ [08:52] Ubuntu bug 807950 in zeitgeist (Ubuntu Oneiric) "zeitgeist-daemon crashed with LookupError in remove_from_connection(): <_zeitgeist.engine.remote.RemoteInterface at /org/gnome/zeitgeist/log/activity at 0xb74ee2cc> is not exported at a location matching (None,None)" [High,Triaged] [08:52] but this is a crash that happens on session end [08:52] ah [08:52] but notified upon the next session start [08:52] kamstrup: I guess we should find a solution before I push fts to oneiric, isn't it? [08:53] seif: we are discussing a crash that occurs in relation to me enabling indexing of CJK characters in the latest fts [08:53] didrocks: :-S yes [08:53] kamstrup, ok count me out [08:53] sorry [08:53] seif: :-) [08:55] didrocks: it's not impossible that it's some race condition in libxapian. I do the indexing in a thread to avoid starving the mainloop of zg - but searches are coming in from the main thread. I am quite sure the xapian API is thread safe, but maybe not so mch :-) [09:00] kamstrup: as you wish, I'm happy with a workaround for blocking the search meanwhile if it touches xapian [09:01] didrocks: I can see, most conveniently, the threadstacktrace includes the line in the fts.py the segfault was at [09:01] didrocks: namely line 251 [09:03] I would be surprised about xapian not being thread safe as well [09:04] didrocks: it's probably trying to write some updates in the worker thread, and then I try to close the index from the main thread [09:06] kamstrup: makes sense :) [09:07] Trevinho, around= [09:07] ? [09:10] smspillaz, around? [09:12] with unity (last revno) when i minimize an app and then i raise it, only the window decorators appear... [09:13] only client side decorator windows (like chrome) work... [09:17] didrocks: Browsing the docs I can't actually see anywhere they explicitly say that the apis are thread safe [09:17] didrocks: maybe I just assume it because I am used to that from Lucene... [09:18] it is single-writer-multiple-reader, but that's not the same as thread safe [09:18] kamstrup: oh really? but anyway, you are in single-writer one-reader mode there, isn't it? === API is now known as Guest88782 [09:19] didrocks: it would look like that, but not really :-) The .close() call is from the main thread, and it may do writing... [09:20] ahah! [09:20] yeah, ok ;) [09:21] didrocks: aha! so I am a dufus... http://xapian.org/docs/overview.html (search for "thread") === zyga-afk is now known as zyga [09:23] kamstrup: you can say you wanted to ensure it's the case ;) [09:24] didrocks: so you say your api is not thread safe?! I don't believe you! ;-) [09:25] kamstrup: heh, "I will proove your're wrong!" :-) [09:27] didrocks: i'll grab some tea and fix it [09:28] kamstrup: I guess you mean "fix xapian to make it thread safe" [09:56] andyrock: rgeat [09:56] *great [09:56] andyrock: I'll fix that, hang on [09:57] smspillaz, thx [10:09] didrocks: fudge! this is gonna be hairy... :-( sqlite is not thread safe - so I can't access the zg db from my xapian worker thread... [10:10] didrocks: aha! I am so clever! I have the elegant solution! === Guest88782 is now known as apinheiro [10:15] kamstrup: \o/ what are you doing to do? [10:17] kamstrup, you are a show off (I wish i had those mighty hacking fingers) [10:51] didrocks: ok can you test r63? [10:51] kamstrup: sure, one sec [10:52] kamstrup: the thing is that I'm already upgraded [10:53] didrocks: one way to test is to kill zg, and then rm -rf ~/.local/share/zeitgeist/fts.index, and restart zg [10:54] which also reminds me... I should add a DBus method to trigger a reindex... [10:54] kamstrup: ok ;) I won't loose all my files indexation that are already indexed? (won't start from an empty database) [10:54] didrocks, nope [10:54] didrocks: nope, the db is left intact, it's just the xapian index [10:55] didrocks, u lose them if u remove activity.sqlite [10:55] ok, doing then :) [10:55] oh so, it will reopens all the files seen by zg (in activity.sqlite) and restart indexing, got it! [11:00] kamstrup: hum, crashed… [11:01] kamstrup: weird, it continued indexing, for a while (can see that), I made some research, still indexing [11:01] and at some point… crash… [11:01] then respawn [11:01] and now, it indexed everything again [11:01] and search works [11:03] qlet me check if the crash is not when I killed it [11:04] no, it's a little bit later [11:06] 8875 [2011-09-02 12:57:24,301] - DEBUG - zeitgeist.fts - Indexing 'application:/ [11:06] 11738 [2011-09-02 12:58:19,247] - DEBUG - singleton - Checking for another running instance... [11:06] 11738 [2011-09-02 12:58:19,839] - INFO - zeitgeist.sql - Using database: /home/didrocks/.local/share/zeitgeist/activity.sqlite [11:07] kamstrup: easier to read like that: http://paste.ubuntu.com/680468/ [11:09] didrocks: a "crash" is a little less well defined because we have multiple threads here. Just because one thread prints stacktrace the others can continue running [11:09] kamstrup: ah, indeed… it seems the whole process stopped for a while though? [11:09] didrocks: odd... [11:09] but at least, it finished its job without me having to restart [11:10] didrocks: if you pull the latest rev I have added a ForceReindex() method you can call over d-feet [11:10] kamstrup: ok :) at least, I think we can go with that one, wdyt? [11:11] took 5 minutes there to reindex everything [11:11] didrocks: wow! [11:11] didrocks: how many items in your log? [11:11] I was untarring Qt at the same time :) [11:11] didrocks: not fair :-) [11:11] so not relevant! :-) [11:12] kamstrup: ok, can u roll a releaz pleaz? :) [11:12] didrocks: i think the latest rev is good. And with the ForceReindex method on DBus we can more easily debug problems that might pop up [11:12] indeed :) [11:12] didrocks: ok, I'll roll it up for you then [11:12] excellent, thanks! :) [11:19] didrocks: ok... third time's the charm https://launchpad.net/zeitgeist-extensions/trunk/0.0.10 [11:20] kamstrup: I prefer when there are only binary digits in a release version :) [11:21] :-) [11:37] kamstrup: hum, there will still be some search stuff to tweak (I guess it's timeouting in the home dash) [11:37] kamstrup: like, I'm searching "foo" in the file lens and I get a foo file (just one result) [11:37] doing the same in the home dash, and I get nothing [11:39] didrocks: is that while it's indexing - or a general problem? [11:40] kamstrup: general problem [11:40] didrocks: hmmm, that's a new one... I ownder if it's somehow related to https://bugs.launchpad.net/unity/+bug/837075 [11:40] Ubuntu bug 837075 in unity "Cannot search for some apps in Dash" [High,Confirmed] [11:40] * kamstrup needs to log out [11:42] kamstrup: already on natty, some people mentionned to get unreliable result IIRC (on the french forum for instance) [11:43] didrocks: odd... I use it all the time 100% reliable... [11:44] kamstrup: I guess it appears more on "slow" machines/harddrives [11:44] didrocks: probably - which tends to indicate that it's a race condition [11:44] didrocks: and only in files? [11:44] yeah, I guess so [11:44] ie not in apps [11:44] not sure, I can reproduceably get it in files [11:45] do you want me to launch the daemon by hand? [11:45] so that I can see if the result is set before/after the timeout [11:45] that'd be nice [11:45] kamstrup: will do in a few (10 minutes), just finishing some qteries first :) [11:46] sure === daker_ is now known as daker [11:52] kamstrup, since you are back ;) what do you think about this bug 820322 [11:52] Launchpad bug 820322 in unity (Ubuntu) "searching for the same term the second time in the dash is slow" [Undecided,New] https://launchpad.net/bugs/820322 [11:53] unity or the daemons bug in your opinion ? [11:57] it seems not be an issue anymore in Oneirc :/ [11:59] om26er: I can't seem to reproduce that here... [12:00] kamstrup: from what I see it's really because the daemon is slow to answer (it answers after the timeout) [12:00] kamstrup, me neither false alarm :( [12:00] kamstrup: I guess it's worse when doing the same search in all lens in parallel [12:00] om26er: but generally I'd say a Unity issue, as in most cases the daemons should onlyt get *faster* on the second query as the caches are warm [12:01] om26er, I fixed it last week I believe [12:01] kamstrup, something certainly changed in the dash which made it little better. [12:01] i was doing something silly...let's not talk about it :) [12:01] and now we know who fixed it ;-) [12:01] :-) [12:02] kamstrup, did we learn about threads today? ;D [12:02] njpatel: watch out - or your life will be hanging by a thread... [12:03] ;-) [12:03] oh, the search result order has been fixed. thanks, i was sick of opening "xchat.png". :P [12:04] kamstrup, hehe [12:04] njpatel, in dash applications are not given the higher priority in the result [12:04] didrocks: you're not depending on specific bzr branches for releases of dee, right? I want to split trunk into a stable 0.5 series and a 0.6 all-things-go series [12:05] om26er: they do? that's what i was talking about. [12:05] I manage to "fix" my duplicated music issue yesterday by accidentally deleting 3/4 of my music :/ [12:05] om26er, 'tis a bug, please file against 4.14.0 and assign to me [12:05] htorque, yes i search empathy and got to gwibber results [12:05] kamstrup: no, it's ok, you can do the split :) [12:05] htorque, om26er it's random [12:05] \o/ rock and roll [12:06] njpatel: oh, in that case i take back the 'thanks' :P [12:07] :) [12:07] eww [12:08] smspillaz, this branch is causing some serious problems https://code.launchpad.net/+branch/~smspillaz/unity/unity.fix_838062 [12:09] combined with https://code.launchpad.net/~smspillaz/unity/unity.fix_output_bug/+merge/73752 [12:09] Ubuntu bug 73752 in Bazaar "Bzr needs a simple way to prevent revision history changing order on shared branches" [Low,Fix released] [12:15] njpatel: ! one can no longer type a URL into the Dash? I did that just last week... [12:17] kamstrup, what the hell? the dash just crashed when I typed in htt [12:17] kamstrup, is that the issue you had? [12:18] njpatel: not exactly. I just expect example.org to give me one hit which launches the url [12:19] kamstrup, we're sending stuff out to the lenses, maybe I broke something there? (it was in files scope, right? I thought that part of the code was fine :/) [12:20] njpatel: I can't recall exactly where that code was - but yeah, it was a simple regex... [12:20] njpatel: hmmm... also seems my files lens has stopped responding [12:20] kamstrup, definitely in files lens, but first I'll take a quick look at why this crashed [12:24] kamstrup, http://www.google.com works [12:24] www.google.com works too [12:24] kamstrup, so, yeah, definitely hang in files place is the issue === MacSlow is now known as MacSlow|lunch|er [12:34] guys, did anybody had experience of creating lenses using python and libunity-4.0? [12:34] om26er: yeah, I'm aware. these problems are complicated though, give me some time [13:17] didrocks, Hi is there a way to override gpu blacklist? [13:19] om26er: indeed, there is one [13:19] * didrocks doesn't remember the variable he added :) [13:19] haha [13:20] om26er: one sec :) [13:20] didrocks, any where i should look? [13:20] ok.. [13:21] om26er: just set UNITY_FORCE_START to something :) [13:21] export UNITY_FORCE_START=1 [13:21] didrocks, thank you [13:22] yw ;) === m_conley_away is now known as m_conley [13:32] mardy: hey I'm Gerry. I was looking at the wallpaper bug, and can (like you) only find nautilus to blame. [13:32] greyback: yep. I also tried with all the unity-components not running, and it still happens [13:34] mardy: I must check to see if it happens with Unity. [13:34] mardy: FYI, I uploaded the metacity change [13:34] mardy: thanks :) [13:35] didrocks: thanks to you! BTW, I also submitted the patch upstream (as a bug), so you might have to remove it from debian/patches/ if it ever gets accepted [13:36] mardy: sure, the usual business :) try to ping marmanel to ensure he can have a look [13:36] mardy: he's sick right now apparently [13:36] didrocks: ok, I'll ping him at regular intervals :-) [14:23] roignac: I think the gwibber lens is in python [14:23] well, the previous version did [14:23] now they are ported to vala [14:23] gah, OK I was looking at their packaging [14:24] vala makes my head spin [14:24] not the language itself, just the c files it makes and compiler errors that report against the c files rather than the vala files [14:25] AlanBell, it's actually way better than it used to be but, yes, sometimes you just want to kill someone [14:25] I would be interested in porting my appmenu lens to python because I think it is overcomplicated in vala [14:25] preferably a vala dev [14:25] appmenu lens? [14:25] do tell [14:26] i'm actually pretty happy with python [14:26] but libunity-4.0 bindings just don't allow me to set filters for a lens [14:26] I took the applications lens and ripped out the zeitgeist stuff and the software centre suggestions and the filters and made it just show applications categorised by category [14:26] http://people.ubuntu.com/~alanbell/appmenu.ogv [14:27] so now I can find and run applications using unity \o/ [14:27] more info here - https://answers.launchpad.net/ubuntu/+source/libunity/+question/169894 - I'd be pretty happy to solve this [14:28] AlanBell is that categorized Unity dash someway official? [14:28] AlanBell, lol, nice [14:28] AlanBell, the filters aren't helping? [14:28] (in 11.10) [14:29] lucazade: nope, it was just me ripping out stuff until I could use it! [14:29] the filters take up half the dash [14:29] AlanBell, thanks.. i was thinking to the same layout for the dash [14:30] so I had a dash with three lines of icons, top being zeitgeist recently used stuff, bottom being adverts of random stuff from software centre right hand side being the filters, so half of the middle line was three of the applications I had installed [14:30] hey didrocks, got a minute? [14:31] didrocks: mardy, greyback and I are looking at some visual glitches that happen during session login; I found 4 of them so far [14:31] code is here https://code.launchpad.net/~alanbell/+junk/appmenu-lens [14:31] heya didrocks [14:31] jcastro suggested I package it for extras, but I have no idea how to do that [14:31] hey jcastro [14:31] didrocks: doing anything? [14:31] * jcastro tries to trap didrocks [14:32] so I was looking at: https://launchpad.net/unity/+patches [14:32] because why not .... [14:32] jcastro: not stopping since this morning :/ [14:32] pinging machine this week [14:32] ah ok [14:32] I can bother someone else [14:32] jcastro: like… dx? ;-) [14:33] yeah, well I was just wondering if a patch gets accepted into ubuntu why it remains open in the upstream project [14:33] but I'll figure it out [14:33] I think it's just an LPism [14:33] jcastro: let me have a look [14:33] (quick one ;)) [14:34] I think if people are just attaching patches then I can have someone grab them and properly propose them [14:38] jcastro: no patch is in ubuntu, I just closed the cjk one [14:39] didrocks: the one that I think you should have a look at is https://bugs.launchpad.net/unity-2d/+bug/839610 because I suspect 02_additional_marge_for_unity.patch patch to Nautilus might be the culprit [14:39] Ubuntu bug 839610 in unity-2d "Background wallpaper briefly appears shifted horizontally by around 15 pixels on startup" [Medium,Confirmed] [14:40] njpatel: could the main applications lens be made to be uncluttered and categorised? [14:41] Kaleo, I believe bug 804435 is a dup of that wallpaper bug [14:41] Launchpad bug 804435 in unity-2d "Wallpaper is loaded twice with different alignment by gnome-session and nautilus (Oneiric)" [High,Confirmed] https://launchpad.net/bugs/804435 [14:41] AlanBell, have you tried out 11.10 (sorry if you've answered) [14:42] njpatel: this is using 11.10 [14:42] AlanBell, allows you to filter per-category pretty easily [14:42] lucazade: no, I don't think so; when I deactivate gnome-session's wallpaper it still occurs [14:42] yes, the filters kind of wokr [14:42] AlanBell, ah, then no, main apps lens are hard-coded in the lens [14:42] Kaleo: the patch only touches the icon [14:42] Kaleo, ok :) [14:42] Kaleo: not the wallpaper [14:42] njpatel: but the filters use up half the room in the dash [14:43] didrocks: ok [14:43] lucazade: I just added clearer steps to reproduce [14:43] if there was an option to turn off the zeitgeist suggestions (they are not too bad) and the software centre stuff (really dislike that) then I would be fine with it [14:43] Kaleo: it's really some separated process, I can have a look next week to confirm anyway [14:44] didrocks: we have 4 different bugs that hinder the visual beauty of the login process [14:44] interesting [14:44] Kaleo: indeed, in particular, the one making everything grey [14:44] didrocks: 2-3 of them affecting both unity-2d and unity [14:45] haha... yesterdays unity release has broken the horizontal renderer again and unity-2d fixed it's horizontal renderer [14:47] kenvandine: do you like it? [14:48] kenvandine: (the u2d's one I mean) [14:49] Kaleo, of course :) [14:50] kenvandine: great, let us know if anything's wrong with it [14:50] Kaleo, the grey text looks a lot better than 3D [14:50] will do [14:55] davidcalle: cool [15:31] jcastro: ok, I'm a little bit more in a quiet time right now, do you need me for anything? :) [15:31] Kaleo: as well ^ [15:31] didrocks: yes :) [15:31] didrocks: me me me me :) [15:32] didrocks: just wondering if it's worth going through all the bugs with patches [15:32] and see if there's any low hanging fruit [15:32] jcastro: I guess that someone on dx should do it, the list is not that long :) [15:33] * jcastro nods [15:33] Kaleo: welcome back! [15:34] jcastro: :) [15:34] jcastro: you mean kamstrup no? [15:34] oh right [15:34] but also, hi to you! [15:38] kenvandine, -> https://launchpad.net/indicator-session/+milestone/0.3.4 [15:38] 0.3.4.2 [15:39] should fix the apt menuitem weirdness [16:19] ronoc, i got it [16:19] any bug associated with it? === MacSlow|lunch|er is now known as MacSlow [16:21] kenvandine, yep [16:21] * ronoc finds bug no [16:23] kenvandine, https://bugs.launchpad.net/ubuntu/+source/indicator-session/+bug/831650 [16:23] Ubuntu bug 831650 in indicator-session (Ubuntu) "indicator-session states 'software up-to-date' although there're updates available" [Undecided,New] [16:23] thx [16:42] gord, if I remember right, did you write the code that speeds up the search and presentation of items in the dash? [16:42] I seem to remember it was you [16:42] it is rocking because the dash feels really zippy now :-) [16:42] jono, not search but yeah presentation, anything drawing related really [16:42] great work! [16:42] jono, still only done half the performance speedups :) [16:42] seriously? [16:42] more to come next week [16:43] so it is going to get faster? [16:43] rocking! [16:43] ~15 minutes until AMA [16:43] I started a hangout if you guys wanna prep [16:43] jcastro, where does it take place, the AMA? [16:43] reddit,com/r/Ubuntu [16:51] jcastro, I'll probably join in in around an hour or so, once i get back home [16:52] njpatel: sweeeeeeeeet. [17:36] we're now answering questions about Unity on reddit! http://www.reddit.com/r/IAmA/comments/k2nvl/we_are_the_ubuntu_unity_development_team_we_3/ [17:41] jcastro, posted to FB too, jcastro [17:43] jono, fixing your music lens bugs :) [17:43] lamalex, you, rock, pal! [17:44] lamalex: you already have a music lens question [17:44] jono: ta for that, I got g+ covered too [17:44] far out [17:50] lamalex: and now a followup [17:50] wow, this got popular fast! [17:58] yah cuz unity da coolest === _dlm_ is now known as dlm [18:37] btw, are you folks aware the files lens is not displaying accurately filtered results? [18:38] https://bugs.launchpad.net/ubuntu/+source/unity/+bug/839807 [18:38] Ubuntu bug 839807 in unity (Ubuntu) "Files lens results filter produces inaccacurate results" [Undecided,New] [18:39]