[01:13] good night everyone === IAmNotThatGuy is now known as M0hi [07:50] thanks, Chipaca === IAmNotThatGuy is now known as M0hi [09:10] Happy New Year, everyone! :D [09:14] hi james [09:20] Hey poolie. :) [11:57] Any1?? [11:58] i need help [12:02] hm , i suppose i'll need to fix the topic [12:03] hm, nothing to fix [12:50] ralsina: ping [12:52] is there any update on a KDE client? [12:52] tarragon: no right now, but we are working on that [12:52] tarragon: what nessita said. [12:53] OK I will hold on. [12:53] tarragon: there is work going on currently to make the lower layers more platform-agnostic so the job for a KDE client is easier === teknico is now known as teknico_away === IAmNotThatGuy is now known as M0hi [13:12] ralsina: ping [13:12] nessita: pong [13:12] sorry I miss the other one [13:12] ralsina: any news on the epic review? the lack of pairing seems to be a bug in the pairing code, not in ubuntuone control panel [13:12] Chipaca: would you have more details on that? ^ [13:13] Other than that I think it works just fine. [13:13] And the code looks good [13:13] One question though: shouldn't ubuntuone-desktopcouch be getting installed too? [13:14] or desktopcouch-ubuntuone whatever it's called ;-) [13:14] ralsina: yes! you're right. But I should do that in another branch, I think [13:14] nessita: fine by me, I will approve it then [13:15] That other branch should be much simpler to review ;-) [13:15] what I think we should do is check for desktopcouch-ubuntuone package, if not present, install that, and that will bring all the needed dependencies [13:15] ralsina: indeed [13:15] nessita: right [13:15] nessita: if that doesn't work, it's a bug on desktopcouch-ubuntuone anyway [13:15] yes [13:16] So, I'll approve it now [13:16] thanks! [13:16] nessita: I see John approved it too, so it's done :-) [13:21] * nessita dances [13:22] nessita: I am late with 2 of the branches you proposed yesterday but I will get to them today [13:22] ralsina: thanks! [13:22] mandel: ping [13:23] hey, does anyone have a minute? I don't seem to be able to sync files with U1 [13:24] u1sdtool --status returns "processing queues" but both --waiting-m and --waiting-c return nothing [13:24] there is also nothing in logs but "NOTE - ---- MARK (state: ; queues: metadata: 0; content: 0; hash: 0, fsm-cache: hit=377 miss=108) ----" [13:25] ah, I'm not subscribed to folders.. hmm [13:26] kklimonda: that's a known bug, facundobatista knows the details [13:26] kklimonda: let me see if I can find the bug # [13:27] ah, not that I've subscribed it's finally doing something [13:29] ah :-) [13:31] I got confused by the fact that both Music and Documents have "Synchronize this folders" checked [13:32] but both have subscribed=False in u1sdtool --list-folders output [13:34] * kklimonda is going to look for bug/report it later [13:58] nessita mandel thisfred vds alecu dobey standup in 3' [13:58] ack [13:58] yessir! :) [13:59] sure [13:59] ok [14:00] me [14:00] me [14:00] me [14:01] hi [14:01] hi dobey! Say me! ;-) [14:01] me [14:01] meh; me [14:02] alecu doesn't seem to be around, so let's start. nessita? [14:02] mandel: remember the no-more-gobject review! [14:02] DONE: bug #697211, but triage, haunted reviewers [14:02] TODO: bug #696782 [14:02] BLOCKED: nopes [14:02] NEXT: vds [14:02] Launchpad bug 697211 in ubuntuone-client (Ubuntu) (and 1 other project) "Provide a specific login D-Bus service (affects: 1) (heat: 6)" [Medium,Triaged] https://launchpad.net/bugs/697211 [14:02] Launchpad bug 696782 in ubuntuone-control-panel (Ubuntu) (and 1 other project) "Start DC service in backend to make that op asynch (affects: 1) (heat: 6)" [Medium,Triaged] https://launchpad.net/bugs/696782 [14:02] s/but/bug [14:02] DONE: web/mobile+ vacation [14:02] TODO: api work - to be defined [14:02] BLOCKED: yes [14:03] we need to plan what to do for the APIs work that is in the planning [14:03] ralsina: please [14:03] DONE: reviewed lp:~nataliabidart/ubuntuone-control-panel/applications-woohoo/+merge/44623 again, reported bug #696968, reviewed lp:/~nataliabidart/ubuntu-sso-client/fix-693531/+merge/44919, reviewed lp:/~nataliabidart/ubuntu-sso-client/no-more-gobject/+merge/44988, helped mandel with a branch, chased people around (including HR). [14:03] TODO: half a dozen more reviews, someday actual coding ;-) [14:03] BLOCKED: no [14:03] Launchpad bug 696968 in desktopcouch "Pairing fails (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/696968 [14:03] mandel? [14:03] DONE: talked with people in TIP and finished the u1trial decorators (1 min ago). Reviewed nessita branches. Got asked by people on TIP why not using testtools rather that unittest (why?). [14:03] TODO: review facundobatista branch. Add decorators to desktopcouch tests so that the test suit can be ran on windows/mac os x [14:03] BLOCKED: No [14:04] * mandel looks at dobey [14:04] me [14:04] λ DONE: work on bug 696968, nightlies updates, reviews [14:04] λ TODO: bug pitti about backports, finish 696968 [14:04] λ BLCK: None. [14:04] thisfred: your roll [14:05] DONE: bugfix for bug #696972 helped beuno with couchdb questions TODO bindwood bindwood bindwood BLOCKED no [14:05] Launchpad bug 696972 in desktopcouch "Wrong import path in bin/desktopcouch-get-port (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/696972 [14:06] Ok, coomments, questions? [14:06] noop [14:06] dobey, vds, everyone else who cares, we need to discuss the API stuff [14:06] ralsina: what would that be? [14:06] Definition of API stuff: https://pastebin.canonical.com/41458/ [14:07] Chad is assigned to that too, but I have not seen him today [14:07] ralsina: was this the bug to talk with rodrigo_? https://bugs.launchpad.net/evolution-couchdb/+bug/673568 [14:07] Launchpad bug 673568 in evolution-couchdb (and 1 other project) "Error modifying contact, other error when saving contacts (affects: 28) (dups: 4) (heat: 132)" [High,In progress] [14:07] ralsina: chad is CardinalFang [14:07] ralsina: I am interested in that as well [14:07] he's away [14:07] ralsina: what API? [14:08] mandel: see pastebin link [14:08] ralsina, nessita, dobey: your input will be very welcome: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/load_test_according_to_platform/+merge/45122 [14:08] ralsina: he, I missed that [14:09] In other news, facundobatista told me they are going to change how queues are handled, so that may impact everyone else's work. He posted a mail about it, IIRC [14:09] ralsina dobey CardinalFang: mumble or irc? [14:09] ralsina: in theory I should be interested since I have to do that on windows at some point, yet I have to have lunch can we talk about it in an hour or 2? [14:09] I have team leads, so +2 hours from now is ok for me [14:09] vds: ^ [14:10] That would be... 4PM UTC? [14:10] should we not discuss it on thursday? [14:10] dobey: vds ould be too bored :-) [14:10] at the normal team call [14:10] I need to go, I'll be back in 30 min [14:10] thisfred, ralsina: wondering about desktopcouch-health-monitor, I think rye wrote something in that direction, also wondering given that is hard to see this from the logs on the server, whether there should be an opt-in way, for us to get those statics regularly from the desktops instead of just for debugging when I problem is found by a user [14:11] This was scheduled to start yesterday. If we delay the planning until thursday, we lose half the window [14:11] * mandel lunch [14:11] pedronis: sounds like a great idea [14:12] ralsina dobey in two hours work for me [14:12] dobey: besides, the weekly meeting aims to be a short as possible and to focus on the roadmap, not specific development [14:12] ralsina: maybe aquarius is also interested in the discussion [14:12] vds: so let's ask him :-) [14:13] aquarius: are you interested? :) [14:13] aquarius: we are having a call about the public APIs definition for 3rd party developers, want to join? [14:13] oh [14:13] * ralsina believes in context ;-) [14:13] awesome bug #539467 is back in Natty :-/ [14:13] Launchpad bug 539467 in pm-utils-powersave-policy (Ubuntu Lucid) (and 4 other projects) "SATA link power management causes disk errors and corruption (affects: 23) (heat: 114)" [Undecided,Fix released] https://launchpad.net/bugs/539467 [14:13] ralsina, yes, yes I am interested. in mumble? [14:13] ralsina, we have team leads in 15 minutes [14:14] aquarius yes, I know :-) [14:14] all meetings should be as short as possible and not focus on specific development problems [14:14] mumble in 1:47 hours? [14:14] ralsina, sure, ping me when you're ready, and I'll be at your disposal [14:14] aquarius, cool [14:14] * ralsina will ping everyone [14:14] * dobey watches the wrench fall through the gears [14:14] hopefully I won't forget anyone this time ;-) === teknico_away is now known as teknico [14:15] aquarius, ralsina I may join, as I wrote in IRC I have some though on dc replication health [14:15] pedronis, cool, adding you :-) [14:15] goody [14:16] dobey is two hours from now good for you? [14:16] assuming it won't be 2 hours long, probably [14:17] dobey: surely not [14:17] dobey: at least drop by so you have some imput on what you are doing the next two weeks, stay as long as you can stand it ;-) [14:18] any other topics? Anyone scared about what I mentioned re: queue changes? [14:19] ralsina: not really :-) facundobatista mentioned they will fake the old behaviour? [14:19] nessita: yes, but it stil may cause subtle problems if you rely on only the top task being running [14:19] or similar things [14:20] nessita, yes, but there is the assumption in the old queues that SD only executes the first item in the queue, and that will be not valid anymore [14:20] ralsina: right, I see the biggest impact on the music store, which uses that to show progress for music downloads [14:20] ralsina: what scares me right now, is this: http://pastebin.ubuntu.com/550243/ [14:20] nessita: that will probably have to be redone to show aggregates [14:20] ralsina: yeah. Do we have a bug # for it? [14:21] not until facundobatista proposes the branch I suppose :-) [14:21] dobey: wasn't thisfred working on that? [14:21] dobey: ask thisfred [14:22] nessita: that doesn't look familiar, I fixed a different bug in the portfinding [14:22] dobey: where did you get that trace from? can you reproduce? is there a bug report for it? [14:22] thisfred: oops then :-) [14:23] which didn't have to do with dbus [14:23] np [14:23] I can help dig for this one [14:23] in my branch i can, but not sure why [14:23] ok then, eom? [14:23] eom! [14:23] although, if i knew why, i guess i wouldn't be worried about it [14:24] dobey: let me know if there's anything I can do to help/falsify/verify [14:24] dobey: if it happens only in your branch, bisect it. [14:25] * ralsina really, really likes bisecting for mistery bugs [14:28] ooh, I didn't even know about https://launchpad.net/bzr-bisect [14:29] so [14:29] if you don't have the desktopcouch package installed, and you try to run desktopcouch from trunk: dbus.exceptions.DBusException: org.freedesktop.DBus.Error.ServiceUnknown: The name org.desktopcouch.CouchDB was not provided by any .service files [14:29] hi everyone! [14:32] so yeah, i can basically reproduce this problem, and now i'm stuck trying to fix it [14:32] brb [14:35] ralsina: unless it's confidential, I'd love to see minutes on what you're discussing on the public API [14:35] karni: sure! [14:35] karni: I don't think it's confidential, but I'll check ;-) [14:35] ralsina: ^ ^ [14:46] ralsina: did you got this? -> "was this the bug to talk with rodrigo_? https://bugs.launchpad.net/evolution-couchdb/+bug/673568" [14:46] Launchpad bug 673568 in evolution-couchdb (and 1 other project) "Error modifying contact, other error when saving contacts (affects: 28) (dups: 4) (heat: 132)" [High,In progress] [14:52] nessita: got it, yes I think that's the one [14:52] ack [15:00] dobey: can you please review (at least) the installation bits for this new (restored?) dbus service in u1client? https://code.launchpad.net/~nataliabidart/ubuntuone-client/auth/+merge/45116 [15:00] * mandel back [15:00] dobey: I added the data/com.ubuntuone.Credentials.service.in [15:01] dobey: and added the lines in Makefile.am to match ubuntuone-syncdaemon's [15:01] dobey: but I would appreciate your review on that [15:04] mandel: on your stabd up summary you said you reviewed my branches... but I don't see you vote on them [15:04] nessita: ping [15:04] nessita: no? I believe i did, it was the no more gobject in sso and some other I cannot remember right now... [15:04] verterok: pong [15:05] nessita: hi [15:05] nessita: I'm looking at https://code.launchpad.net/~nataliabidart/ubuntuone-client/auth/+merge/45116 [15:05] mandel: you could please confirm? links are https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/fix-693531/+merge/44919, https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/no-more-gobject/+merge/44988 and https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/really-errback-on-error/+merge/45016 [15:05] verterok: yes [15:05] mandel: no need to review all three, but those are the pending list [15:05] nessita: should this go into ubuntuone/platform/linux? [15:05] nessita: you are right by vote is no there… I even remember talking with you about the return value... [15:06] verterok: hum... not sure, this is like another independent python package that may turn out into being a separated ubuntu package and all [15:07] nessita: what was the command for hal to tell be the reviewlist? I forgot -_- [15:07] mandel: @reviewlist [15:07] verterok: maybe? I honestly don't know, I would guess ubuntuone/platform/linux was for linux specififc buts for syncdaemon [15:07] bits* [15:07] nessita: mandel might know better, as he's the win guru [15:08] nessita: ok, is the credentials module going to be imported by syncdaemon? [15:08] verterok: I'll ask [15:09] mandel: so, I built this dedicated login dbus service for u1client that abstracts the caller of setting all the specific bits for ussoc [15:09] nessita: do you need me for anything? [15:09] mandel: right now I added that into ubuntuone/credentials [15:09] mandel: is a pure dbus service that depends only on ussoc [15:09] branch is https://code.launchpad.net/~nataliabidart/ubuntuone-client/auth/+merge/45116 [15:09] * mandel looks [15:10] mandel: shall that go into ubuntuone/platform/linux ? I think not, since we may want to build a separated ubuntu package for it [15:10] mandel, nessita: I was wondering if it will affect the windows client as it's a module in the ubuntuone namespace [15:11] verterok: I would hope that the python packages can be filtered when building the window stuff just like we do on linux [15:11] from a same source tree we generate serveral ubuntu packages and even leave stuff outside (such as test suites) [15:12] nessita: I have no idea, that's why I'm asking :) [15:12] yeah, thanks for pointing this out [15:12] verterok, nessita: I'd say leave it there, do not worry to much about windows, I'll make sure I clean it after [15:13] mandel: k, thanks! [15:13] I'm also writing a dbus compatible API for windows, so we might not need to touch a thing :P [15:15] mandel: yeah! [15:15] nessita: did you manage to take a look at the decorators branch? [15:17] I'd like to start moving desktopcouch to use them so that hte tests can run on windows, which I think is imperative [15:21] mandel: a little, is there a merge proposal for it? [15:22] nessita: yes, https://code.launchpad.net/~mandel/ubuntuone-dev-tools/load_test_according_to_platform/+merge/45122 [15:22] nessita: you should be happy with the tests of the decorators ;) [15:22] Aloha [15:22] mandel: I will review after lunch [15:22] nessita: sure [15:22] hi czajkowski [15:23] nessita: ups, that sounded terribly sarcastic, it was not :) [15:23] I've been having an issue for the lat few days wiht music syning to my folder, I know there were issues before Christmas but the status of a few songs are like this sinc eI bought them a few days ago http://twitpic.com/3mzger/full [15:24] czajkowski: best things is that all the song title could be applied to an error hehe [15:24] el mas duro => to hard to download [15:24] mentere => I know I have to download [15:24] :P [15:25] czajkowski: are those songs in your account if you visit the web ui? https://one.ubuntu.com/files/ [15:27] mandel: nessita yes they are [15:27] mandel: oi smartie go work on your bugs! [15:28] czajkowski: so you have all your songs in the web ui but not in your computer? is that the issue? [15:28] czajkowski: hehe I could not resist :) [15:29] nessita: all songs are down, just still getting that message in rhythmbox [15:29] mandel: I'll assign all the bugs to you if you don't behave! [15:30] my big annoyance with U1 is that it's after seperating all of one album into many albums as it's a main artist featuring another artist which is a rater annoying === hito_jp0 is now known as hito_jp [15:30] * mandel behaves [15:31] czajkowski: that's odd, I'm pinging another dev wil more knowledge about the rhythmbox plugin [15:32] http://twitpic.com/3mzjfw/full [15:32] thats one album of Cheryl cole split into different albums as it's featuring another artist [15:32] mandel: :) [15:33] that is indeed a PITA [15:34] czajkowski, I get what you mean. The very idea of an artist owning an album doesn't make sense. We should have something like book anthologies, where the catalog lists it by editor. [15:34] mandel: yup, more annoying than that was not being able to buy it in Ireland but had to come back to UK to buy it! :s [15:35] CardinalFang: or just put the entire album in one album even if it features other artists. [15:35] as it;s the one song in the main album [15:35] hi muffinresearch, czajkowski is having issues with the RB plugin and/or server part [15:35] muffinresearch: he sees his rb like http://twitpic.com/3mzger/full but all his songs are already on the web ui and on his computer [15:36] nessita: hi I'm Laura :) [15:36] czajkowski: ah! sorry :-D [15:36] nessita: no worries, happens a lot [15:36] *her* computer :-) [15:36] czajkowski: your nickname doesn't look female, sorry :-/ [15:36] czajkowski, the problem is that there's no albumness any more. You get to buy individual songs, and there's nothing binding those songs together. The idea that one can look at the song artist doesn't really work, as you're finding out. [15:37] nessita: really it's fine :) it happens at least once a week, I clearly need to blog more or something:) [15:38] nessita: wow! I hope mine does :) [15:39] nessita: it looks polish? [15:39] CardinalFang: there are no tags to identify an specific album? [15:39] karni, got the latest apk for me to test? [15:40] CardinalFang: well it does, look at the album here http://twitpic.com/3mzlcp/full puts it into the album folder, the issue is when an artist colaborates with an another artist on the album http://twitpic.com/3mzlcp/full [15:40] czajkowski: if you put an 'a' on the end of your nick, all the spanish will think it's female at least :) [15:41] dobey: I am not changing my nick :) [15:41] beuno: i'm polishing minor things, but I'll provide you the link in a minute. [15:41] karni, awesome, thanks [15:41] i'd change my nick, but irc and unicode don't mix well [15:42] muffinresearch: ping? [15:42] dobey: it happens with the verb put, if you put an a at the end, we all think is a female :) [15:42] dobey: problem is in ireland 30+years ago, not many poles and I already had an non irish name I was not going to have a different name to my dad, plus my nick is very me now :) [15:42] dobey: hi there, did you got my messages re new dbus service for u1client? [15:43] czajkowski: hi, I'll fille a bug; we'll need to look into it further to understand why there's a disconnect happening. /cc nessita [15:43] muffinresearch: thanks! [15:43] muffinresearch: thanks! [15:44] glad to see it's not just me going batty :) [15:44] thisfred, can you please come to mumble for a minute? [15:44] will do [15:44] muffinresearch: can you add me to the bug,- launchpad id czajkowski also thanks [15:44] czajkowski: i was mostly just poking fun at my spanish comrades :) [15:45] czajkowski: sure, will do [15:45] dobey: oh I understand that, tis easily done to do it to mandel [15:45] * ralsina points at his nick for dobey [15:45] nessita: maybe, but i am trying to solve this drastic problem with desktopcouch [15:45] dobey: ok, I'll wait [15:46] like, why the heck is desktopcouch not showing up on the bus at all :( [15:47] beuno: http://ubuntuone.com/p/Wee/ There's a problem downloading from Shares. Apart from that, only minor glitches (work in progress) [15:47] beuno: Let me know if you see your UDFs/Shares (/me prays) [15:48] mandel, there could be some identifier. MusicBrainz metainfo would be best. But ID3 tags in files have "album" as a 30 character string. There's nothing there to say a song's album named "X" is the same album as for another song with album "X". Imagine the headache of grouping all songs together that have album "Greatest Hits", for instance. [15:48] beuno: uninstall required [15:48] beuno: I might have shuffled few values (FileStatus) in content provider [15:48] CardinalFang: +1 on musicbrainz id [15:49] dobey, I saw something like that this weekend. I'm trying to eat more of my own dogfood, so abandoning all the sqlite for couchdb. [15:49] ralsina: when, you are not a woman!!! why did I take those pict them... [15:49] dobey, The next connection attempt succeeded, FWIW. Didn't see the cause yet. [15:50] mandel: you saw me in person. if I am a woman, I am incredibly ugly :-) [15:50] CardinalFang: i'm having trouble finding the cause too; but i can reproduce the lack of dbus easily. see #desktopcouch [15:50] CardinalFang: uh, I though that ID3 tags had more info than that.. I should read more about that [15:51] mandel, id3v2 is smarter. I hope we're using that. [15:51] ralsina: well, I went back to spain saying terrible things about the ralsina the woman I met in argetina… I even sang you "I feel preaty…" what I waste! [15:52] CardinalFang: I'm so out of date about those things… I'm slightly embarrased about that [15:52] cardinalfang: hello! Are you available for mumble in... 9 minutes? We need to discuss the 3rd party API work with a bunch of people, you included :-) [15:53] ralsina, yes, I'm in. [15:53] mandel, confirmed, we're writing ID3v2 (2.3.0). [15:53] CardinalFang: cool. I'll ping everyone else in a few minutes [15:53] aquarius: Wanna testdrive? U1F: http://ubuntuone.com/p/Wee/ Download from Shares broken. Apart from that, initial sync / upload / download in place. [15:54] aquarius: I'm mostly interested if (after sync finishes) you see content of UDFs and Shares [15:54] karni, testing! [15:54] beuno: thank you [15:54] ralsina: mumble? [15:54] mandel: yes [15:56] aquarius mandel vds pedronis cardinalfang dobey nessita : mumble for 3rd party APIs talk in 5 minutes [15:56] beuno: in case you want to send logs, please enable verbose logging in Menu->Settings->Debug settings [15:56] karni, ack [15:56] initially, you're running ProductionLogger, with less I/O overhead [16:00] czajkowski, if you're comfortable with it, I'd like to see the result of running id3v2 ~/.ubuntuone/Purchased\ from\ Ubuntu\ One/*/Recovery/* [16:00] czajkowski, you may need to install package "id3v2" first. [16:00] karni, looking *great* so far! [16:00] beuno: got you your UDFs?? [16:00] karni, 57% progress [16:01] waiting [16:01] CardinalFang: sure just tell me what you need me to do so [16:01] but the account info loads perfectly [16:01] and smooth so far [16:01] aquarius mandel vds pedronis cardinalfang dobey nessita : mumble for 3rd party APIs talk now [16:01] beuno: ah, ok. /me has fingers crossed. you never got to UDFs, which worried me back then. [16:01] CardinalFang: you sure that's the right package? [16:02] czajkowski, no. [16:02] nm typo [16:02] yes. [16:02] installing now [16:02] beuno: btw we'll improve that sync speed with less I/O using transactions, like mentioned before. not implemented yet. [16:02] aquarius: sorry :-/ [16:03] CardinalFang: ok installed [16:03] karni, still no udfs, but got some shares and it's still syncing [16:03] czajkowski, From a terminal. $ id3v2 ~/.ubuntuone/Purchased\ from\ Ubuntu\ One/*/Recovery/* [16:03] It might be a few pages worth. [16:03] beuno: now that's a mystery to me.. // i'm fixing download from shares, just found the problem. [16:04] CardinalFang, FWIW, we're a bit backed up with file scanning on the server [16:04] should be back to normal in about 2 hours [16:04] CardinalFang: nope wants me to add stuff to it, let me pm you with the output [16:04] karni, it seems to be downloading shares first [16:05] so once it finished that, I may get udfs [16:05] also, we need to speed this up 1000x :) [16:05] will transactions do that? [16:05] beuno: I hope to reach 5-6x. What worries me more is authentication, it's too heavy. It's always been to heavy.. [16:06] beuno: I'll implement transactions soon [16:06] karni, awesome [16:06] this is good enough for now [16:06] just throwing it out there [16:06] beuno: and well see what we've got [16:07] karni, sent logs. no udfs [16:07] it's stuck at 100% origress [16:07] beuno: thank you. feels like time to file a bug myself. [16:07] beuno: don't clear the notification yet [16:08] beuno: it might be the service/java future. notification is only for us, give it one moment [16:08] ooops, closed the app and opened up again [16:08] beuno: nah, no problem :) [16:08] * karni cheks download fix [16:08] no progress, no udfs [16:10] karni, and I didnt get all shares [16:10] http://pastebin.ubuntu.com/550288/ [16:10] is the output that CardinalFang asked me to get from my U1 [16:11] beuno: the sync could have been interrupted. I'll make it a foreground process, I'll take care of it. [16:11] czajkowski, line 229 seems to show what the "album" key should be, at least for what we, Ubuntu One, produces. [16:11] mandel, ^ [16:12] * CardinalFang is now on a phone call. [16:15] verterok: this doesn't look good :< I get NPE when downlading from shares http://paste.ubuntu.com/550291/ [16:15] verterok: line 44 -- should 3rd parameter to new QueryItem be the shareId ? (share volumeId?) [16:16] beuno: thanks for testing. I'll get back to you with a custom verbose logging build some time soon. [16:17] beuno: anything else heavily failed? [16:18] karni, you should not be using Query, as you'r using generations [16:19] verterok: I thought so. [16:19] :< [16:19] verterok: You think you could find a little time for documentation? Or direct me for documentation of python sources perhaps? [16:19] * kklimonda hugs karni [16:19] you should get a medal for writing Java code ;) [16:20] verterok: I'm kinda coding blid :< [16:20] karni, you have all the required info in the node deltainfo objects [16:21] kklimonda: It's not that. It's just u1-java-sp has never been a real project (still laying around in verteroks +junk :D ), so I've got little issue with docs, and bothering verterok all the time (He's really patient) [16:21] verterok: oh crap... right [16:21] karni, there are no docs for the python stuff either [16:21] karni, only code :/ [16:22] karni: meh, I just find Java.. repulsive ;) [16:22] verterok: I just realized I have everyting I need.. I can simplify that code! [16:22] verterok: ack :( [16:23] verterok: Tell me one thing. Sometimes I see in method signatures 'share'. Is it a second name for a volume? Or shares have some another identifier? [16:23] verterok: For instance, in that QueryItem(node, hash, share) [16:23] kklimonda: thanks for thumbs up ;D [16:24] karni, nope. I uploaded something, but didn't get feedback [16:24] I think you know about that, though [16:24] beuno: yes. however I'll ping you with a more verbose apk, I wanna now every detail that happens on your side :) [16:25] cool [16:25] beuno: plus, turns out I can simplify some parts, generations are still new thing, and quite often I end up dropping another part of code and use generations. [16:26] karni, shares are a kind of volume [16:26] verterok: aha? [16:27] A share id is a volume id :) [16:27] verterok: oh.. so whenever I see shareId, I should provide volumeId ;) [16:27] Sounds good. [16:28] Karni, yes share id is a legacy stuff from the protocol :/ [16:28] verterok: roger === beuno is now known as beuno-lunch [16:35] dobey: ping [16:35] karni, I am testdriving :) [16:36] karni, the notification area says "Progress (per volumes) 53%" and is counting up, slowly. What's it doing? [16:36] aquarius: if you've got much content (and I know you do ;)), sync is slow. we'll boost that later. authentication is slow.. (protocol/auth fault :<). I'm interested in UDFs in particular. [16:36] aquarius: fetching meta of the files [16:36] karni, also, I went into UDFs -- the spinner spun for a while, and then stopped, but the screen is blank [16:37] aquarius: it's slow, because each sqlite is transactioned [16:37] karni, perhaps it's still syncing [16:37] aquarius: yes, still syncing probably [16:37] *each write [16:37] karni, but this is now 2 minutes or so -- it takes that long just to show me the names of my UDFs?? [16:37] aquarius: any progres indications? [16:37] karni, not in the app. he notification area is not up to 72% [16:38] aquarius: it's syncing ~u1, udfs, and shares in that particular order [16:38] aquarius: progres indication is per volume, so it'll jump to 0 few times (yup, not good. that's slow.) [16:39] karni, so basically I just have to live with the pain the first time i start the app? [16:39] notification area now at 85% [16:39] aquarius: meta of each element consists of: volume, parent, node, hash, modified, name, etc, etc. it's the writes that slow it down, at least to some extent [16:39] karni, my Files are listed, if I go into Files [16:39] UDFs still blank [16:40] aquarius: yes. I had already a screen with a demo slides (empty slides ;P) sliding in, to spend that time in less pain [16:40] (note: UDFs is internal terminology. They're called "Synced Folders", although I'm sure beuno will have already mentioned that :)) [16:40] aquarius: in theory, you can start using it already. however not all items are yet visible (due to ongoing sync) [16:40] nessita: yes? [16:40] aquarius: ah right.. have to fix the dashboard label [16:41] aquarius: ok, so what's the main catch. [remember, we'll speed up the sync by factor of 3x - 5x maybe] the catch is that [16:42] dobey: can you please tell me what you mean with "The translation bits are wrong here. The correct way is already done in clientdefs.py.in"? you mean that I should use that Q_ function? [16:42] aquarius: instead of waiting for each folder, like AU1 did, you (indeed) spend that terrible time waiting for first sync [16:42] aquarius: but then you can browse all your content in a blast. all meta is there, on your phone. [16:42] nessita: yes, do the Q_ thing, and don't do gettext.textdomain() [16:42] aquarius: you can download stuff, will be able to pick what you want to sync, etc [16:43] dobey: most of the UI I've seen have the gettext.textdomain(GETTEXT_PACKAGE) at module level.. why is that wrong? [16:43] karni, do you really need to sync everything the first time? I mean, can you sync the names of my UDFs first, so they show up, and *then* sync the content metadata? Also, having the UDFs screen say "still syncing" rather than being blank would be good -- I can't tell whether it's blank because it hasn't finished syncing or because it's broken :P [16:43] aquarius: or get to a file that is particularly deep in folder hierarchy - not having to wait for each folder to load. [16:43] dobey: even preferences doesn't use the Q_ function [16:43] nessita: because it screws up other applications if they import that module [16:43] dobey: ah... right [16:43] nessita: preferences is all done in the script [16:43] the Q_ method is for use in library code [16:43] dobey: so, for this particular case, who will be doing the setting of the textdomain? [16:44] aquarius: right, I'm sorry :) initial sync is one of those this that needs some love. initially we thought of even *not* showing the dashboard at all, but some demo content, whatsoever [16:44] nessita: nobody. Q_ passes in the correct domain to use [16:44] aquarius: the notification is only for debugging, we'll have much better sync notification implemented. [16:45] karni, ok, "progress" in the notification area is now at 100% (it went back to 0 a few times, indeed), and going into UDFs is still blank but it has a spinner [16:45] dobey: ok, changing that then [16:45] aquarius: and the anwer to your question is -- it's your backend :( the getDelta returns *all* file meta. we can't just fetch the filenames, unless you guys would implement that :( [16:45] aquarius: the spinner definitely means it's still working. [16:45] karni, ah, i don't mean fetching any contents at all, but you can, I think, get the names of the volumes and show them in the UDFs list quickly -- so I can see the names of my UDFs at least [16:45] aquarius: (or that's a bug ;d but the spinner is 99.6% reliable, the notification is crap ;d) [16:46] aquarius: aha, that's what you mean. [16:46] karni, just so the app isn't totally blank for the first 10 minutes :) [16:46] aquarius: so. the first thing a client does is fetch the volumes [16:46] aquarius: hehehe [16:46] dobey: I've fixed and pushed the changed for https://code.launchpad.net/~nataliabidart/ubuntuone-client/auth/+merge/45116 [16:46] aquarius: then, for each volume, we fetch the meta. one of those elements is an unnamed element, the volumes 'node', right [16:47] aquarius: the ContentProvider has a Files table, that contains all the file data you see. [16:47] aquarius: getDelta returns stuff in parent -> child order. so what I could do is implement something like a 'shallowSync' [16:48] aquarius: that would fetch only those unnamed nodes, which would appear in the Files lists (~U1, UDfs, Shares) [16:48] aquarius: but still we'd need to call the real sync method (which takes 10 minutes, at least for you, and for now). [16:49] aquarius: we can definitely work on the order of sync, remembering it's a DFS, afaik [16:50] aquarius: coming back to bleeding edge testing - you see any udfs/shares ? [16:52] karni, I still have the spinner on the UDFs screen [16:52] verterok: hmm... getDelta is volume-based, right? currently, we can't drop the download of delta halfway, right? [16:53] aquarius: I'll send you guys a pre-compiled verbose version. also, I'm sure you have sdk installed. you can jump in $ adb logcat to see how's it doing [16:53] karni, yes getdelta is per volume [16:53] aquarius: this night (last night) the service was unbelivably slow, but I heard guys where aware of that issue. [16:54] karni, ENOPARSE the drop download part [16:54] verterok: I'm starting to think I'm going have to heavly customize the Client for mobile/speed optimization.. [16:54] nessita: i'll look at it again as soon as i can. i feel like i am chasing schroedinger's cat right now [16:54] verterok: imagine we only want to get the name of first (top) folders of the volume [16:55] verterok: say, content's of ~/Ubuntu One. can we drop (not download) the rest of the meta? [16:55] verterok: that would have to be implemented in the Client, right? [16:55] dobey: thanks, and good luck [16:55] karni: what do you mean with "Client"? [16:56] finally, no more mumble for today. Maybe :-) [16:56] karni: no, you can get a partial view of a volume [16:56] verterok: storageprotocol.Client.java , or even the Request(s) themselves [16:56] verterok: can? or can't? (to be clear) [16:56] karni: hmmm [16:57] karni: sorry, *can't* [16:57] aquarius: ↑ [16:57] karni: what would you need that? [16:57] or why...aactually [16:58] * verterok reads the backlog [16:58] * karni takes a breath, that's some good brainstorming [16:58] verterok: no need. wait, I need to collect my thoughts. [16:59] karni: so, the problem is the android client being slow? [16:59] aquarius: ok, even if I'd make the UDFs/Shares names appear (but not the ~U1 names, which are already volume content), would that do any good? [16:59] verterok: sync is partially slow due to untransactioned sqlite writes. however, [17:00] verterok: I'm not sure how much I'll cut the time using transactions (3-5x at most) [17:00] karni, it'd show that the app works. At the moment, I don't know whether the U1 app is working fine and is just slow, or whether there's actually a bug in it [17:00] one second... [17:01] ralsina: I forgot to ing you about https://code.launchpad.net/~mandel/ubuntuone-dev-tools/load_test_according_to_platform/+merge/45122 [17:01] verterok: we still use getVolumes method, while using generations. to get the volumes information, right? (Just wanna be clear about that, as generations have changed much) [17:01] karni: yes, you request the volumes, and the for each volume a getDelta [17:01] mandel: I'll queue it :-) [17:01] verterok: thank you [17:02] ralsina: haha there are to many [17:02] aquarius: brb checking [17:02] mandel: better too many than too few :-) [17:03] ralsina: yes,usually it is like that [17:03] nessita: I approved a couple of your branches in sso, you need to change the status [17:03] mandel: THANKS! [17:03] karni, still got the spinner. [17:03] aquarius: we can show UDFs/Shares name quite fast. [17:04] nessita: tell me if there are any of those that are urgent and I'll take care [17:04] aquarius: drop it. kill the app :) I'll get back with another apk. I'm sorry. [17:05] aquarius: beuno-lunch: ok, so current focus is -- transactioned sync writes to database (we'll see how much we cut), and showing UDFs/Shares names soon after sync starts. Sounds good? [17:06] aquarius: If we don't cut enough, I'm seriously considering a separate *activity* for the initial sync. Even a good, foreground notification (foreground = persistent) won't be good enough for first few-minute meta sync. [17:07] karni, have killed and uninstalled [17:07] * nessita -> lunch [17:08] aquarius: good. I'm in the middle of download modification, so I'll ping you later, ok? [17:09] karni, no problem. It looks *great*, I have to say :) [17:09] aquarius: I'd prefer it to _work_ great.. I'm really still unsatisfied :< But thank you for your kind words. [17:09] aquarius: I'll do my best. [17:26] * nessita is back [17:27] dobey: when you have a moment, I would like to ask advice on how to handle the migration preferences -> control panel (in particular, how to handle package deps/conflicts for the binary of our icon) === beuno-lunch is now known as beuno [17:28] karni, sounds GREAT [17:29] beuno: focus set. I'm on it. [17:30] nessita: ok; i'm slowly but surely making progress on this desktopcouch + dbus bug, at least === IAmNotThatGuy is now known as M0hi === Franxesk is now known as Franxesk_afk [19:02] nessita: no-more-gobject has the two approvals now. Moving to really-errback-on-error... [19:02] ralsina: awesome! [19:05] Ouch, nessita, I just noticed no-more-gobject has a few pylint warnings. [19:06] ralsina: it does? I'll fix before marking it as approved [19:06] But I noticed after I already approved [19:06] nessita: It's these: https://pastebin.canonical.com/41489/ [19:08] ralsina: you don't have disutils-extra installed [19:08] ralsina: that's why I didn't get any lint warning :-) [19:08] PASSED (skips=2, successes=416) [19:08] nessita@dali:~/canonical/ussoc/review_no-more-gobject$ [19:08] nessita: makes sense :-) [19:08] nessita: is that marked as a build dep? [19:09] ralsina: there is not meta package for build dep for ussoc [19:09] only for u1client [19:09] nessita: ok then [19:09] (and this is ubuntu-sso-client) [19:33] anessita: approved really-errback-on-error [19:34] aralsina: gracias! [19:34] ade anada :-) [19:35] mandel: approved load_test_according_to_platform I am assuming it's not worse than when I tested it this morning ;-) [19:46] ralsina: is not :) [19:46] is even better it has not lint issues [19:46] mandel1: then you just need to find another fool^w reviewer ;-) [19:47] * nessita is reviwing that [19:48] * ralsina is being told XML is human editable. Yeah, right. [19:52] nessita: thx :) [19:52] mandel1: pep 8 issues! [19:52] nessita: arg, really? dammed! [19:52] ./ubuntuone/devtools/testcase.py:64:13: E301 expected 1 blank line, found 0 [19:52] ./ubuntuone/devtools/tests/test_decorators.py:42:41: E202 whitespace before ')' [19:52] ./ubuntuone/devtools/tests/test_decorators.py:56:13: E301 expected 1 blank line, found 0 [19:53] nessita: besides those, is the rest ok [19:53] * mandel1 goes to fix them [19:53] mandel1: don't know, I first run the tests and then review code :-) [19:53] uh, ok hehe [20:06] dobey: you around? [20:07] nessita: errors fixed, what do you think about the rest? [20:07] mandel1: I'll look now, I was waiting that push [20:07] nessita: oh, ok [20:10] mandel1: all the try-except for the imports should go at the top, with the other imports [20:11] nessita: I wanted to put them close to the test so that it made sense to read it, but if that is an issue I can be easily changed :) [20:11] mandel1: yes please. Also, the docstring [20:11] """ [20:11] Decorate the test so that it is skipped. [20:11] """ [20:11] should be [20:11] """Decorate the test so that it is skipped.""" [20:12] Is that ok, grammatically? Just curious. [20:13] not sure, but I'm no english teacher... I would personally write: """Decorate the test so that it gets skipped.""" [20:13] Both are great. [20:14] yes, I think they are the same [20:15] but my english got worse in Manc so I'd not be surprised if I write seomthing bad :P [20:15] :-) [20:22] nessita: do you want me to push the changes or are you ok with me waiting to see if you have any other comment? [20:22] mandel1: I'm done with the code, the rest is a +1 [20:22] :) [20:23] I'll push the style fixes then [20:39] beuno: aquarius: Debug is on by default, but you'll have to check it to send the logs (aq, you can send the logs from settings). Please test 2 things: initial sync speed, and presence of UDFs. Sync progress is gone temporarily (optimizations!), but the notification will disappear once sync has finished. http://ubuntuone.com/p/WhE/ (uninstall first!) [20:40] karni, on it! [20:40] beuno: aquarius: having 141 files in cloud, I've cut time from 55s down to 7s (7.8x, not bad) [20:41] but I know you guys have tons of files ;) [20:41] karni, you should add in a few thousand files [20:41] feel the pain a bit more [20:41] beuno: PS you can poke around the app, but don't leave it before sync finishes (I haven't moved the service to foreground yet, and I don't want it to get killed) [20:42] will d [20:42] do [20:42] beuno: I imagine. Right now we're using *raw* sql, so if we need more speed - you have to modify the servers ;) [20:42] karni, lets give it a try [20:43] I'm aware ppl sync tens of gigs. [20:43] another alternative is to sync in top-level items [20:43] and bring in the rest on demand, right? [20:43] beuno: we can't.. [20:43] we can tweak the server as well, I'm sure [20:43] beuno: 1 sec [20:43] not sure how much margin we have there [20:44] karni, syncing! [20:44] beuno: yes, I think we can 'sync' just top level -- a list of UDFs and Shares only. no ~U1 [20:44] * karni awaits [20:45] youre now at 150+ files [20:45] karni: I'm looking at the code...and it seems it's always doing a rescan from scratch...o every volume. is that right? [20:45] beuno: btw I was testing on wifi, which is preferred for initial sync. naturally.. [20:46] verterok: I've changed stuff in last 2 hours. [20:46] karni, it finished *much* faster [20:46] verterok: initial sync now makes no queries/updates whatsoever [20:46] like in less than 2 minutes [20:46] and I have udfs [20:46] beuno: 1:22 ;) [20:46] verterok: I use transactions, raw sql, and inserts only [20:47] karni, you get logs automatically? [20:47] verterok: so we're 7.8x faster ;) [20:47] mandel1: your branch, got my approve. After last pushes, you approve the proposal itself [20:47] beuno: not really, no. you have to send them, if you like [20:47] karni: cool [20:47] mandel1: cool, thx! [20:47] karni: I'm looking at the volumeNewGenerationCallback client callback [20:47] karni, it all works [20:47] beuno: please send whenever you have any comments/bugs/etc [20:47] shares, udfs [20:47] * mandel1 is ready to start last touches of desktopcouch on widnows :) [20:47] beuno: yes! [20:47] I have nothing to complain about! [20:47] finally [20:48] karni: and isn't using the generation number on the notification, following the code, it ends up calling getdeltafromscratch [20:48] verterok: verterok, my dear friend. definitely the volumeNewGenerationsCallback is just a test ^ ^ haha [20:48] karni: ah, ok [20:48] verterok: please, don't mind that sync invocation [20:48] karni: that will put a lot of stress on the servers, and make all go slower [20:48] verterok: I was focusing on sync speed only [20:49] karni: oh...mmm. yes, that's quite related to sync speed or I'm missing something? [20:49] verterok: tell me what I should do. if we can a volumeNewGenerationsCallback, I simply request delta for that volume? [20:50] verterok: no, I'm talking particuarly about *initial* sync. the client does not get that callback upon first connection [20:50] verterok: what it does it uses getVolumes [20:50] verterok: or actually it's zippy version that I wrote today [20:50] :D [20:50] verterok: before I push it, I'd like to clean things up [20:51] karni, crashed like a very crashy thing indeed [20:51] karni, http://pastebin.ubuntu.com/550378/ [20:51] aquarius: whoa crap, tnx [20:51] I should note I'm on Android 2.3. :) [20:51] verterok: http://paste.ubuntu.com/550377/ this gives us the boost [20:51] aquarius: aha! xD [20:51] aquarius, http://ubuntuone.com/p/WhE/ [20:51] karni: nm, I see you end calling getDelta [20:51] is that the version you're on? [20:52] beuno, rev19-sync-boost [20:52] right [20:52] yes [20:52] *end up [20:52] aquarius: could you please send the logs :)? [20:52] I'd like to have them. [20:52] karni, send them where? that's what that pastebin is [20:53] aquarius: ah, nvm :) that's good. however, you've got 'DEbug settings' in the Settings of the app [20:53] aquarius: those logs are really verbose, and might prove helpful. now, I'll have a look at that paste. [20:53] karni, will try again [20:53] aquarius: strange o_O your database is locked, and it's just calls to CREATE TABLE [20:54] aquarius: I'm pretty sure that part is rock solid [20:54] karni, ok, it now works perfectly. I shall uninstall and reinstall :) [20:54] aquarius: ah, right :) great [20:54] aquarius: let me know how do you feel about the sync speed [20:55] * karni thinks it was an awesome day. I need a shower, lunch and sleep, but it was definitely worth it. [20:55] ok, launching app (after clearing, uninstalling, reinstalling) for the first time now [20:55] ok [20:55] * karni likes aquarius'es irc-verbose testing [20:55] :) [20:55] crash again [20:56] not good [20:56] 2.3 is buggy! ;P [20:56] but seriously. ok, so you can't send in the logs.. I'll test it in the emulator :) [20:56] ok, can't send the logs because it crashes [20:56] right [20:57] alogcat FTW? [20:57] does it verbosely log to somewhere magic? or to the android syslog? [20:57] aquarius: ok, thanks for testing. will get back to you (some time later than today) with updates on 2.3 [20:58] aquarius: it's writing to app_folder/logs/here*.log [20:58] karni, hm, app now works (it seems to crash until sync finishes, and hten the app works OK), but my debug verbose setting is turned off [20:59] karni, so I don't think I have debug logs. Something weird going on there. [20:59] aquarius: i needed better logging 'framework' than what I knew about Log. so I wrote it. Plus, I can filter private data, etc, the way I want. [20:59] aquarius: you have. I've sent you guys a debug build, doesn't matter it was off. [20:59] aquarius: but indeed it was strange to crash before sync started [21:00] karni, I'm not sure; it might have crashed after sync started. I started the app and pressed Files -- the screen slid over to a black screen and hung there for a while (with U1 in the notification area) and then I got the Force Close window [21:00] karni, and after sync had finished and I went back into the app (and it was working), the "verbose logs" checkbox was unticked, and the "send logs" entry was disabled. [21:00] aquarius: Aha. I think I know the reason.. [21:01] aquarius: right, you have to tick it to send the logs. but this build logs them anyway. this checkbox is for production purposes [21:01] aquarius: normally, I would send the 'production'build, but I wanted your logs this time. so it was logging right away. [21:01] karni, I suggest you leave verbose logs on by default for now [21:02] beuno: right, I'll fix that (and the checkbox) [21:02] aquarius: so, about the FC [21:02] karni, ok, so...should I reinstall, make it crash, wait for sync to finish, go back in, tick the box, and then send the logs? [21:02] (ticking the box won't erase any already-existing logs, will it?) [21:02] aquarius: there are transaction writes (45 writes each) during sync, this produces such overhead, that the Files activity couldn't query the content provider for the contents :> [21:03] karni, you're DoSing *yourself*? haha. :) [21:03] aquarius: 1. it won't erase 2. just tick and send :) they sould be in place [21:03] nessita: hey, sorry. what's up? [21:03] aquarius: what I mean is, I squeeze all the power I can during sync fron the sqlite [21:04] aquarius: we can try different values, I tried 30 and 60 for the transactioned writes. 45 turned out to be faster than both of these. [21:04] aquarius: I'll probably write tests one day to test that with the framework. [21:05] beuno: you finally see UDFs :) I've spotted few glitches during today's coding, so.. that's good. I'll make sure all runs fine. [21:06] karni, yeah, I see everything! [21:06] aquarius: out of curiosity, how much Gigs of content meta have you just synced with your phone? [21:06] aquarius: i.e. how much you hold in the cloud, if that's a sensible question [21:06] karni, I've got about 18GB of files in U1 [21:06] awesome [21:07] dobey: I wasnted to ask about ubuntuone-dev-tools, is that dep going to some meta package? Also, I was hoping to get some guidelines from you on the .desktop file thing between preferences and control panel [21:07] ok, I am EOD [21:07] bye bye! [21:07] karni, ok, sync has finished, went back into app, did Send logs [21:07] bye ralsina [21:07] aquarius: thank you! [21:08] nessita: no, not really in a metapackage. [21:08] ralsina: bye bye! (so the APIs are confidential ;) !) [21:08] dobey: shoudln't it be? [21:08] karni, they don't seem very useful, though. [21:08] :D [21:08] nessita: what are the differences exactly? do we expect -preferences and cp to both be there? [21:09] nessita: i don't think so, why would it be in a metapackage? [21:09] karni, lots and lots of android.database.sqlite.SQLiteException: database is locked in the logcat [21:09] dobey: nopes, cp should replcae preferences since natty. I would like to make the transition smoothly [21:09] aquarius: roger. will definitely look into that [21:09] karni, after SELECT locale FROM android_metadata failed [21:09] dobey: like the deps for building u1client and such. [21:09] aquarius: ok, now that's strange.. [21:09] nessita: apt-get build-dep does that [21:10] aquarius: it only contains one entry 'en' like English, so I wonder if it's actually not android playing with us [21:10] karni, http://pastebin.ubuntu.com/550389/ [21:10] dobey: yeah, but verterok says that apt-get build-dep will not install ubuntuone-dev-tools [21:11] aquarius: I'll check that. However, it's not been once that we had issues with bugs on android :< [21:11] aquarius: hm.. looks like I keep it to busy with the writes :D ! [21:11] nessita: at least not in lucid (using nightlies) [21:11] nessita: depends on the project. ubuntuone-client doesn't use it yet [21:11] and yes, nightlies needs to be added [21:11] dobey: ah, my credentials branch does uses it, I saw no point on adding new code accessing the "old" contrib [21:11] karni, I assume by "it's not been once" you mean "it's been lots more than once that we've found android bugs", rather than "it's not even been once, in fact it's been ZERO times"? :) [21:12] aquarius: You get those during normal app operation? (not during sync?) [21:12] aquarius: Right. I actually mean at least 2 (I thought I'd fail GSoC because of one! ;P ) [21:12] karni, after the app force closes, if I start the app again while sync is happening, the app runs for about a second and then force-closes again. [21:12] karni, once sync has finished, the app works fine [21:12] nessita: oh, i think we should do that switch in a different branch. but yes, things don't build-dep on stuff they don't yet require [21:12] aquarius: aaaaah ok ok ! my man, sorry [21:13] aquarius: solution is: [21:13] karni, it might be an android 2.3 issue. it might also be the phone -- I've got a nexus S, which is really stupidly fast :) [21:13] aquarius: this build performs full sync on each start :D ! it's because I only wanted you guys to test it once, the sync speed ;) [21:13] dobey: I'm not migrating existent code to use devtools, but I think new code should use it [21:13] to ease the future migration [21:13] aquarius: you can't operate the app well really, it might be because of _initial_ sync every time you start it [21:14] nessita: i don't think we should write new tests that use different test runners than the project is already using [21:14] aquarius: good :) fast phones are good :) [21:14] but anyway, adding a build-depends on something that isn't actually required yet is wrong [21:14] aquarius: anyhow, I'll ping you tomorrow/the day after with a better build. that doesnt use "zippyInitialSyncVolumes()" every time ;) [21:14] karni, ok, cool, good work! [21:14] aquarius: thank you! [21:15] dobey: ok, I can change it, should be trivial. Speaking of which, would you please re review that branch? I'll change the runner in the mean time [21:15] beuno: would you like me to list/report my recent activities anywhere today? [21:17] nessita: as soon as i finish this desktopcouch fix. i had an appointment that went over a bit, so i'll be working a bit later today to make up for that [21:17] karni, sure, send a basic introduction to the -users list [21:17] and a the list [21:18] beuno: you mean androidu1-users? [21:18] or ubuntuone-users? [21:18] nah.. I probably got it wrong [21:18] karni, androidu1-users [21:18] beuno: they know who I am, it's just been a very little trafic for some time ^_^ [21:19] beuno: ok [21:19] dobey: hum, I have no DbusTestCase in contrib... any idea where that comes from on u1-devtools? [21:21] nessita: I think it's DBusTwistedTestCase or something like that. it's in contrib/testing/testcase.py [21:21] dobey: right now is on tests/platform/linux/test_dbus.py [21:21] which sucks :-/ [21:22] can't have the credentials module isolated [21:22] I think they move it to be able to do the multiplatform thing [21:23] nessita: it should be in contrib, unless it was moved in the "big move for win32 support" [21:23] verterok: it was, apparently [21:23] nessita: then I agree...it sucks :( [21:23] i think it should still be in contrib, but perhaps it should have been more isolated or something [21:24] dobey, verterok: I think in this case I should use the one from devtools, since mandel1is already making that framework multiplatform [21:26] dobey: since we're moving to devtools anyways, I will not put that much energy on this [21:27] ok [21:27] dobey: let's add the new dep to u1client's dep and I leave the devtools import as is [21:29] when your branch lands, i'll add the build-dep [21:29] for the moment though; dbus hates me [21:29] but i think i know what's wrong [21:30] ok [21:30] I'll eod now, need to run some errands [21:31] see ya tomorrow crowd [21:36] muffinresearch: did you log the bug earlier on ? [22:10] Have a good evening everyone, EOD for me. === JanC_ is now known as JanC