[11:01] <karni> Good morning team!
[11:17] <gatox> good morning!
[11:21] <karni> Morning gatox
[11:22] <gatox> karni, o/
[11:24] <gatox> karni, if you have a moment please :D https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/share-indicator/+merge/131973
[11:24] <gatox> karni, also, if you have any branches pending for reviews..... feel free to through it to me
[11:27] <karni> gatox: I have one coming very soon, not just yet. But thanks :) Yeah, I'll have a look.
[11:28] <karni> gatox: Does that mean we're moving out from the notifications tray menu?
[11:32] <gatox> karni, nono..... this is a functionality that is intended for windows and mac..... but if you want to test it, you can use it on linux too with the --with-icon arg
[11:32] <karni> I see.
[11:33] <karni> gatox: +1'ed
[11:33] <gatox> karni, the notification tray menu is from u1 client..... and this is integrated with control panel..... which is always running on windows and mca..... not the case for linux
[11:33] <karni> gatox: I trust the tests pass on your machine, I still haven't gone through the developer set up for the desktop.
[11:33] <gatox> karni, thx
[11:33] <karni> gatox: I see. np!
[11:34] <gatox> karni, great
[13:55] <dobey> whee conf wifi
[13:56] <aquarius> :)
[13:58] <gatox> karni, another review (really small one) for when you have a moment (no rush): https://code.launchpad.net/~diegosarmentero/ubuntuone-client/menu-new-options/+merge/132106
[13:58] <aquarius> actually trying a different irc client
[14:01] <karni> gatox: sure
[14:02] <gatox> karni, thx!
[14:02] <karni> aquarius: Be right there with you.
[14:04] <karni> gatox: /s/Ubuntu One Website/Ubuntu One website or even 'web site' ? You'd have to ask a native speaker.
[14:05] <gatox> karni, that is in that way in the design document
[14:05] <karni> gatox: Looks like a capitalization convention you're using there consistently. "Open the Ubuntu One Folder" (Folder), "Share a File" (File).
[14:05] <karni> gatox: roger that!
[14:05] <gatox> karni, i just copy the string from design
[14:05] <karni> gatox: k :)
[14:06] <karni> gatox: "webbrowser.open(self._syncdaemon_service.get_rootdir())" -- this contains full "rootdir" url?
[14:06] <karni> gatox: such as http://one.ubuntu.com/files/~/Ubuntu One ? (or similar)
[14:07] <gatox> karni, that contains the absolute path to the user ubuntu one folder in the disk
[14:07] <karni> aha
[14:07] <gatox> as: /home/gatox/Ubuntu One
[14:07] <karni> gatox: So it'd spawn the default file browser, although it says "webbrowser" ?
[14:07] <gatox> karni, yap
[14:07] <karni> neat
[14:08] <gatox> karni, that implementation actually works on windows and linux....... for mac you need to add file:// at the start.... but THIS sync menu is only for linux.... so....
[14:09] <karni> gatox: This one's for linux, yep?
[14:09] <karni> gatox: +1'ed
[14:09] <dobey> ugh
[14:09] <gatox> karni, awesome, thx
[14:10] <karni> np
[14:10] <gatox> brb..... need to buy some vegetables in the fair downstairs
[14:11] <gatox> s/in/at
[14:21] <gatox> back
[14:42] <mmcc> hi folks
[14:43] <gatox> mmcc, hi
[14:45] <gatox> mmcc, when you have a time, can you re-review this one? https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/share-indicator/+merge/131973
[14:45] <mmcc> gatox, yep, it's on my list.
[14:45] <gatox> mmcc, great, thanks!
[14:51] <karni> hi mmcc
[14:52] <mmcc> hi, karni
[15:00] <karni> me
[15:00] <gatox> me
[15:00] <briancurtin> me
[15:01] <rockstar> me
[15:01] <karni> dobey: me?
[15:01] <karni> mmcc: ↑
[15:02]  * karni is not sure who's around and who's not, sorry! :)
[15:02] <karni> DONE: shuffle, repeat TODO: pause when headset unplugged, pause/resume on incoming call NEXT: gatox
[15:02] <gatox> DONE:
[15:02] <gatox> Couple of branches to improve the Control Panel Share tab. Add Share tab and Ubuntu One folder access from the U1 sync menu indicator and the windows/mac implementation, both in control panel and ubuntu one client.
[15:02] <gatox> TODO:
[15:02] <gatox> Keep fixing control panel bugs, and figure it out how to fix the failing test on nightlies.
[15:02] <gatox> BLOCKED:
[15:02] <gatox> No
[15:02] <gatox> briancurtin, go
[15:03] <briancurtin> DONE: reviews for share branch, ubuntuone-client test debugging since 110 tests are failing
[15:03] <briancurtin> TODO: fix ubuntuone-client tests so i can continue with the release and get an installer to QA
[15:03] <briancurtin> NEXT: rockstar
[15:03] <rockstar> DONE: convert hand crafted view into xib
[15:03] <rockstar> TODO: continue work on the v2 api
[15:03] <rockstar> BLOCKED: No
[15:03] <rockstar> NEXT: ?
[15:03] <karni> UDS has kicked off, dunno who might be there as well.
[15:04] <gatox> briancurtin, when you have a moment.... this branch has been updated with the latests changes: https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/new-share-design/+merge/131397
[15:04] <gatox> eom?
[15:05] <karni> eom
[15:05] <briancurtin> gatox: yeah i saw that update yesterday. i ran the tests and everything is fine, but im having trouble running on windows (unrelated to your branch - SD doesnt want to run)
[15:05] <dobey> oh snap
[15:05] <briancurtin> gatox: once i get ubuntuone-client tests cleaned up and running, i'll try it again. working on it now
[15:06] <gatox> briancurtin, thx
[15:06] <mmcc> whoops. completely forgot standup, was making coffee…
[15:06] <dobey> DONE: script to tweak series/branch statuses
[15:06] <dobey> TODO: create next stable branches
[15:06] <dobey> BLCK: None.
[15:07] <mmcc> DONE: tests/polish on mac menu — TODO: continue, reviews — BLOCK: no
[15:08] <karni> np, thanks guys
[15:26]  * gatox lunch
[15:33] <dobey> lunch time, bbiab
[15:53] <mmcc> briancurtin - are you still having hangs because syncdaemon won't load? I'm testing gatox' branch on windows now and syncdaemon is complaining that one of the tritcask files is owned by another process (even though all the other u1 processes are dead) - is that familiar?
[15:54] <briancurtin> mmcc: i'm not seeing that particular issue right now, but i think this whole thing is related. i was seeing that for a while, then i was seeing a KeyError inside tritcask during local rescan, and now nothing obvious shows up in the logs but SD starts disconnected when i start U1CP
[15:55] <briancurtin> also u1client tests still fail big time on windows, so im trying to figure that out first
[15:55] <mmcc> it starts disconnected when you start CP with credentials or without credentials?
[15:55] <briancurtin> with
[15:55] <mmcc> ok, so probably not related to my cloud-folders fix
[15:55] <mmcc> oh crap, that still needs to be merged
[15:56] <briancurtin> mmcc: we never got back to fixing that up. it just seemed to hang forever for me IIRC
[15:56] <mmcc> right… ok I'll re-review that one myself now. assuming I can get things to start again :\
[15:58] <mmcc> ok, trying again with nuking all u1 processes got me a working CP
[15:59] <mmcc> (sans my cloud-folders fix)
[16:00] <diogobaeder> Guys, I found a problem. None of the songs from this album are loading for preview: http://goo.gl/0qytW
[16:01] <mmcc> diogobaeder that does sound like a problem (I see it too) - did you mean to send to u1-internal?
[16:02] <diogobaeder> mmcc, I was not sure what would be the best channel for it, so I sent it here. Will send further problems to #u1-internal next time :-)
[16:03] <mmcc> diogobaeder - this channel's pretty quiet wrt server-side stuff, even though I think most people are in both
[16:04] <diogobaeder> mmcc, ack
[16:04] <diogobaeder> I'll resend the message there
[16:08] <mmcc> briancurtin, gatox_lunch - on windows with current u1-client trunk and https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/share-indicator/+merge/131973 , the systray menu 'share a file' menu item will open the shares tab, but the shares tab doesn't load… hung on the overlay, no useful log messages or tracebacks in either CP or SD
[16:08] <mmcc> gatox - share-indicator should have the code to use the new search_files stuff in u1-client, right
[16:08] <mmcc> ?
[16:09] <gatox_lunch> mmcc, no, this part doesn't depend on that
[16:12] <mmcc> gatox_lunch - ok, so the share-indicator branch doesn't contain the changes from search-shared-files, I see…
[16:12] <gatox_lunch> mmcc, aaja
[16:14] <mmcc> which might be why the shares tab doesn't load when I'm testing share-indicator… although I expect it to still work, since it was doing file scanning itself before you switched to the new api…
[16:14] <mmcc> anyway, I'll look at it. sorry to bug you during lunch
[16:17] <gatox_lunch> mmcc, no problem..... let me know if you find something
[16:21] <mmcc> oh this is frustrating. last time I quit U1CP on windows, it left the python icon in the taskbar. when I started it up again, it didn't put a new one in, and the old one doesn't work. and using the systray menu to bring up the CP window worked once but now it doesn't work
[16:21] <briancurtin> windows sucks
[16:21] <karni> +1
[16:21] <karni> :P heheh
[16:22] <mmcc> oh it was minimized or whatever? there's some real mismatch between how windows works and how I expect it to work
[16:22] <mmcc> I'm sure some of this is just me not knowing what it's supposed to be doing
[16:22] <karni> mmcc: I appreciate you guys do this (so that we don't have to do it ;P)
[16:22] <briancurtin> ah, windows is bad about knowing when to clear out taskbar icons of closed/crashed/etc applications
[16:22] <briancurtin> i'll look in there and it'll have 10 U1 icons
[16:25] <mmcc> karni - "you guys". the irony here is that I'm the Mac guy. :}
[16:27] <mmcc> I was about to say "it even says that in my job title", but it actually says "Mac/Win", so there you are
[16:28] <karni> mmcc: ha ;D (I am aware youre the 'Mac guy', still, you're helping out with Windows. Or it looks like so ;) )
[16:29] <mmcc> briancurtin - I'm seeing the thing where it's disconnected when it starts up. and the thing where the account info doesn't load, and when you minimize it to the taskbar and then re-show it, it tries to connect (?!)
[16:30] <mmcc> lots of fun on windows today
[16:31] <briancurtin> yeah the whole thing is just a ton of fun
[16:31] <mmcc> happy fun windows (do not taunt, etc. etc.)
[16:31] <briancurtin> ahhah
[16:35] <briancurtin> i guess this is all failing due to IPC
[16:37] <mmcc> kind of looks like syncdaemon is not getting the credentials found signal
[16:37] <mmcc> IPC, sigh
[16:42] <mmcc> no, I guess it got the creds ,since it pushes the USER_CONNECT event… but SD hangs at state='with user not network'
[16:47] <mmcc> looks like two different parts of the code are concurrently trying to popen ubuntu-sso-login
[16:51] <mmcc> and the last call to interaction_interfaces's connect() appears to have autoconnecting=False, even though my settings show that it should be true… (unless I'm misunderstanding that argument's meaning
[16:51] <mmcc> )
[16:51] <mmcc> brb, have to watch the kid for a minute
[17:22] <mmcc> briancurtin, so when I start syncdaemon first in a separate console, it doesn't complain, and CP seems to be respecting the auto-connect setting
[17:23] <briancurtin> hm, let me try that
[17:23] <mmcc>  - but it only gets to 'file sync starting' - it doesn't make it to idle
[17:26] <mmcc> yeah, controlpanel.backend.connect_files doesn't appear to be returning
[17:29] <briancurtin> nice
[17:29] <briancurtin> yeah i see the same thing here, file sync starting
[17:30] <karni> I'll run an errand, afk maybe 30"
[17:33] <mmcc> same issue, calling credentials.CredentialsManagementTool().login() from request_token is just returning a deferred that I guess is never firing
[17:35] <mmcc> although if you put a print after the line in interaction_interfaces in connect(), where it says token=yield self._request_token, it will get called once when you start syncdaemon, but *not* when you then later start CP
[17:37] <mmcc> no, now I'm not sure that's true
[17:38] <mmcc> yeah, that wasn't it
[17:38] <mmcc> it does get the right token from _request_token both times
[17:38] <briancurtin> i'm looking back through revisions to see when this stuff might've changed. this was working early last week
[17:41] <dobey> hrmm
[17:48] <mmcc> gotta love how all the comments in ipc_service.py reference dbus
[18:04] <mmcc> I'm getting a hunch that we are registering for signals incorrectly in ipc_service.py (in ubuntuone/platform/credentials) but I don't really understand what's going on in RemovableSignal, which we're calling with the signal name 'on_credentials_found_cb', and grepping for that in all the projects doesn't give any useful results
[18:11] <mmcc> trying a little instrumentation to see what's going on there
[18:13] <mmcc> it looks like we're registering for 'on_credentials_found_cb', and sso is sending 'CredentialsFound'
[18:13] <mmcc> there's a mapping between the two in ipc_service.py, but it's not clear why
[18:13] <briancurtin> is that affecting mac as well, or is this only on windows?
[18:13] <mmcc> that is, a mapping from CredentialsFound (and similar) to 'on_credentials_found_cb' and similar
[18:14] <mmcc> briancurtin only windows, afaict. that's a big WTF, for sure.
[18:15] <mmcc> A difference is that we're using unix sockets on mac, and tcp on windows, so maybe there's something different about how it handles multiple connections that makes it work on macs despite this
[18:15] <gatox> dobey, i just answered to your need info
[18:17] <dobey> ok
[18:19] <mmcc> whoops, have to go for lunch now.
[18:29]  * briancurtin back in a bit
[18:33] <dobey> have to run for a few, brb
[18:39] <karni> Was slighly longer. Will make it up now.
[19:17] <dobey> hey karni
[19:17] <karni> hi dobey
[19:29] <gatox> back..... little internet issue
[19:38] <dobey> gatox: and we are not showing the newly accepted shares in there at all are we?
[19:39] <gatox> dobey, no..... i suggested that..... but it was decided that doing that seems to be adding a feature just to fill the space.... and it's true
[19:39] <gatox> so it was decided to not do tha
[19:39] <gatox> that
[19:40] <dobey> well that's good at least; having that in 2 places would be weird
[19:41] <dobey> meh; between uds and holidays, finding the right people to talk to about some things isn't exactly easy :-/
[19:42] <gatox> dobey, if you have any question about this..... ask me.... maybe i can help you.... my last branches in particular are based in a lot of stuff we have been discusing lately
[19:44] <dobey> i have some questions about launchpad/bzr; was writing an awesome script to make my life easier, but i guess i'll have to make stable branches the hard way for now :-/
[19:45] <gatox> ahh
[19:46] <dobey> i guess i still have some questions about your branch (and i guess, the indicator menu in general)
[19:47] <gatox> dobey, shoot
[19:48] <dobey> gatox: what happens if a user clicks on that "share a file" entry, and control panel is already running (and is behind a bunch of firefox windows or something)?
[19:51] <gatox> dobey, in the u1-client or the control panel implementation?
[19:51] <gatox> dobey, basically..... is going to do the same as it does when you choose the "open ubuntu one" optoin...... but moving the current tab to the "share" tab
[19:52] <dobey> well i imagine it will be less problematic in the control panel implementation.
[19:52] <dobey> gatox: yes, but my question is about when cp is already open
[19:53] <dobey> gatox: in which case it seems the desired behavior is broken
[19:53] <gatox> dobey, on linux at least..... it just shows the icon in the launcher jumping to notify that the application is trying to request the focus
[19:54] <dobey> gatox: not here it doesn't (quantal)
[19:54] <gatox> dobey, mmm.... i'm in quantal too
[19:55] <dobey> i get no shaking icon
[19:56] <gatox> ahhhh right..... for the open ubuntu one.... i get the icon..... for share a file the application just popup to the front
[19:56] <dobey> see bug #1063927
[19:57] <dobey> gatox: i don't get the icon doing anything special, for either case
[19:57] <dobey> the window does get raised (if it's not minimized) though; however, it doesn't have focus
[19:58] <dobey> not sure how/why that is, but we should fix it
[19:58] <gatox> dobey, either way..... i think that is not related to this branch....... we should fix that in the part from control panel or whatever where the process is spawned
[19:58] <mmcc> briancurtin, any new news on that windows IPC issue?
[19:59] <gatox> dobey, i mean..... the one which receive the call
[19:59] <briancurtin> mmcc: none from my end, im working on getting SD tests fixed up at the moment
[19:59] <mmcc> briancurtin ok, I'll keep poking then. will let you know if I learn anything useful
[19:59] <dobey> gatox: even so, there is no way to fix the --switch-to case, without making the control panel a service
[20:00] <gatox> dobey, mmh
[20:00] <briancurtin> dumb bzr question: how can i tell what revision i'm currently at? all i can tell right now is that im not at the tip
[20:00] <briancurtin> oh there we go, bzr revno
[20:00] <gatox> dobey, yes....
[20:00] <dobey> gatox: this will also be an issue on mac, where the menu will be a separate process, and not part of cp
[20:01] <mmcc> dobey - right, but we're going to have to use a separate mechanism to bring cp to the front, and I'm not sure yet how to handle showing the shares tab
[20:01] <gatox> dobey, wait..... which branch are you reviewing..... the one from u1-client is only valid for linux
[20:02] <mmcc> er, that is - dobey is right, *and* it's my problem, not gatox's :)
[20:02] <dobey> mmcc: the mechanism isn't just "run the control panel" like it is in this code?
[20:02] <briancurtin> ha, well either bzr revno is wrong or my tree is wrong...so rm -Rf i guess
[20:03] <gatox> dobey, for this branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/menu-new-options/+merge/132106 the behavior is only valid for linux..... because that is to interact with the sync menu indicator..... which we only have on Ubuntu
[20:03] <dobey> gatox: no, the behavior is supposed to be the same across all the menu implementations (which we now have to have 3 of, which also sucks)
[20:04] <mmcc> briancurtin - comparing bzr revno to the trunk revno is problematic… can be higher than trunk revno if you have lots of local commits
[20:04] <gatox> dobey, no, because the other menu implementations are in the control panel, the one in u1-client is just for ubuntu
[20:04] <briancurtin> mmcc: i'm just going backward from trunk with no changes at all. i thought i was on 1350, it claimed 1355, now i have no idea. whatever, bzr is awful.
[20:04] <gatox> dobey, because the other one is using qt..... which we don't have in u1-client
[20:04] <mmcc> briancurtin bzr qlog  is probably your friend here. revnos are not usually helpful IMO
[20:06] <dobey> gatox: no, the mac menu won't be the cp --with-icon thing. mmcc has written a proper menu with pyobjc for mac.
[20:06] <gatox> dobey, that is because on linux we have u1-client running all the time...... and on windows/mac we have also the u1-cp running all the time..... and in those platforms, cp is the one who adds the syncmenu
[20:06] <mmcc> dobey: causing CP to run / come to the front is basically "run CP", but making it do that, then open the share tab, even if it's already running, is not easy. I can't do what he's doing in open_share_tab, because I'm in a different process…
[20:06] <gatox> dobey, ah yes....... but i don't see how is that related to this
[20:07] <mmcc> gatox  -well, for one, we won't have CP running all the time any more
[20:07] <mmcc> on osx
[20:07] <dobey> mmcc: what he's doing in open_share_tab is running "ubuntuone-control-panel-qt --switch-to share_links"
[20:07] <gatox> mmcc, you can do what i do on u1-client..... spawn the u1-cp process with: --switch-to tab-share
[20:07] <dobey> mmcc: the problem is that argument is ignored in cp, if it's already running
[20:07] <gatox> mmcc, sorry..... share_links
[20:07] <mmcc> I must be looking at a different branch…
[20:07] <briancurtin> ooh, qlog is nice
[20:08]  * dobey does wish this one branch was 2 branches instead
[20:08] <mmcc> I was looking at this: ~diegosarmentero/ubuntuone-control-panel/share-indicator
[20:08] <gatox> mmcc, dobey is talking about this one: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/menu-new-options/+merge/132106
[20:08] <mmcc> oh
[20:10] <mmcc> so yeah, what happens when you do that and cp is already running? …
[20:11] <mmcc> also, not new in that branch, but open_go_to_web has "Ubuntu" misspelled in its docstring :)
[20:11] <dobey> basically, not the desired thing :)
[20:11] <mmcc> right…
[20:11] <dobey> so it does
[20:12] <gatox> dobey, ahhhhh now i see what you mean... the problem wasn't if it opens or don't u1..... the problem was that it doesn't change to the propre tab
[20:12] <gatox> dobey, i wasn't understand what you mean
[20:12] <dobey> gatox: it doesn't change to the tab; nor does the window grab focus
[20:13] <gatox> dobey, well..... in my case doesn't grab focus..... but the icon in the launcher get highlighted, and in awn start jumping
[20:13] <dobey> gatox: well, it's supposed to grab focus. if it's not already running, then the right thing happens, and the window has focus
[20:14] <karni> rockstar: Do you also have shuffle play in iOS music app? I just realized the "prev" button in case of shuffle enabled doesn't do the job well.
[20:14] <karni> rockstar: If you do, how do you approach it? Randomizing a list of songs? I try to keep as little in memory as possible. Even my play queue is persisted and queried (SQLite) only on demand.
[20:14] <gatox> dobey, ok..... but the place to fix that wouldn't be u1-cp?
[20:14] <rockstar> karni, trunk might, but this 2.0 branch doesn't currently.
[20:15] <karni> rockstar: trunk - whatever is on App Store?
[20:15] <karni> urbanape: Hi hi!
[20:15] <karni> urbanape: Do you also have shuffle play in iOS music app? I just realized the "prev" button in case of shuffle enabled doesn't do the job well.
[20:15] <dobey> gatox: well i don't think the tab not switching is fixable in just u1-cp at least
[20:15] <karni> urbanape: If you do, how do you approach it? Randomizing a list of songs? I try to keep as little in memory as possible. Even my play queue is persisted and queried (SQLite) only on demand.
[20:16] <dobey> gatox: admittedly i don't know that code all that well either
[20:16] <rockstar> karni, yes. I'd check, but i have my current branch on my phone, so it would require deleting that first.
[20:17] <dobey> gatox: also, i'm not really sure we should have these things in the menus
[20:17] <dobey> gatox: also also, one can't share a file, anyway. one can only share folders :)
[20:17] <karni> rockstar: Whenever/if you have time. I'm actually more interested how that's solved than if it is in the app at all (if it is) :) If you know what I mean.
[20:17] <rockstar> karni, as I recall from the web player, a good shuffle is relatively non-trivial.
[20:17] <karni> rockstar: I'll try to grap urbanape.
[20:17] <dobey> so there is some language issues i think we need to take up with design
[20:17] <karni> rockstar: Indeed.
[20:18] <urbanape> karni: I'm heading out just now, but I'll be back later tonight
[20:18] <karni> urbanape: Ah cool! no problem :)
[20:18] <karni> Talk to you then.
[20:18] <rockstar> Basically, you'd have to keep track of the songs that you've played, in the order, and when it chooses another random song, it needs to ensure that the song it picked hasn't played until they've all played.
[20:18] <gatox> dobey, well....... share a file..... technically will be "publish a file"..... but....... for that...... and include or not include that in the menu we should talk with someone from the design team....... did you take a look at the document i mentioned? where it specifies what the menu should have and do?
[20:19] <dobey> i did
[20:19] <dobey> but it doesn't mean the design itself doesn't need a bit more review :)
[20:20] <gatox> dobey, yes....... i understand...... so....... are you suggesting to freeze this until we have roberto or alecuu here to discuss it further?
[20:20] <dobey> also that doc doesn't really say anything about *why* the various things need to be there
[20:20] <dobey> just that they should be there
[20:21] <gatox> dobey, all the documents i got from design are like that :P
[20:21] <dobey> :(
[20:22] <gatox> dobey, ok.......  do you want to freeze this to discuss it with the rest of the team later and take a well informed decision?
[20:22] <mmcc> hrm. "later" is kind of far off, you're basically saying "London"…
[20:22] <dobey> well there are things that need fixing otherwise too, i think
[20:23] <mmcc> I'm going to have to make some choices for the mac menu sooner than that
[20:23] <gatox> yes..... later is really far away
[20:23] <dobey> mmcc: well, there's tomorrow; i don't think all of design are on vacation, or in .dk
[20:24] <gatox> dobey, yes...... but at least lisette that was working on this is not in our team anymore
[20:24] <mmcc> dobey, that's true. as for 'open control panel and switch to the share files tab', I am happy tabling that until London, since it'll need some CP changes I think
[20:25] <gatox> dobey, ok....... we can try to start pinging people tomorrow..... let me know when you are here..... i'll be earlier
[20:25] <gatox> 'll be here earlier
[20:25] <dobey> i'll probably just e-mail
[20:25] <gatox> dobey, can you cc me?
[20:25] <dobey> yes
[20:26] <gatox> dobey, thx
[20:26] <mmcc> me too, please…
[20:26] <gatox> dobey, i don't have any problem if it is decided that this shouldn't be included..... but..... i did it because it was in the document, and was dicussed a couple of time that it should be done after the freeze.... that's why
[20:26] <mmcc> gatox, what else is on your plate now? Are you familiar with the Windows IPC stuff? can you see how things are working for you in windows land - see if you get similar problems with trunk to what brian and I have been seeing?
[20:28] <gatox> mmcc, nothing major here....... so i can take a look at that....... probably tomorrow because i already pass my eod here....... it would be very helpful if you or briancurtin can sent me an short email telling me which are the problems..... i'm familiar with the ipc part so i can try to figure it out if i know what i should be looking for
[20:29] <mmcc> gatox, ok. here's the email: "running control panel from source hangs on the loading overlay at various places. most recently in the shares tab, for me."
[20:29] <mmcc> I will send more details when I'm done poking at it today
[20:30] <gatox> mmcc, awesome, thx!! i'll start looking at that first time tomorrow.... if you find any other thing please send me an email
[20:30] <gatox> ok..... eod here! i'll specting some emails tomorrow from you :D
[20:30] <gatox> byeeeeeee
[20:30] <mmcc> ok, thanks gatox :) bye
[20:41] <mmcc> I wonder if this is related to the problem mandel had to hack around on osx with the 'credentials found' signal not being sent when you first log in -- there are apparently a few different ways that we're returning results in IPC calls - sometimes via signals, sometimes via return values…
[20:42] <briancurtin> am i missing something or does this change look to be fairly harmless? http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntuone-client/trunk/revision/1351 - this is where the tests picked up 110 failures
[20:44] <karni> rockstar: Right. I'll tell you what my problem is. First of all, I don't want to do the shuffle in memory. So I flag the songs which have been already played, and pick the next random one from the rest.
[20:44] <karni> rockstar: As you can see, it's easy to pick the next random one, but I'd have to track the previous one as well, so that you can hit "prev".
[20:45] <karni> rockstar: SO, I can use this column that I have, shuffle_order, which contains random ints, and sorting by it would give me the order of the queue.
[20:45] <rockstar> What if they hit "prev" twice, or three times?
[20:45] <karni> rockstar: Even worse. No way. But listen up.
[20:46] <mmcc> briancurtin, yeah, that seems very harmless… hm
[20:46] <karni> rockstar: The problem/disadventage is that if the user hits "shuffle" I should update that column for all tracks on that list, which would be equivalent of doing something like Collections.shuffle(playQueue) on the playQueue (in memory).
[20:46] <rockstar> Yeah, I think that's the solution you want.
[20:47] <karni> rockstar: So, if I want prev/next, I think it's either in memory solution (easier, but more memory), or update that shuffle_order column (slighly more complex, but memory efficient).
[20:47] <mmcc> briancurtin - what are the failures that's causing?
[20:47] <karni> rockstar: Yeah.. k, I wanted a sanity check. Thanks.
[20:48] <briancurtin> mmcc: they're all in test.platform.test_tools. no real indication of whats happening except that the defer never fires, or something..."still running at 5.0 secs"
[20:48] <mmcc> briancurtin oh yeah, I saw those too
[20:48] <mmcc> so if you revert that change, everything's OK? whoa
[20:49] <briancurtin> mmcc: yep, going back to the previous revision, everything's cool
[20:51] <mmcc> briancurtin weird.
[20:54] <mmcc> kind of does seem like the network manager should be mocked out for tests though. It's spawning threads for each of these tests, and doesn't need to
[20:55] <mmcc> not sure why that's changed though
[20:55] <mmcc> I mean, why that rev would change things
[20:56] <mmcc> briancurtin - fwiw, the corresponding changes in sso are here: https://code.launchpad.net/~dobey/ubuntu-sso-client/lint-roller/+merge/131016
[20:57] <briancurtin> yeah i saw those, and made sure i have an up-to-date sso to match
[20:59] <mmcc> well, one thing is that networkstate/windows.py has its own internal definitions of ONLINE,OFFLINE,UNKNOWN still -- despite those being defined in networkstates now
[20:59] <mmcc> (that's my fault !)
[21:00] <mmcc> so if the tests are comparing the return values of networkmanager anywhere to networkstates.ONLINE, for instance, it won't work as expected
[21:00] <briancurtin> good catch. hmm.
[21:02] <mmcc> oh, not really totally my fault! I didn't create networkstates.py… I had plenty of missed chances to notice though :(
[21:02] <dobey> hmm
[21:03] <mmcc> ok, I'm going back to looking at IPC. as usual, asking the question "how does this ever work?"
[21:18] <briancurtin> i pushed that little networkstates branch at https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/networkstates-update/+merge/132206 if anyone can do a super-small review
[21:19] <briancurtin> i need to head out a bit early and pick up my girlfriend. i'll either come back later on or more likely start work earlier tomorrow
[22:00] <mmcc> good news - that windows networkstates branch also fixed the ipc errors I was seeing. I guess the problem must have been that it thought there was no network, so it decided it couldn't connect.
[22:00] <karni> mmcc: \o/
[22:00] <mmcc> we don't handle telling the user about that error condition very well (ie, not at all)
[22:06] <dobey> briancurtin: lint issue in that branch :)
[22:06] <briancurtin> dobey: just pushed a fix
[22:06] <dobey> great
[22:29] <dobey> alright, need to get away from here for a bit. still need to send that e-mail though, so will be back later
[22:29] <dobey> later
[22:29] <briancurtin> see ya
[22:32] <mmcc> ok, I have to go carve a pumpkin before it gets dark. I'll be back later too…
[23:04]  * karni triages an SQL update issue
[23:21] <karni> Something's up with resolving playlist content, but I've temporarily worked it around using full (verbose) song URIs, which works.
[23:21] <karni> Premature optimization is the source of all evil, isn't it.
[23:24]  * karni EODs, night all!