[00:26] <chaselivingston> ping mmcc: around?
[00:55] <mmcc> chaselivingston: EOD a little while ago, just checked in and saw your ping. I'll be back a bit tonight, maybe email me? or I'll just catch you tomorrow
[00:56] <chaselivingston> mmcc: cool, i'll email you
=== tamotsukun is now known as ramon
=== mmcc_ is now known as mmcc
=== Chipaca` is now known as Chipaca
[09:59] <JamesTait> Happy Friday, folks! :-D
[10:25] <ramon> Happy Friday ;)
[10:26] <czajkowski> mandel`: mandel two of you!
[10:29] <mandel> czajkowski, yes, bip server issues :)
=== Guest71222 is now known as adorilson
[12:01] <alecu> mandel: hello!
[12:07] <gatox> good morniing
[12:31] <ralsina> good morning!
[12:32] <gatox> ralsina, hi
[13:03] <gatox> people, ralsina is having issues with his internet connection, and he can't leave his home at this moment.. just to let you know
[13:17] <gatox> the sync menu is looking gooood! :D
=== m_conley_away is now known as m_conley
[13:35] <mandel> alecu, hello!
[13:35] <mandel> alecu, sorry I was out for lunch
[13:37] <alecu> hola mandel!
[13:37] <alecu> mandel: any news about the demo?
[13:38] <alecu> mandel: did you got my latest branch?
[13:39] <mandel> alecu, yes, I saw your latests branch, so far I have no idea about the demo, we should ash ralsina
[13:44] <dobey> gatox: so what does it look like?
[13:45] <gatox> dobey, like it works :P
[13:45] <gatox> (the new one)
[13:46] <alecu> mandel: have you pushed the unity branch that fixes the state machine?
[13:47] <dobey> gatox: i was hoping for pictures :)
[13:47] <gatox> dobey, let me add the progress bar and i'll have pretty pictures :P
[13:48] <mandel> alecu, not yet, let me finish a fix I have and send it to you
[13:59] <dobey> gatox: how about a reivew of https://code.launchpad.net/~dobey/ubuntuone-client-data/move-icons/+merge/124425 ? :)
[13:59] <gatox> on it
[14:04] <mandel> alecu, pushed to lp:~mandel/avani/u1-payment-preview there are BIG problems with the new layout of things and the new fonts, fixing them at the moment
[14:04] <gatox> dobey, +1
[14:04] <mandel> alecu, you will see that the transition to the new preview is broken, so we have to add extra code to DashView.cpp to know how to do it correctly
[14:06] <alecu> mandel: what about the password entry box?
[14:06] <alecu> mandel: does it still show the password contents???
[14:07] <mandel> alecu, yes, major bug there.. even doing the following: lp:~alecu/avani/unity-lens-music.ubuntuone-purchase does not work
[14:07] <mandel> alecu, I need to debug nux :(
[14:07] <mandel> alecu, I don't think it really supports it, but is easy to implement
[14:09] <alecu> mandel: right
[14:09] <alecu> mandel: the thing is that to complete a purchase right now you need to use your real password, so make sure to change it after your demo :-)
[14:09] <mandel> alecu, sorry, pasted the wrong code, I meant doing: password_entry_->text_entry()->SetPasswordMode(true)
[14:10] <alecu> mandel: ah, "even doing"... the above.
[14:16] <mandel> alecu, yes, and we just found the error, a bug in nux
[14:17] <mandel> alecu, doing a workaround right now
[14:23] <mandel> alecu, if you pull you will have a copy with no public password
[14:23] <alecu> mandel: great, thanks!
[14:24] <mandel> alecu, I'll keep improving the ui to make the layout work.. we lack a grid layout or a form layout in nux so getting the things right with HLayout and VLayout is a hard thing to do :-/
[14:39] <gatox> alecu, ping
[14:39] <alecu> gatox: pong
[14:39] <gatox> alecu, pm
[14:45]  * dobey wonders when ralsina will have internet again
[14:53] <gatox> dobey, do you need me to send him a sms?
[14:53] <dobey> no
[14:53] <dobey> and i can do that if i need to anyway :)
[14:54] <dobey> irc-sms gateway == win
[14:56] <mmcc_> good morning everyone
[14:56] <gatox> mmcc_, hi
=== mmcc_ is now known as mmcc
[14:57] <mmcc> hi gatox!
[14:58] <ralsina> dobey: here I am!
[14:58] <facundobatista> does anybody know where the "temp-directory" is set for the tests when I run "make test" in the client?
[14:59] <ralsina> I am changing ISPs on monday, so hopefully this will not happen again
[14:59] <ralsina> yeah right
[14:59] <dobey> heh
[15:00] <thisfred> me
[15:00] <gatox> me
[15:00] <dobey> facundobatista: it's set to the default trial directory; do you need to change it for a test run?
[15:00] <dobey> me
[15:01] <facundobatista> dobey, where the default trial directory is set? my problem is that it's being set to an unicode path and this break an assertion in code
[15:01] <mandel> facundobatista, isn't it in devtools?
[15:01] <dobey> in trial
[15:01] <facundobatista> mandel, can not find where
[15:01] <mandel> facundobatista, sets the env variable, right dobey?
[15:01] <mmcc> me
[15:01] <alecu> oh, me
[15:01] <dobey> it's os.path.join(os.getcwd(), '_trial_temp') by default
[15:02] <gatox> thisfred, go 12:02
[15:02] <facundobatista> dobey, where's that? if I run that in a clean python env, I get bytes path
[15:02] <thisfred> DONE: being sickish TODO: finish private u1db server API | make playlists work with that API BLOCKED: no NEXT: gatox
[15:02] <gatox> DONE:
[15:02] <gatox> Moving forward with the SyncMenu implementation in u1-client. Trying to get the timer to work to update the progress of the files being uploading.
[15:02] <gatox> TODO:
[15:02] <gatox> Finish with the sync menu implementation.
[15:02] <gatox> BLOCKED:
[15:02] <gatox> No
[15:02] <gatox> dobey, go
[15:02] <dobey> facundobatista:         trial_temp_dir = os.environ.get('TRIAL_TEMP_DIR', os.getcwd())
[15:02] <dobey> DONE: team call, stopped u1client-data FFe/UIFe, researched webapps packaging, bug #1050923
[15:02] <ubot5> Launchpad bug 1050923 in ubuntuone-client-data stable-4-0 "Icons not getting installed" [Undecided,Triaged] https://launchpad.net/bugs/1050923
[15:02] <dobey> TODO: reviews, new icons?, webapp package
[15:03] <dobey> BLCK: None.
[15:03] <dobey> mmcc: go
[15:03] <mmcc> DONE: helped lisette with daemon issues, root daemon version checking / upgrading
[15:03] <mmcc> TODO: daemon event processing bug with move
[15:03] <mmcc> BLOK: daemon upgrade runs into apparent bug in apple code or docs,
[15:03] <mmcc> NEXT: alecu
[15:03] <alecu> DONE: lens now contacts all required webservices and completes a purchase
[15:03] <alecu> TODO: handle lots of errors and corner cases, lots more tests
[15:03] <alecu> BLOCKED: no!
[15:03] <alecu> NEXT: ralsina, when he gets back
[15:03] <dobey> facundobatista: that bit is in dev-tools
[15:03] <alecu> oh, he's here!
[15:04] <facundobatista> dobey, ok, that I get bytes
[15:04] <facundobatista> dobey, but next line does:
[15:04] <facundobatista>         homedir = os.path.join(trial_temp_dir, options['temp-directory'])
[15:04] <facundobatista> and that options['temp-directory'] is unicode for me
[15:04] <facundobatista> and I can not find where options['temp-directory'] is set
[15:04] <dobey> facundobatista: it is set in trial itself
[15:05] <dobey> facundobatista: but in dev-tools we use unicode_literals, and maybe something is causing that to be unicode as well
[15:05] <briancurtin> DONE: banging on tests, loving python3
[15:05] <briancurtin> TODO: keep on porting, hopefully get to a point where the devtools branch can be merged without breaking SSO
[15:05] <alecu> briancurtin: standup!
[15:06] <ralsina> me (and hey, I am just connecting ;-)
[15:06] <dobey> facundobatista: oh, sorry it's set in dev-tools
[15:06] <dobey> facundobatista:                      ['temp-directory', None, '_trial_temp', None],
[15:07] <ralsina> DONE: calls, fought the ISP man, and the ISP man won, 1-1s, some reviews, planning TODO: catchup, decisions BLOCKED: not yet
[15:22] <facundobatista> dobey, yeap
[15:23] <facundobatista> alecu, dobey, so, we're setting the _trial_temp directory to unicode, and then in expand_user() in the client we assert isinstance(result, str)
[15:24] <facundobatista> alecu, dobey, what should we do?
[15:25] <facundobatista> alecu, dobey, if we remove the
[15:25] <facundobatista>                      ['temp-directory', None, '_trial_temp', None],
[15:25] <facundobatista> line,
[15:25] <dobey> facundobatista: and a test in client is failing?
[15:25] <facundobatista> it uses the same value from twisted, that is bytes
[15:25] <facundobatista> dobey, yeap
[15:26] <facundobatista> dobey, u1trial -t TestRequestQueue.test_with_one_run tests/syncdaemon/test_action_queue.py
[15:26] <facundobatista> a bunch in there, but only because they use expand_user()
[15:26] <facundobatista> I mean, not the test is failing
[15:26] <facundobatista> it uses expand_user, production code, that has an assert that explodes to the temp dir that is being set when running the tests
[15:29] <dobey> huh, why is it not failing in tarmac or nightlies?
[15:30] <facundobatista> dobey, it's a very good question
[15:33]  * gatox lunch + errands
=== gatox is now known as gatox_lunch
[15:35] <facundobatista> dobey, maybe something is not really updated?
[15:36] <facundobatista> dobey, in nessita's machine (I'm here with her), that has an english machine and all updated, it fails also with the same
[15:36] <nessita> facundobatista: you wanted this output? http://pastebin.ubuntu.com/1204997/
[15:36] <dobey> facundobatista: no, everything is updated in tarmac/nightlies. it fails for me doing the u1trial -t to only run that test, but if i do u1trial tests/syncdaemon/test_action_queue.py it passes
[15:37] <facundobatista> dobey, oh
[15:37] <dobey> facundobatista: so perhaps there is some issue with isolation in that file?
[15:37] <facundobatista> ugly indeed
[15:37] <facundobatista> dobey, or maybe it just changes the order in how options are taken
[15:38] <facundobatista> dobey, remember that u-dev-tools are setting (in unicode) the same setting that twisted does (in bytes)
[15:38] <facundobatista> so, if u-dev-tools one comes second, all breaks, if twisted comes second, all works
[15:39] <facundobatista> (the "comes first/second" that I mean is in u-dev-tools/devtools/utils.py, _parse_arguments)
[15:40] <dobey> facundobatista: i don't think that's possible?
[15:40] <facundobatista> dobey, the order change?
[15:41] <dobey> yes
[15:44] <facundobatista> dobey, I'm checking that right now
[15:45] <facundobatista> dobey, no order change: if you don't run with the "-t", the u-dev-tools are not set into the game
[15:46] <facundobatista> dobey, so, without "-t" (even running the whole test file, not the whole test suite), only the twisted's 'temp-directory' option is taken
[15:46] <dobey> if i print repr of the default value, u'_trial_temp' is printed second in both cases
[15:46] <dobey> facundobatista: how did you determine that?
[15:47] <facundobatista> dobey, in u-dev-tools' utils.py, put this print:
[15:47] <facundobatista>         print("========= adding", class_obj, class_obj.__dict__.get(attr, []))
[15:47] <facundobatista> in line 44 (before list_obj is extended)
[15:47] <facundobatista> dobey, it tells you where the options are, and from where they came
[15:53] <dobey> facundobatista: both are added in both cases, for me, afaict
[15:54] <dobey> facundobatista: http://pastebin.ubuntu.com/1205030/ is what i see
[15:55] <facundobatista> dobey, which print did you put?
[15:55] <dobey>         print('==add===', class_obj, class_obj.__dict__.get(attr, []))
[15:56] <facundobatista> dobey, really? I see a lot more of addings, not only those two each time
[15:56] <facundobatista> ah, you're grepping, sorry
[15:56] <dobey> right :)
[15:59] <dobey> facundobatista: i need to get something to eat, but i think the first thing we need to do is figure out why it isn't failing in the case of running the whole suite, and fix it to ensure it always fails. will poke a little more after lunch :)
[16:00] <dobey> bbiab
[16:16] <facundobatista> dobey, I also put a print in the expand_user, call, see this:
[16:17] <facundobatista> when running only that test: https://pastebin.canonical.com/74581/
[16:17] <facundobatista> when running everything: https://pastebin.canonical.com/74582/
[16:17] <facundobatista> in the second pastebin, the _trial_temp is broken, but if you go to line 1060 you see that expand_user is not being called for that same test!
[16:18] <facundobatista> dobey, *probably* because the expand_user() call comes from a chain of calls regarding configuration, and the config is already up from another value?
[16:32] <mandel> alecu, just pushed a version to lp:~mandel/avani/u1-payment-preview with a decent text input (first version, I need to clean the code) but does have the bits we need, looks like the search bar and has password support
[16:33] <mandel> alecu, in case you want to take a look
[16:37] <mandel> EOD here, see you all!
[16:37] <mandel> EOD here, see you all on monday!
[16:37] <mandel> have a great women
[16:47] <ralsina> "have a great women"? WAT
=== gatox_lunch is now known as gatox
[16:55] <briancurtin> nice, the whole vmware program froze...
[17:05] <ralsina> briancurtin: detailed emulation.
[17:12] <briancurtin> :)
=== trijntje_ is now known as trijntje
[17:26] <dobey> facundobatista: ah, possibly
[17:28] <dobey> facundobatista: though weird that it's not unicode in any of those other tests, no?
[17:35] <facundobatista> dobey, indeed
[17:37] <dobey> facundobatista: ah, i wonder if it's one of the things that patches user_home?
[17:39] <dobey> ah-ha!
[17:39] <dobey> indeed
[17:40] <alecu> "have a great women" indeed, mandel!
[17:40] <dobey> facundobatista: it's because BaseTwistedTestcase in u1client (in contrib.testing.testcase) patches ubuntuone.platform.user_home, so when you run that one test alone, it doesn't happen because it doesn't use that test case, but only the twisted test case
[17:41] <facundobatista> dobey, great you found it!
[17:43] <dobey> facundobatista: but now, how to fix it? kill the patching of user_home in that test case?
[17:44] <dobey> facundobatista: indeed, without that, i get a LOT of failures :)
[17:47] <facundobatista> dobey, *one* way to fix it is to remove this line from u-devtools
[17:47] <facundobatista>                   ['temp-directory', None, '_trial_temp', None],
[17:48] <dobey> facundobatista: or make it be b'_trial_temp' instead rather
[17:48] <dobey> facundobatista: but that also doesn't fix the problem of the tests not failing when it should :)
[17:48] <facundobatista> dobey, ah, yes :)
[17:49] <dobey> well i made the tests fail, then pass again; only changing u1client, but i don't know if we want an .encode() there
[18:18] <dobey> mmcc: ping
[18:19] <mmcc> what's up dobey?
[18:19] <dobey> mmcc: hey. i'm making a branch to windows-installer to add ubuntuone-client-data to the sourcedeps branches; i see in the plist file a couple of references to .icns files, but don't see them anywhere. are they just not being included yet?
[18:20] <mmcc> looking…
[18:22] <mmcc> yeah, the icns is just on my system. I didn't include it because I knew you were working on icon stuff
[18:23] <mmcc> now that you point it out, the plist has a redundant entry for the icon :) only one of those needs to exist
[18:24] <dobey> ok. i know basically nothing about plist files. can i point it to a file inside the tree of a sourcedep?
[18:25] <mmcc> the plist is only read by the app once it's bundled, so it only needs the filename, and setup-mac will have to copy from wherever the sourcedep is into the app wrapper
[18:26] <mmcc> feel free to assign me a bug for making it do that, just let me know where to find client-data in the buildout - I'm assuming parts/ubuntuone-client-data/ if you're just adding it like the other sourcedeps
[18:28] <dobey> ah ok
[18:28] <dobey> i am just adding it like the other sourcedeps
[18:35] <dobey> brb
[19:38] <dobey> hrmm, maybe i should put it under sources, rather than sourcedeps
[19:39] <dobey> mmcc: where inside the tree does it need to be copied to, to add it to the .app?
[19:41] <mmcc> dobey: not sure what you're asking - are you changing setup-mac or something else?
[19:41] <dobey> mmcc: yeah, to copy the ubuntuone.icns in
[19:42] <mmcc> ok, actually no need to change code in setup-mac -- as long as ubuntuone.icns is in the buildout/scripts/data/ directory, setup-mac will grab it and put it in the right place
[19:43] <mmcc> oh, but if there's a setup in client-data, setup-mac should call that in the prepare step, then copy the result in to scripts/data
[19:43] <dobey> ah, it's in a separate project. i'm fixing u1-win-installer to pull that project with buildout, and pull the icons for mac/win
[19:43] <mmcc> make sense?
[19:44] <dobey> well the .ico and .icns are just stored in vcs; setup.py doesn't isntall them. and generating the icons on mac/win will be painful
[19:44] <mmcc> is the icns generated by client-data or just stored?
[19:44] <mmcc> ah ok
[19:44] <dobey> both
[19:44] <dobey> if you change the .svg and run the build_png or build_icns command, it will want to regenerate the PNGs, and then regenerate the icns file
[19:45] <dobey> well build_icns will regenerate the icns file
[19:45] <dobey> which we will need to do on mac at some point
[19:45] <mmcc> ok, but if I just grab a new branch, the icns will be there?
[19:46] <mmcc> then we just need to add something like what we do for the certs in setup-mac.py:407 -- just copy the .icns into data/ in the prepare step
[19:46] <dobey> yes
[19:47] <mmcc> yeah, just a simple shutil.copyfile and you're set
[19:47] <dobey> but i need to change client-data a bit, becuase the png2icns tool on linux doesn't support the fancy @2x sizes and such
[19:47] <dobey> ah ok, to data/ then
[19:47] <mmcc> yep
[20:02] <gatox> eod here!! need to start preparing the things for the PyDay Cba... i'll see some of you tomorrow! bye! :D
[20:08] <dobey> mmcc, briancurtin: i'll request you both as reviewers for this branch…
[20:08] <mmcc> sounds good
[20:10] <dobey> mmcc, briancurtin: https://code.launchpad.net/~dobey/ubuntuone-windows-installer/add-data/+merge/124509
[20:10] <briancurtin> looking
=== salgado` is now known as salgado
[20:18] <mmcc> argh crashy emacs
[20:23] <mmcc> dobey, in setup-mac.py, the line "+        os.chdir(os.path.join("sources", "ubuntuone-client-data"))" should have os.path.join(self.source_dir, "ubuntuone-client-data") instead
[20:23] <mmcc> usually the sources aren't in sources/, they're in devsetup/parts
[20:23] <mmcc> which is the default for source_dir…
[20:28] <dobey> ah. i was actually editing the plain setup.py when i made that change, and copy/pasted it; i wonder if it should be using self.source_dir too, or is that different between darwin and win?
[20:30] <mmcc> I think it's different
[20:30] <mmcc> yeah, looks like the windows setup assumes everything's in sources/
[20:31] <dobey> yeah, it does. wondering if that is broken though
[20:33] <briancurtin> dobey: i have a change you'll need to make in conf.py to include client-data in the fetch command for the build_installer.py script
[20:33] <briancurtin> re-running it right now then will post it
[20:34] <dobey> ah, i see
[20:37] <dobey> briancurtin: fixed/pushed
=== yofel_ is now known as yofel
[20:55] <briancurtin> dobey: NameError on sp_revno...but it doesn't appear in trunk either
[20:56] <briancurtin> although it doesnt seem to fail on trunk either, at least as far as i know. trying now
[20:57] <mmcc> pyflakes tells me there's also a (not new to this branch) typo'd "URL" instead of "url" on line 376 of setup.py…
[20:58] <dobey> briancurtin: oh; i see the windows setup.py isn't building storage-protocol either
[20:58] <dobey> briancurtin: why is that?
[20:59] <briancurtin> no idea
[21:00] <mmcc> I don't think it needs to
[21:01] <dobey> where does it get the _pb2.py files which are built by protobuf from?
[21:01] <dobey> setup-mac.py does build it
[21:02] <briancurtin> dobey: that might be a part of the environment setup, since i know i had to add a way to download and locally install the protoc compiler. i
[21:02] <briancurtin> i'll check on that
[21:03] <alecu> ok, this is an early EOW for me. bye all!
[21:07] <dobey> briancurtin: ok. i can remove that one for now as it wasn't there before. i had presumed it was overlooked (didn't actually check the code)
[21:08] <briancurtin> dobey: i'm running it without sp_revno right now. that appears to have been the last thing
[21:09] <dobey> briancurtin: removed the sp_revno usage, and pushed
[21:15] <mmcc> dobey, works for me, +1
[21:15] <mmcc> brb
[21:16] <briancurtin> dobey: works, approving now
[21:17] <dobey> great, thanks guys
[21:30] <mmcc> ok, lunchtime here. if you guys leave before i get back, have a great weekend. or a great women, your choice
[21:31] <dobey> both! :)
[21:31] <dobey> have a great weekend all!
[21:31] <briancurtin> hahaha
[23:07] <mmcc> ok, heading out. if I hear back from apple's dev support guy about this code signing issue I'm having I might be back later to fix it. great women and men to everyone!
=== m_conley is now known as m_conley_away