[08:18] <davidcalle> Chipaca, ping
[08:54] <JamesTait> Good morning all! :)
[12:02] <gatox> brb...... need to restart
[12:02] <gatox> nessita, hi o/..... brb...... rebooting.....
[12:02] <nessita> gatox: hola!
[12:04] <nessita> hello everyone!
[12:04] <gatox> ni
[12:04] <gatox> hi
[12:19] <nessita> hola gatox
[12:19] <gatox> nessita, buenas
[12:37] <InBar> hi all, is the notes sync service always available in ubuntu one ?
[12:39] <rye> Hello all, we are currently experiencing a filesync service outage and investigating the reason
[12:39] <gatox> nessita, did you get my email yesterday?
[12:39] <gatox> about the checkbox
[12:40] <nessita> gatox: yeap, I did... but I still find those solutions too hacky, and I think that the best thing to save time is ask ralsina which one is the "best" option
[12:40] <gatox> nessita, ok
[12:56] <urbanape> morning, all
[12:56] <gatox> urbanape, hi
[13:07] <dobey> InBar: syncing notes from tomboy should generally be working, yes
[13:17] <davidcalle> Chipaca, ping
[13:34] <InBar> dobey, doesn't work for me because : https://one.ubuntu.com/notes is not found :/
[13:35] <dobey> InBar: the web ui was removed. the instructions on https://one.ubuntu.com/help/tutorial/install-and-setup-notes-sync/ should still work
[13:36] <dobey> make sure you have the trailing / in the URL in the config
[13:47] <cielak> hello everyone
[13:49] <cielak> I have problems with syncing files, is that a global, temporary, known issue, or does anyone have spare few minutes to help me investigating the problem?
[13:50] <dobey> cielak: yes, and it's being looked into
[13:50] <cielak> okay, thanks!
[14:07] <Chipaca> davidcalle: pong
[14:08] <Chipaca> davidcalle: i'm on holidays these two weeks, so you should drop me an email if you need me
[14:08] <Chipaca> at some point i'm going to stop checking irc :)
[14:08] <davidcalle> Chipaca, oh ok sorry about that :) One quick question, is there an eta on the videosearch server change?
[14:09] <Chipaca> davidcalle: no updated eta. But, I'll ask.
[14:09] <davidcalle> Chipaca, thanks :) enjoy your holidays!
[14:14] <facundobatista> mandel, did we release a new client yesterday?
[14:15] <dobey> facundobatista: to ubuntu? yes
[14:15] <facundobatista> dobey, it was a Precise beta or something?
[14:16] <dobey> facundobatista: it was the scheduled releease for all our client projects
[14:16] <facundobatista> dobey, ok, thanks!
[14:16] <dobey> facundobatista: there will be another release next tuesday, and next thursday is precise final freeze
[14:51] <gatox> nessita, dobey when you have a few minutes (small branch): https://code.launchpad.net/~diegosarmentero/ubuntuone-client/syncdaemon-q/+merge/100984  (i think you are the only ones around)
[14:51] <nessita> gatox: ack!
[14:53] <dobey> lies
[14:53] <briancurtin> i'm around...just fighting til death with windows firewall
[14:55] <nessita> briancurtin: I hijacked mandel's branch re: autoupdate, since the branch he proposed  can not be landed as is (comment in MP explains why). WOUld you be able to review it and IRL test it?
[14:55] <briancurtin> nessita: yep. do you have a link?
[14:55] <nessita> briancurtin: in a couple of minutes
[14:58] <dobey> i presume we're skipping the call today?
[14:59] <briancurtin> may as well just to the text standup
[15:02] <nessita> briancurtin: +1
[15:02] <nessita> dobey: yeah, let's just do the IRC standup
[15:02] <nessita> so, me?
[15:02] <briancurtin> me
[15:03] <dobey> thisfred and urbanape are around too no?
[15:03] <urbanape> me
[15:03] <thisfred> me
[15:03] <dobey> meh
[15:03] <thisfred> DONE: started on split words mappin
[15:04] <thisfred> oops
[15:04] <nessita> DONE: ubuntu releases, bug #973689, IRL tested Ubuntu One in a french clean precise VM
[15:04] <nessita> TODO: pickup mandel's branch regarding integration autoupdate to the controlpanel
[15:04] <nessita> BLOCKED: nopes
[15:04] <nessita> NEXT: briancurtin
[15:04] <briancurtin> DONE: hacking on windows firewall stuff for an issue elopio found in the installer (bug #909389)
[15:04] <briancurtin> TODO: get the windows firewall actually working...it's being difficult
[15:04] <briancurtin> BLOCKED: none
[15:04] <briancurtin> NOTE: i have a doctors appointment to go to in 1.5 hours
[15:04] <briancurtin> NEXT:
[15:04] <nessita> gatox: standup?
[15:04] <gatox> me
[15:04] <urbanape> DONE: Got some work done on the iOS Music crashers, not a lot on the Mac port. TODO: Get at least a simple proposal in today, since I'm off tomorrow and next week for Lex's spring break. BLOCK: None
[15:04] <urbanape> thisfred: go
[15:04] <thisfred> DONE: started on u1db split words mapping TODO: finish u1db split words mapping NEXT: dobey
[15:04] <dobey> λ DONE: finished uploads, reviews, bug #604810
[15:04] <dobey> λ TODO: more bug #968555, more of #969262
[15:04] <dobey> λ BLCK: none.
[15:05] <dobey> gatox: go
[15:05] <gatox> DONE:
[15:05] <gatox> Propose a branch for Bug #907479, working on Bug #973830
[15:05] <gatox> TODO:
[15:05] <gatox> Finish with error after removing current device
[15:05] <gatox> BLOCKED:
[15:05] <gatox> nO
[15:05] <nessita> comments anyone?
[15:05] <gatox> nop
[15:06] <nessita> reminder: I'm starting a rotation into consumer software since next Monday, for 3 months, so I will be around in IRC but will not be the teach lead during that period. The new teach lead will be alecu :-)
[15:06] <nessita> so, any questions or concerns you should ping ralsina/alecu
[15:06] <urbanape> gracias, nessita, and good luck!
[15:06] <gatox> nessita, yep..... good luck!
[15:06] <nessita> thanks! :-)
[15:06] <nessita> ok, eom?
[15:07] <gatox> nessita, yep.... and start of llunch for me :P
[15:07] <nessita> ack!
[15:07] <dobey> "reminder" ?
[15:07] <dobey> that is new to me :)
[15:08] <nessita> dobey: really?!?!? we talked about this in a weekly call
[15:08] <nessita> like a month ago?
[15:08] <urbanape> yeah, even I knew that
[15:08] <dobey> i don't recall it
[15:11] <thisfred> nor do I, but then I probably was not on that call
[15:12] <thisfred> sounds like fun though, hope it is
[15:12] <dobey> it was probably mentioned during some talk about windows details and i tuned out :)
[15:13] <thisfred> "yes, dear"
[15:13] <nessita> :-)
[15:21] <nessita> briancurtin: the branch I mentioned https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/autoupdate/+merge/100989
[15:21] <dobey> alright, need to get lunch. bbiab
[15:30] <briancurtin> nessita: ack, i'll take a look
[15:30] <nessita> briancurtin: thanks!
[16:00] <briancurtin> urbanape: i'm doing another doctor/lunch combo so i'll be leaving in a bit, for a bit...but this afternoon do you want to mumble for a bit before you head out for your days off?
[16:00] <urbanape> briancurtin: yeah, that sounds good
[16:01] <briancurtin> i'm jammed up with trying to figure out windows firewall stuff, but may as well do a handoff, which i can then include mandel on when he's back next week
[16:05] <duanedesign> nessita: helping a user on Windows. Logs seem a little sparse. Could this be a clue to what his issue is? https://pastebin.canonical.com/63880/
[16:06] <nessita> duanedesign: looking... what's the symptom?
[16:07] <duanedesign> nessita: "at the setup page, when I log in, I get the "Please wait, getting information" message and there it waits endlessly."
[16:07] <nessita> duanedesign: syncdaemon is definitely stuck trying to load the metadata, which is perhaps crashing (since we have an opened bug for logging when the metadata load fails)
[16:08] <nessita> duanedesign: has this user been testing U1 since the early beta?
[16:10] <duanedesign> nessita: hmm, does not say. "I had to re-install software following a re-build,  I did a complete uninstall and registry clean before installing again"
[16:10] <duanedesign> nessita: hmm, the credentials log has "DEBUG - Handling signal_name: 'on_credentials_not_found_cb', app_name: u'Ubuntu One'."
[16:11] <nessita> duanedesign: my money is that this user has corrupted metadata
[16:11]  * duanedesign nods
[16:11] <nessita> duanedesign: so he should quit eveything u1 related, remove the folder where the metadata is, and re install
[16:12] <nessita> duanedesign: the problem is that the uninstaller does not clean metadata (and is not clear we want it to do it, since is very sensitive)
[16:12] <duanedesign> thank you
[16:12] <nessita> anytime!
[16:23]  * briancurtin doctor+lunch
[17:07] <nessita> lunchtime on this end
[17:07] <nessita> brb
[17:14] <dobey> thisfred, briancurtin, gatox, nessita: anyone care to review https://code.launchpad.net/~dobey/libubuntuone/system-oauth/+merge/100977 ?
[17:14] <thisfred> sure
[17:15] <gatox> dobey, ok, on iit
[17:15] <dobey> thanks
[17:23] <gatox> dobey, +1
[17:28] <nessita> gatox: could you please review https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/autoupdate/+merge/100989 ?
[17:28] <gatox> nessita, yep!
[17:28]  * gatox reviewing....
[17:28] <nessita> gatox: also, when you can, would you answer the question from dobey here? https://code.launchpad.net/~diegosarmentero/ubuntuone-client/syncdaemon-q/+merge/100984
[17:31] <gatox> nessita, dobey i answer here to avoid delay..... we can't change the if to: "if running and options.quit:" because you will be leaving out a lot of options....... i can avoid printing the message and just exit if you think is best..... i added the message to show some kind of feedback and the user knows that actually something was executed
[17:31] <nessita> gatox: can you please add the explanation to the merge proposal? so we can keep track of the conversation
[17:32] <gatox> nessita, yes, of course
[17:32] <nessita> thanks!
[17:32] <dobey> i don't understand the "you will be leaving out a lot of options"
[17:34] <gatox> nessita, if you change the if that way..... you are saying that the code is only going to be executed if sd is running..... and the user select the "quit" option....... but you can select also: start, status, etc
[17:34] <gatox> nessita, sorry
[17:34] <gatox> dobey,  ^^
[17:37] <dobey> oh, that same block of code is used for all those options? but then you could just have it be "if running:" no, and leave out the options.quit check there?
[17:39] <gatox> dobey, but maybe you want to execute: --start....... and in that case, sd probably is not running
[17:39] <dobey> ah right
[17:40] <gatox> you could do: not (not running and options.quit)
[17:40] <dobey> that code really needs to be modularized better :)
[17:40] <gatox> if you want to avoid the print message
[17:40] <dobey> eh
[17:41] <gatox> dobey, but my question is...... do you really think that it's better to just don't print anything? i mean, i added that because i thik is useful for the user to see that actually something was executed
[17:41] <gatox> nessita, opinion?
[17:41] <nessita> gatox: haven't looked at the code yet, let me do that
[17:41] <nessita> (I'm trying to finish a branch ASAP :-))
[17:42]  * nessita branches
[17:42] <dobey> gatox: u1sdtool -c doesn't print anything
[17:42] <dobey> gatox: should it wait until connected (or until failure), and print something?
[17:43] <gatox> dobey, ok, either way....... is fine by me..... i just thought that a message would be useful....... but is not something really critical
[17:43] <gatox> dobey, sorry.... i didn't understand your last question
[17:43] <dobey> gatox: eh, i didn't realize the code was so conflated at that point
[17:44] <nessita> so, without using this branch, in linux I get:
[17:44] <nessita> nessita@dali:~/canonical/client/review_syncdaemon-q$ u1sdtool -q
[17:44] <nessita> ubuntuone-syncdaemon stopped.
[17:44] <dobey> gatox: it was about u1sdtool -c not printing anything currently
[17:44] <nessita> using this branch, I m getting:
[17:44] <nessita> nessita@dali:~/canonical/client/review_syncdaemon-q$ PYTHONPATH=. bin/u1sdtool -q
[17:44] <nessita> ubuntuone-syncdaemon not running.
[17:44] <dobey> nessita: that sounds correct, assuming it's not running
[17:44] <nessita> is not running, indeed
[17:45] <gatox> correct
[17:45] <nessita> gatox: what was causing the "    sync_daemon_tool = SyncDaemonTool(bus)" at the beginning of main?
[17:45] <dobey> gatox: anyway, i approved, but with the comment that we should probably split the logic up for that code later
[17:45] <nessita> gatox: my question being, why you moved it?
[17:45] <gatox> nessita, but we are talking with dobey if the message is really necessary or  i sohuld remove it
[17:45] <nessita> gatox: I know
[17:45] <nessita> gatox: but first, I wonder why you moved the SyncDaemonTool creation
[17:46] <nessita> since in linux that does not start the service
[17:46] <nessita> if the simple fact of creating an instance of SyncDaemonTool starts the service, then that's a bug
[17:46] <nessita> and we need to solve that
[17:46] <gatox> nessita, because SyncDaemonTool was starting the proxy, sd and sso in windows.....
[17:46] <nessita> gatox: makes sense?
[17:46] <nessita> gatox: then that's  the bug we need to fix
[17:46] <nessita> gatox: because it shouldn'y
[17:47] <nessita> t*
[17:47] <dobey> nessita: it's a bug we need to fix
[17:47] <nessita> dobey: yes
[17:47] <dobey> nessita: i think moving the instantiation of the variable to where it's needed, rather than at the beginning, when it may not be needed, is also the right thing to do
[17:48] <dobey> nessita: so while it might mask that bug, it is a separate bug, and i don't think the change shouldn't be made, simply for that reason
[17:48] <nessita> dobey: SyncDaemonTool should be the instance to ask if the service is running, instead of having a separated method... so, considering that, I think we should solve both
[17:49] <dobey> if there is a compelling reason it shouldn't be moved, then i might agree, but i don't see any compelling reason not to move it in the branch as-is.
[17:49] <gatox> nessita, eh? we are not asking SyncDaemonTool is sd is already running
[17:49] <nessita> gatox: I know, I'm saying we should
[17:49] <nessita> gatox: but that's a separated issue
[17:49] <dobey> nessita: i think that should be done in a second branch perhaps then
[17:49] <nessita> sure
[17:49] <nessita> but I don't want this to land wthout having the real problem solved
[17:49] <nessita> after that, we can land this as well
[17:51] <dobey> then vote needsfixing and block it until the other issue is fixed. i'm not picky about it myself, and don't think it matter which order they're fixed in :)
[17:53] <nessita> dobey: ack
[17:53] <nessita> gatox: do you understand my point
[17:53] <nessita> ?
[17:53] <gatox> nessita, yes
[17:53] <nessita> gatox: so is like we have dirt in the floor, and instead of wiping it, we put it under the carpet
[17:54] <gatox> nessita, i'll look for the other issue, and add the solution in this branch
[17:54] <nessita> gatox: thanks!
[18:05] <dobey> gatox: if it's more than a couple lines, could you please do it in another branch?
[18:05] <gatox> dobey, ok
[18:06] <dobey> gatox: i know it can be annoying, but it's nice to keep branches small and focused on a single bug, particularly if we need to go back and look at or revert things in the future :)
[18:30] <thisfred> and with bzr pipeline it's not even that annoying
[19:01] <briancurtin> nessita: sorry for the delay, but IRL the autoupdate branch works. scanning the diff right now
[19:01] <nessita> briancurtin: no problem. Branch is approved and landed, but please let me know if you want me to change anything, I will quickly propose another one
[19:02] <nessita> briancurtin: if branch looks good, would you be able to build a new installer including it and (perhaps) also adding your firewall fix?
[19:02] <nessita> briancurtin: so we can send that to QA
[19:03] <briancurtin> nessita: i have an installer right now with your autostart branch, but the firewall fixes do not seem to be working, at least not consistently
[19:03] <nessita> briancurtin: :-(
[19:03] <nessita> any idea why?
[19:03] <urbanape> briancurtin: also, good news: I'll be around tomorrow morning at least, if not all day.
[19:04] <briancurtin> nessita: no clue. it doesnt really make sense since i only changed the code once, saw it work on XP, then it didnt work, then it worked, then it hasnt worked since yesterday
[19:04] <gatox> nessita, ping
[19:04] <nessita> gatox: pong
[19:05] <briancurtin> nessita: i've also reset the firewall settings, tried it all over...no luck. i dont really know what to say :/
[19:05] <nessita> briancurtin: gosh, I have no idea what to advice
[19:05] <briancurtin> urbanape: cool. since i might have to take the firewall thing back to the drawing board, mind if we call tomorrow?
[19:06] <urbanape> yeah, no worries
[19:06] <nessita> briancurtin: is this under an admin or not admin account?
[19:07] <gatox> nessita, ahhhh.... wait.... nothing.... i need to check something further
[19:07] <nessita> gatox: ok! :-)
[19:07] <briancurtin> nessita: it'll always be admin since the installer requests/requires elevation
[19:07] <nessita> briancurtin: right, but then I'm confused since the wizard is not run under an elevated scenario, no?
[19:07] <nessita> briancurtin: so, in my head, elevation goes like this (please correct me):
[19:08] <nessita>  user downloads our exe
[19:08] <nessita> user double clicks it
[19:08] <nessita> windows asks elevation, user grants
[19:08] <nessita> installer finishes, elevation is "finished"
[19:08] <nessita> user is presented with the controlpanel
[19:08] <nessita> at that last point, user has not elevated perms?
[19:08] <nessita> right?
[19:09] <nessita> so, when the controlpanel actually runs, can't "access" the firewall exception since is a regular user?
[19:09]  * nessita may be saying nonsenses
[19:09] <briancurtin> nessita: this is all happening inside of the BitRock installer, so it's occuring within an elevated context. i actually see that the firewall exceptions are correctly being added, but something about them is not working properly
[19:10] <briancurtin> so while elevated, we mess with the firewall, then move onto non-elevated where CP starts and *should* already have access through the firewall.
[19:10] <nessita> briancurtin: what about that last bit I mentioned? something like firewall exceptions being added only for admin users (is that even possible?)
[19:10] <briancurtin> nessita: nah it's just a per application basis. you just say "C:\foo.exe" is allowed
[19:10] <nessita> briancurtin: sorry if I'm saying stupid things, just trying to think crazy options
[19:11] <nessita> briancurtin: what about the path for the exe being wrong?
[19:11] <briancurtin> nessita: it won't add it if the path doesn't exist - i've done it both manually and through the installer
[19:11] <nessita> briancurtin: ah, ok
[19:11] <nessita> :-/
[19:11] <nessita> briancurtin: did you try restarting in between? (another uninformed suggestion)
[19:12] <nessita> in between installer finishes and controlpanel is opened
[19:12] <briancurtin> nessita: yeah i've tried that as well
[19:13] <nessita> briancurtin: is this only for XP? does w7 works?
[19:14] <briancurtin> nessita: its for both - neither work. the cases where i've seen it work have been on XP from the installer, and i've seen it work on Win7 when everything is running from python.exe (ie. not py2exe'd binaries)
[19:14] <briancurtin> but it worked on XP once or maybe twice, none since then
[19:14] <nessita> briancurtin: in any case, please prioritize w7 in your debugging
[19:14] <briancurtin> nessita: that's what i started to do, especially since its easier to work on
[19:15] <nessita> briancurtin: since XP is not officially supported (though we try really hard to have it working)
[19:15] <nessita> briancurtin: question: does the firewall exception specifies a port number?
[19:16] <briancurtin> nessita: you can do specifics like that - for now i have it set to just allow all access
[19:17] <nessita> briancurtin: another question, just to see if my uneducated-in-windows intuition comes up with something, what's the "command" that we run to add this exception?
[19:19] <briancurtin> for XP: netsh firewall add allowedprogram "C:\Program Files\dist\ubuntuone-syncdaemon.exe" "Ubuntu One" ENABLE
[19:19] <briancurtin> for vista/7: netsh advfirewall firewall add rule name="Ubuntu One" dir=in action=allow program="C:\Program Files (x86)\dist\ubuntuone-syncdaemon.exe" enable=yes
[19:19] <briancurtin> the vista/7 one you need to run again with "dir=out" to set both inbound and outbound allowances
[19:19] <nessita> briancurtin: does the "C:\Program Files (x86)\dist\ubuntuone-syncdaemon.exe" is calculated at installation time?
[19:20] <briancurtin> nessita: yep. i've looked in our installer logs as well as the actual firewall exception list to ensure that the proper path is being added
[19:20] <briancurtin> plus the installer will fail if that command fails, so if you don't get a CP after install then something went wrong, but i figured that part out early on
[19:22] <nessita> briancurtin: question aside: are we creating a 'dist' folder right at the program files dir root?
[19:23] <nessita> don't we have an 'ubuntuone' folder in between?
[19:23] <nessita> last time I tested the installer (before you joined the company :-)) we had paths like C:\Program Files\ubuntuone\dist\blah
[19:23] <nessita> so perhaps that changed?
[19:23] <briancurtin> nessita: yeah sorry, it's "C:\program files (x86)\ubuntuone\dist\ubuntuone-syncdaemon" -- i typed it out by hand because its actually "${installdir}\dist\ubuntuone-syncdaemon.exe" in the code
[19:24] <nessita> ah, ok
[19:24] <nessita> so, dead end there
[19:24] <nessita> :-/
[19:24] <nessita> briancurtin: last annyoing question: does the name has to match the directory name under Program Files?
[19:25] <nessita> perhaps ubuntuone != Ubuntu One is breaking the firewall rule (doesn't make much sense, I know, but is windows)?
[19:25] <briancurtin> nessita: nah it doesnt have to match, it's just a friendly identifier - none of the others in there match their executable name
[19:26] <nessita> ok, so I'm out of ideas
[19:26] <briancurtin> nessita: im almost out too. i might try another way to add the exception and see if that does anything better
[19:27] <nessita> briancurtin: ok... if we don't add the exception, does the user gets the firewall prompt at least?
[19:28] <briancurtin> nessita: well they do but not during the installer setup portion. it will just show the progress dialog forever at the cloud-to-computer setup, then you hit next through the installer, then you'll be prompted for the exception
[19:28] <nessita> briancurtin: do you know why the user does not get prompted sooner? like, as soon as syncdaemon tries to connect
[19:28] <nessita> that will happen right at the cloud to computer page
[19:29] <briancurtin> nessita: i'm not really sure yet. i'll see if i can find a way to make it come up sooner or something
[19:29] <nessita> briancurtin: perhaps that helps us debugging why the exceptions does not work. "perhaps" the exception list is not queried in that page, which will explain why the exception is not working
[19:30] <nessita> I mean, if we have no exceptions at all, and even so the user is *not* prompted in the first wizard page, that sounds like the firewall is not being "inspected"?
[19:30] <nessita> is like the firewall (and its exceptions) are queried too late?
[19:31] <nessita> too late for us, I mean
[19:33] <briancurtin> nessita:  i'll try to step in and see where an outbound connection is being made, and why it's not succeeding at this point
[19:33] <nessita> briancurtin: good luck :-)
[19:59] <nessita> will grab some snacks, brb
[20:01] <gatox> ok.... EOD here..... see you tomorrow (if there is someone here)
[20:01] <gatox> nessita, enjoy your rotation!
[20:07] <dobey> hrmm
[20:07] <dobey> nessita: https://bugs.launchpad.net/ubuntu/+source/ubuntu-sso-client/+bug/961563
[20:23] <nessita> dobey: will debug
[20:24] <dobey> nessita: do you know what the lp project for the sso server side/service is?
[20:26] <nessita> dobey: yes, canonical-identity-provider
[20:26] <dobey> thanks
[20:26] <nessita> prego
[20:29] <nessita> dobey: the bug report for sso makes no sense... we have UI_PROXY_CREDS_DIALOG defined in ubuntu_sso/__init__.py:39
[20:30] <nessita> so the import error smells like corrupted pycs
[20:30] <dobey> i didn't file it or look at it really. i just saw it and thought you should look at it :)
[20:30] <nessita> dobey: of course, was mainly braindumping with you
[20:30] <dobey> if you think it's due to busted install, feel free to suggest some dpkg-reconfiguring :)
[20:32] <dobey> on the other hand, the bug i just filed, is a real bug :)
[20:32] <dobey> bug #974606
[21:01] <dobey> also, this: https://bugs.launchpad.net/ubuntu-sso-client/+bug/974637
[21:01] <dobey> which is extremely frustrating right now :-/
[21:01] <nessita> dobey: agreed
[21:03] <dobey> nessita: it is particularly frustrating for me, because i am debugging a problem where the token is getting deleted, and the register dialog pops up
[21:04] <dobey> and there is no "i just want to log in already" button
[21:04] <nessita> dobey: who is raising the register dialog?
[21:05] <dobey> nessita: libubuntuone/libsyncdaemon (music store)
[21:06] <nessita> dobey: oh right, because is calling register
[21:06] <nessita> because the register method use to show a screen with both options :-/
[21:06] <nessita> used* to
[21:06] <nessita> when we used the GTK+
[21:06] <dobey> yes
[21:06] <nessita> dobey: so, this bug will also affects our music store users, right?
[21:06]  * dobey misses the nice simple gtk+ dialogs
[21:07] <dobey> nessita: well it affects anyone who wants to do anything that needs sso
[21:07] <nessita> right :-(
[21:07] <dobey> at least, anyone who installs ubuntuone-control-panel-qt
[21:07] <nessita> :-(
[21:07] <dobey> (or ubuntu-sso-client-qt if they just manually install that)
[21:08] <dobey> because imagine you click the u1 icon, install the control panel, register or log in, then quit, go open the software center, and try to rate something, and it asks you to register; and there's no way to click "just log in"
[21:09] <nessita> dobey: well, software center is using the gtk UI
[21:09] <nessita> so the register screen is showing the login option
[21:10] <briancurtin> nessita: i just got cloud-to-computer to work...once! rather than have u1cp start up SD and sso, i started them manually, then started up u1cp after they were running. it worked 5 minutes ago. same thing is not working right now. WTF.
[21:10] <dobey> nessita: eh? sso pops up the qt ui by default
[21:10] <dobey> nessita: or is it somehow explicitly requesting gtk+ ui now?
[21:10] <nessita> dobey: nopes, sso pops up the gtk by default
[21:10] <nessita> dobey: u1.credentials pops up qt by default
[21:11] <briancurtin> nessita: this isn't even firewall related, actually. i have no firewall exception and this worked 5 minutes ago without it. i guess the firewall exception will be needed at a later point, but its not the current issue FWIW
[21:11] <nessita> briancurtin: WTF indeed.... is there such thing as 'firewall exception being deleted after a single use'?
[21:11] <nessita> briancurtin: ouch
[21:14] <dobey> nessita: ok, well it's ok then, but maybe Deja Dup isn't since it has the u1 thing
[21:14] <nessita> dobey: agreed to that
[21:14] <dobey> nessita: either way, this is quite annoying, particularly given that the gtk+ and qt dialogs are also quite different :-/
[21:14] <briancurtin> i wish the sso login dialog would just exit once you're logged in, rather than forcing you to click to move on
[21:15] <dobey> indeed
[21:15] <nessita> briancurtin: agreed to that, but there are some drawbacks that are not easy to solve to have a "better" ux  experience
[21:15] <nessita> so, it needs design
[21:15] <dobey> what drawbacks?
[21:15] <nessita> mostly the whole sso needs design
[21:16] <dobey> well, i guess it's hard if the login bit is actually a separate process, yeah
[21:16] <nessita> dobey: if the caller does not behave "properly" after requiring sso services, the user may see a window popping up asking for credentials, the that windows will just dissapear
[21:17] <nessita> and, again, if the caller crashes or does not behave, the user may be very very confused
[21:17] <nessita> about how the process ended
[21:17] <dobey> well, that's a solvable problem, actually
[21:17] <nessita> dobey: meaning?
[21:17] <dobey> we can exit if the calling process goes away
[21:18] <nessita> dobey: you, as sso service that opened the UI, will never know that
[21:18] <dobey> though i agree things are much easier if it's in-process
[21:18] <dobey> nessita: you will if one of the required arguments is the PID, and you do some fancy things to know when the process crashed
[21:19] <nessita> dobey: perhaps, but there is a simpler case where the caller did not crash, but just hangs
[21:19] <nessita> or misbehaves
[21:19] <nessita> anyways, I agree the "finish" screen sucks, though I don't see a way, that does not suck in a different way, to solve that
[21:19] <dobey> which is why we need in process API
[21:20] <dobey> but python is not a good way to do that
[21:21] <dobey> i see a way, but it means a lot of work in sso client
[21:23] <dobey> but of course, there are deeper and deeper problems
[21:32] <dobey> ok, am off. have a good evening :)
[21:39] <nessita> bye dobey
[22:00] <briancurtin> nessita: in u1cp, we hang at self.sd_client.refresh_volumes() inside volumes_info of ubuntuone/controlpanel/backend.py when u1cp is started and needs to start up sso and SD by itself. when you start SSO and SD manually, then u1cp, that refresh_volumes succeeds and cloud-to-computers works nicely
[22:03] <nessita> briancurtin: one sec, otp
[22:05] <nessita> briancurtin: back. Any idea why refresh_volume sis hanging? it can take a little while, but should not hang, afaik
[22:06] <briancurtin> well i left my VM running overnight and it never finished, likely from being stuck at refresh_volumes in this case
[22:07] <nessita> briancurtin: that's odd
[22:07] <briancurtin> nessita: i didn't look further into it yet. doing that next
[22:07] <nessita> briancurtin: ack, thanks for the heads up
[22:07] <briancurtin> at least i now know where it's getting stuck
[22:07] <nessita> briancurtin: right. We can certainly disable that check if we need to release and haven't solved this yet, but it will be interesting knowing what the heck is going on :-)
[22:08] <nessita> and by "check" I meant 'refresh_volumes'
[22:08] <briancurtin> yep, i'm EOD'ing shortly but i'll spend a little time poking into refresh_volumes and then see how it works without it
[22:09] <nessita> briancurtin: ack, I'm eoding as well
[22:09] <nessita> briancurtin: I guess you can follow up on this with alecu next Monday, tomorrow will be a difficult day to get help :-/
[22:10] <briancurtin> nessita: i'm doing a mac call with urbanape in the morning then i'll try to push this forward and maybe get an installer ready for rmcbride if it all works out
[22:10] <nessita> briancurtin: good luck :-)
[22:11] <rmcbride> briancurtin: cool. Definitely ping me. I'm fixing a lot of stuff in the old sd integration suite, so I wn't necessarilly be watching irc directly
[22:11] <briancurtin> rmcbride: will do, thanks
[22:33] <nessita> dobey: if you are around later, or tomorrow, would you please review https://code.launchpad.net/~nataliabidart/ubuntuone-windows-installer/stable-3-0-update-2.99.92/+merge/101045
[22:33] <nessita> when that lands I will release the tarball
[22:35] <nessita> ok, I'm gone
[22:35] <nessita> bye all!!!