[01:07] <dobey> aplund: it is a known issue, and is unrelated to file size it seems. it's been reported already and we're looking into it
[02:24] <aplund> dobey, Gah! I seem to have no chance of syncing this ~100mb file
[02:24] <aplund> and it just chews my download quota as it gets to 10ish mb and restarts
[02:24] <aplund> shame it cannot just resume the transfer
[03:05] <aplund> dobey, is there a bug number or something I can keep an eye on?
[08:28] <JamesTait> Good morning, folks! *8OD
[10:13] <karni_> Mornings :)
[11:15] <gatox> good morning
[11:20] <gatox> reboot...... brb
[11:54] <mandel> gatox, buenas!
[11:54] <gatox> mandel, buenas
[11:54] <mandel> gatox, dude, I hate hitting old qt bugs..
[11:55] <gatox> mandel, what happend now?
[11:55] <mandel> gatox, proxy is not working on O because of the old version
[11:56] <gatox> ahhhhhh
[11:56] <mandel> gatox, if you use the QNetworkAccessManager the second request you do will not fire the signals.. WTF?!
[11:57] <gatox> mandel, crap
[11:59] <mandel> gatox, yes..
[11:59] <mandel> gatox, so I'll be on this bug first since is high and then will move back to fix the broken tests of my sso branch
[12:00] <gatox> mandel, ok..... i'll try to have something of fsevents for today.... i think it's possible
[12:02] <mandel> ajmitch, ck
[12:02] <mandel> ajmitch, ups, sorry, brain fail
[12:02] <mandel> gatox, I meant to say ack
[12:03] <gatox> mandel, jejejee yeeeeeahhh, i'm not the only one who do that!
[12:03] <gatox> jeje
[12:24] <mandel> gatox, I'm off to have lunch
[12:24] <gatox> mandel, ack
[12:25] <alecu> hola gatox, mandel
[12:25] <gatox> alecu, buenas!!
[12:26] <gatox> alecu, did you enjoy your long weekend?
[12:27] <alecu> gatox, in fact: no. My sciatic nerve decided that there would be no long weekend for me. :P
[12:27] <gatox> alecu, ouch :S
[12:27] <alecu> gatox: I had to get a shot, and now I'm taking some pills...
[12:28] <alecu> gatox: the most awful part is that this friday I'll be travelling like too many miles.
[12:28] <gatox> alecu, that happens when you practice extreme sports! :P
[12:28] <gatox> alecu, oh right
[12:28] <alecu> gatox: right: sitting is my sport.
[12:28] <alecu> gatox: extreme sitting.
[12:28] <alecu> :-(
[12:29] <alecu> gatox: anyway: I should be dialing the phone of a very good chinese doctor, to see if he can put my bones in shape before I leave...
[12:30] <gatox> alecu, well.... the good news is that SSO is working in Mac OS..... u1-client/platform is almost completely refactored and we are already working in fsevents (????) if that makes you happy jeje
[12:30] <gatox> alecu, chinese doctor ftw
[12:30] <alecu> gatox: awesome!
[12:31] <alecu> gatox: those are really good news. Congrats!
[12:31] <alecu> gatox: what about the SSL error? do you know if there was any progress on that?
[12:32] <gatox> alecu, no idea....
[12:48] <dobey> bad gatox
[12:48] <gatox> dobey, yes.... i realize..... but it was too late
[12:48] <gatox> dobey, can you abstain?? the problem was already fix in mmcc branch
[12:48] <dobey> gatox: no, both problems are not fixed in his branch. you still need to fix the other error, which is introduced by that branch
[12:49]  * gatox checking....
[12:49] <gatox> oh right..... :S
[12:52] <dobey> gatox: also, you shouldn't make big changes in a branch after it's been approved, like you did for the other branch which resulted in the issue mmcc fixed
[12:52] <gatox> dobey, the change was asked in the need fixing comments
[12:53] <gatox> s/was asked/was requested
[12:53] <ralsina> good morning!
[12:53] <gatox> ralsina, hi!
[12:54] <dobey> gatox: no it wasn't
[12:54] <gatox> dobey, well..... so it wasn't, but was something that we speak with mandel on irc
[12:54] <gatox> sometihng like that was
[12:55] <ralsina> gatox: dobey is right, don't approve changes without re-review.
[12:56] <gatox> ralsina, nono, the problem was that i made the changes :P
[12:56] <ralsina> gatox: I know, and then you clicked "apprve" on the branch :-)
[12:56] <gatox> ralsina, ahh right
[12:57] <dobey> gatox: right, those changes should have resulted in a "resubmit" of that proposal, or should have been in a separate branch. also, mandel shouldn't have voted approve if he had requested you to make changes
[12:57] <dobey> so bad mandel too
[12:57] <gatox> dobey, ok..... lesson learned for the next time
[12:58] <ralsina> yeah, so, moving on... what did I miss?
[12:58] <dobey> the rise and fall of an empire
[12:58] <alecu> hola ralsina.
[12:58] <ralsina> hola alecu
[12:59] <ralsina> alecu: so, back pain?
[12:59] <alecu> ralsina, I hear that SSO is fully working on mac
[12:59] <ralsina> alecu: so did I!
[12:59] <ralsina> which is great news
[12:59] <alecu> ralsina, right: my sciatic nerve decided to cut short my long weekend.
[12:59] <alecu> ralsina, I had to get a shot, and now I'm with some pills
[12:59] <ralsina> alecu: ouch
[12:59] <ralsina> alecu: if you are in pain, take the day off
[12:59] <alecu> ralsina, and trying to get some chinese doctor to put my bones right before I travel on friday.
[13:00] <ralsina> alecu: sitting down is *bad* or sciatic
[13:00] <alecu> ralsina, I'm using the nadachair that nessita gave me.
[13:00] <alecu> ralsina, in fact, I'll be using it on friday, or else I will stay home :-(
[13:01] <alecu> ralsina, and yes, perhaps I'll take half day off today, but first I'm trying to find out the state of the SSL bug.
[13:01] <ralsina> alecu: ack
[13:02] <ralsina> alecu: I am on windows today to try build a 3.0.0 with the same versions I used for 2.0.3 if needed
[13:02] <ralsina> alecu: and if that works, I will get everything's numbers and we can retry on brian's machine or something
[13:03] <alecu> ralsina, awesome.
[13:04] <alecu> ralsina, building an installer on your machine sounds like the best plan
[13:04] <ralsina> alecu: at least it's the last known "good" configuration. I hate magical building machines.
[13:05]  * alecu hates magicians
[13:05] <ralsina> and mimes
[13:05] <ralsina> and mimetypes
[13:06] <alecu> btw: I thought we catered for bug 993111. I'm looking at the code to see what's happening.
[13:06] <ralsina> alecu: no, we never fixed it
[13:06] <ralsina> alecu: we knew it would bite us eventually
[13:08] <alecu> oh, ok. That makes a lot of sense then :-)
[13:08] <ralsina> alecu: I don't even know how to fix it
[13:08] <ralsina> alecu: we can't make different mappings for all names like we do for illegal characters
[13:09] <dobey> i think the only way to fix it is to document that we can't reasonably support it
[13:09] <ralsina> dobey: or we can fail to sync o the linux side
[13:09] <dobey> or just don't support it on linux either
[13:09] <ralsina> dobey: but then we are breaking a working linux feature
[13:09] <dobey> right
[13:10] <ralsina> so, basically crap solutions abound.
[13:10] <alecu> ralsina, dobey: I remember spending a few hours discussing the char mapping issue with __lucio__ and facundobatista, and coming up with an alternate solution
[13:11] <alecu> it must be in some old bug... somewhere.
[13:11] <ralsina> alecu: if you remember it, share because I am drawing a blank :-)
[13:11] <alecu> I think it involved having a per-computer local table where server filenames were translated into local filenames.
[13:12] <dobey> alecu: that won't really work
[13:12] <ralsina> alecu: yikes
[13:12] <alecu> dobey, why not?
[13:12] <dobey> how do you deal with "foo.jpg" vs. "foo.JPG" vs. "foo.Jpg" vs. Foo.jpg" etc?
[13:13] <ralsina> dobey: yes, it would work. It would create those exact names on linux and some bad alternative on windows.
[13:13] <__lucio__> the thing is that if tyhe mapping is not one to one, bad situations arise, as server side names are only one. and case sensitive.
[13:13] <alecu> dobey, on a case sensitive filesystem, you don't deal with it. On a case insensitive, if there's a file with a similar name, you rename it before storing on disk.
[13:13] <ralsina> dobey, alecu: OTOH, if Foo.jpg maps to "xxx.jpg" and then, with syncdaemon down, I delete it and create a xxx.jpg then I will end overwriting Foo.jpg on linux. Unless
[13:14] <ralsina> No, unlessnothing, it just won't work :-(
[13:14] <dobey> ralsina: see :)
[13:14] <ralsina> dobey: it is a much smaller corner case than we have now thugh
[13:14] <alecu> ralsina, it's already corner, right :-)
[13:14] <dobey> i think this is a very small corner case
[13:15] <ralsina> yeah, and partial solutions are probably not worth it
[13:15] <dobey> it really should only ever happen for people who intentionally break it
[13:15] <dobey> we should either document "be mindful of what the OSes you're syncing between support" or just not support it on anything
[13:16] <ralsina> only "real solution" is to break it on the linux side. Or at least show a warning.
[13:16] <ralsina> Like "folder blah contains files with names too similar for windows"
[13:16] <ralsina> which is a crappy message
[13:21] <alecu> We will surely hit this same bug on osx, where some filesystems are case insensitive.
[13:22] <alecu> so, the old bug was... https://bugs.launchpad.net/ubuntuone-client/+bug/368626/comments/7
[13:22] <alecu> and it was about utf-8, not about case sensitivity.
[13:22] <alecu> but it's got some common elements.
[13:24] <alecu> ralsina, we should at least fix the data corruption. If we decide to allow this on linux, we should make it so the windows client does not delete the similarly named files it will decide not to sync. And show some alert about those unsyncd files on the windows side too.
[13:24] <ralsina> alecu: we would have to fail the syncing and stop syncdaemon, because we will never be able to reach the same generation on that volume
[13:25] <ralsina> alecu: so, we would have to do things like unsubscribig windows UDFs where this happens, and if it's ~/Ubuntu One, stop syncing at all
[13:25] <alecu> ralsina, too drastic :-)
[13:25] <ralsina> alecu: we could have it on linux, too, since you *could* have an insensitive home :-)
[13:26] <alecu> only an insensitive clod would do that :-)
[13:26] <ralsina> alecu: I suspect syncdaemon has no way to say "this volume is synced except for this file, which is not there locally"
[13:26] <__lucio__> ralsina, broken nodes
[13:26] <ralsina> __lucio__: we can mark a node as broken? Ok then
[13:27] <__lucio__> ralsina, "mark"...
[13:27] <__lucio__> maybe... log?
[13:27] <ralsina> __lucio__: he
[13:27] <ralsina> __lucio__: how is it going to be any less confusing, though? Less data loss, is a win, though.
[13:27] <__lucio__> ralsina, nothing a rescan from scratch would not fix
[13:58] <ralsina> rebooting because of windows wretchedness. Will probably take about 20 minutes because it has to reconfigure updates. Will be back when it's done.
[14:06]  * mandel back
[14:06] <mandel> alecu, ralsina, it looks like proxy support is broken on O, if we have a proxy creds error the second request (the one with the correct creds) does not fire the finish signal..
[14:06] <mandel> so, bloody PUTA
[14:07] <alecu> mandel, O or P?
[14:07] <mandel> alecu, O
[14:07] <alecu> mandel, O with nightlies?
[14:07] <mandel> alecu, bingo
[14:09] <mandel> alecu, ralsina, the  good news is that this: https://code.launchpad.net/~mandel/ubuntu-sso-client/use-unix-domain/+merge/104244 is sso working on mac
[14:09] <alecu> mandel, weird. What about this same thing with P?
[14:10] <mandel> alecu, with P it works, but we use diff versions of pyqt
[14:10] <alecu> mandel, yes, gatox told us about that branch. And he took credit for it too :-)
[14:10] <gatox> lie!
[14:10] <alecu> mandel, oh, right. PyQt again.
[14:10] <mandel> alecu, this is bug 990048
[14:10] <alecu> mandel, awesome.
[14:11] <mandel> alecu, yes.. I have played with the idea of doing: reply.finished.connect(lambda: self._handle_finish(reply)) after reading this: http://www.johanpaul.com/blog/2011/07/why-qnetworkaccessmanager-should-not-have-the-finishedqnetworkreply-signal/
[14:11] <mandel> alecu, but it does not fix it.. I'm tempted to try an implementation with a nam per request and see what happens
[14:12] <mandel> alecu, which would also as a side effect fix the issue in which we use always the same proxy no matter the request (http/https)
[14:14] <gatox> mandel, have you been looking at filesystem events in mac?? it seems i can not detect when a file is opened
[14:14] <mandel> gatox, a little but not in details yet
[14:15] <alecu> mandel, "an implementation with a nam per request" <- that sounds like a nice thing to try in our case.
[14:15] <mandel> gatox, we cannot on windows so don't worry, the event nanny won't work and no more problems
[14:15] <gatox> mandel, ahhh okok
[14:15] <mandel> alecu, I'll try that path and will see what happens
[14:15] <mandel> gatox, if you want more details of why I can tell you :)
[14:16] <gatox> mandel, jejeje no need
[14:16] <mandel> alecu, ralsina, also, people are stupid: https://bugs.launchpad.net/ubuntuone-client/+bug/993111
[14:16] <alecu> gatox: regarding fs events, you should read this article: http://arstechnica.com/apple/reviews/2007/10/mac-os-x-10-5.ars/7
[14:16] <gatox> alecu, ok...... reading
[14:25] <ralsina> mandel: I would not say the reporter is stupid.
[14:26] <mandel> ralsina, the reporter is rye :)
[14:26] <mandel> ralsina, but I feel that people should not be naming files AB, Ab, aB, ab
[14:26] <ralsina> mandel: which is why I would not call the reporter stupid :-)
[14:27] <mandel> ralsina, because is a stupid thing to do, I do agree that we have to solve it in a better way
[14:27] <ralsina> mandel: sometimes it's not the user's fault, it's just something he got
[14:29] <gatox> dobey, can you re-review this: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/refactor-notification/+merge/104164 ?
[14:30] <dobey> gatox: soon as the diff updates :)
[14:36] <ralsina> anyone need any reviews before I start doing a build and curse in mumbles?
[14:39] <dobey> ralsina: briancurtin was ill yesterday. not sure where we're at for windows fixes right now
[14:39] <ralsina> dobey: ack
[14:40] <ralsina> briancurtin: are you here now?
[14:48] <briancurtin> ralsina: yeah i'm here, i was out the first half of yesterday (i'll add it on canonicaladmin)
[14:50] <ralsina> briancurtin: please
[14:50] <ralsina> briancurtin: hope nothing serious!
[14:50] <ralsina> briancurtin: also, any progress on SSL issues ?
[14:50] <ralsina> also, I apparently have a 14 seconds lag
[14:51] <ralsina> @ping
[14:51] <dobey> lag lag
[14:52] <briancurtin> ralsina: i got stuck on making the bitrock installer work - it now works flawlessly on upgrade installs (stops U1, runs uninstall, does new install, all without reboot), but is broken on clean installs (doesn't create uninstaller, doesn't create menu shortcuts...). im switching to SSL now. i'll try to figure out how to debug from alecu's email
[14:52] <dobey> ralsina: we got a report about ssl issue on linux
[14:52] <dobey> ralsina: bug #992708
[14:52] <ralsina> dobey: looking...
[14:52] <dobey> and it's in sso
[14:52] <dobey> fun! :)
[14:53] <ralsina> dobey: sporadic SSL errors on SSO usually are dueto performance issues
[14:53] <dobey> yep
[14:53] <ralsina> Also, the throttling couldn't be good for that. Maybe ask if it's still happening?
[14:56] <ralsina> There I asked.
[14:56] <thisfred> standup in 5!
[14:56] <thisfred> 4
[14:57] <dobey> 3
[14:58] <gatox> 2
[14:58] <thisfred> nope
[15:00] <gatox> me
[15:00] <briancurtin> me
[15:00] <mmcc> me
[15:00] <thisfred> me
[15:00] <mandel> me
[15:00] <gatox> ralsina, alecu dobey ?
[15:00] <alecu> me
[15:01] <dobey> me
[15:01] <ralsina> me
[15:01] <gatox> DONE:
[15:01] <gatox> Reviews, fix some branches, start working on filesystem notifications for MAC OS.
[15:01] <gatox> TODO:
[15:01] <gatox> Keep working in filesystem notifications.
[15:01] <gatox> BLOCKED:
[15:01] <gatox> No
[15:01] <gatox> briancurtin, go
[15:01] <briancurtin> DONE: had to take more medicine and rest through the morning. debugged some of the bitrock issues, but its still broken. 100x better than before, but broken :(
[15:01] <briancurtin> TODO: debug SSL issues on the older Qt
[15:01] <briancurtin> BLOCKED: none
[15:01] <briancurtin> NEXT: mmcc
[15:01] <mmcc> done: almost done with refactoring merge, learned more about workflow
[15:01] <mmcc> & tools - thx gatox ;P
[15:01] <mmcc> TODO: get reviews for branches, finish up last branch, figure out
[15:01] <mmcc> what's next, get windows tests working
[15:01] <mmcc> blocked: nonenext
[15:01] <mmcc> next thisfred
[15:01] <thisfred> DONE: Bug 987412, major refactoring of index expression parsing in C TODO: Boolean index mapping, Continue with u1todo app BLOCKED: no MANDEL: mandel
[15:01] <mandel> DONE: fixed bug 990048 by using a QNetworkAccessManager per request to work around issues in older versions of Qt. Read about fsevents
[15:01] <mandel> TODO: Land the sso branch for mac (is working) after fixing failing tests on some windows machines.
[15:01] <mandel> BLOCKED: no
[15:01] <mandel> alecu, please
[15:02] <alecu> DONE: misc national holidays. Managed to sprain sciatic nerve, had to take a shot, now on pills. Got appointment for a chinese doctor.
[15:02] <alecu> TODO: I'm taking half day off, starting now.
[15:02] <alecu> BLOCKED: kinda
[15:02] <alecu> NEXT: dobey
[15:02] <dobey> λ DONE: more backports work
[15:02] <dobey> λ TODO: more backports work, SRUs, u1db packaging/buildsys
[15:02] <dobey> λ BLCK: none.
[15:02] <dobey> ralsina
[15:02] <ralsina> DONE: implemented a website! looong weekend. Catchup, switching to windows (which is broken) TODO: get a grip on SSL bug and kill it BLOCKED: no NEXT dobey but he's early
[15:03]  * alecu saw dobey's "me" before ralsina's
[15:03] <thisfred> me too
[15:04] <mmcc> me three. ralsina has latency issues today?
[15:04] <ralsina> my lag acting up I guess
[15:04] <ralsina> I do have them. I blame windows.
[15:04] <ralsina> Also, my cat scratched me, and I blame windows.
[15:04] <alecu> ralsina, cheer up. A lot better to have latency issues than latent issues.
[15:05] <mandel> or talent issues..
[15:05] <mmcc> alecu, do you know of a doctor for lag?
[15:05] <mmcc> also, hope you feel better soon alecu
[15:06] <mandel> ralsina, alecu, can I have a review for https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-oneric-tests/+merge/104394
[15:06] <dr_lag> take a benzedrine and call me in the morning
[15:07]  * mmcc found another dr lag: http://static.usenix.org/event/hotos11/tech/slides/rumble.pdf
[15:07] <mandel> dobey, this fixes the O issues: https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-oneric-tests/+merge/104394
[15:12] <alecu> mmcc, I know of some chinese doctors that lay fiber to your house. that usually fixes latency issues. But it sounds expensive...
[15:14] <alecu> mandel, print 'Result d %s' % d
[15:14] <mandel> alecu, ups, leftover from debugging, sorry
[15:14] <alecu> mandel, I won't be doing a full review, but make sure that your reviewers do IRL tests for this on both linux and windows.
[15:15] <alecu> mandel, and that in one of those cases they use an authenticated proxy.
[15:16] <mandel> alecu, yes, it is fundamental since there is a big change in the inner workings..
[15:17] <alecu> mandel, also: make some wall clock tests to see that the time it takes the control panel to open and get all the info is not severely affected by using multiple nams
[15:18] <alecu> mandel, and another thing that could break: the proxy password dialog may open too many times in this case...
[15:18] <alecu> mandel, (multiple nams means that control panel api calls cannot reuse a single http connection)
[15:20] <alecu> mandel, other than that, the code looks good, and it's a very good idea to try this as it will likely fix the issue with cached proxy credentials.
[15:20]  * gatox lunch
[15:21] <mandel> alecu, ok, I'll jot that down in the description so that the reviewers know about this things
[15:30] <briancurtin> is there no way to enter a half sick day in canonicaladmin? or should i just do half holiday? (i have no idea)
[15:31] <dobey> briancurtin: you can enter number of days as 0.5 no?
[15:32] <briancurtin> dobey: ha, i guess i could do that. the calculate button doesn't calculate it for that, so i figured it meant it wasn't applicable
[15:32] <dobey> briancurtin: it still shows the "Half Day" dropdown for me if i choose "Sick"
[15:33] <dobey> briancurtin: so i guess pick "Half Day: AM" ?
[15:33] <briancurtin> yeah that's what i have, and entering 0.5 and submitting tells me i have to enter a value for number of days. form validation is off.
[15:33] <dobey> briancurtin: if you do that, then click calculate, it should say 0.5 :)
[15:34] <dobey> briancurtin: ah, just have it be 1.0 then
[15:34] <dobey> briancurtin: i don't think it matters that much
[15:35] <briancurtin> ah, the problem was that i didnt choose an end date...i thought it would figure that out from it being a half day. it worked with 0.5
[15:38] <dobey> ah
[15:42] <mandel> dobey, one question, when we try to use nightlies in an older version like O, what version do we use of PyQt?
[15:43] <mandel> dobey, the bug seems to be related to that and I don't want to screw up or code to o around old issues
[15:43] <dobey> mandel: it's using the pyqt that's in oneiric
[15:44] <dobey> mandel: the build log shows what specific version of pyqt is being installed
[15:44] <mandel> dobey, I saw the version and is a PITA.. so I guess i'll have to do funky code for this to work :(
[15:45] <dobey> mandel: if it's a bug in qt or pyqt, we can discuss how to get a fix shipped for it
[15:45] <mandel> dobey, I'll try to get the latests pyqt on my O machine, if that works, it means that it is pyqt
[15:46] <dobey> mandel: well, it doesn't mean it's pyqt necessarily
[15:46] <mandel> dobey, is probably Qt more than pyqt
[15:46] <dobey> mandel: it means changing pyqt changed something that made it work. pyqt could be different enough to affect the timing, if it's a race issue for example
[15:47] <mandel> dobey, the problem that we face is that in the version that we find in O we cannot make two requests with the same QNetworkAccessManager because the second one wont fire the finished signal
[15:50] <dobey> mandel: likely there is a change for that specific issue in either qt or pyqt, which we can probably get into an SRU, then
[15:50] <dobey> mandel: it would be good to find that, instead of working around it in our code
[15:50] <dobey> mandel: because the same problem probably exists in 11.04 and 10.04 as well
[15:50] <mandel> dobey, I think so too, I'll look close to it and see if getting the latest pyqt fixes the issue
[15:51] <ralsina> mandel: i it's in PyQt, they are very responsive. We had a bug open for years that they fixed in 24 hours after reporting
[15:51] <dobey> mandel: ok, thanks. maybe change your branch to "work in progress" until more debugging is done then?
[15:51] <mandel> dobey, yes, well do!
[15:52] <dobey> hrmm
[15:54] <dobey> ok, need to get lunch. bbiab
[16:24] <joshuahoover> ralsina: so, i may have an ubuntu user or 2 who are reporting similar symptoms as the ones we were seeing in windows 3.0.0...waiting for logs...i noticed that these users are languages other than english, maybe that's a clue? not sure
[16:24] <ralsina> joshuahoover: I don't see how
[16:25] <joshuahoover> ralsina: any update on 3.0.0? i'm afraid we're going to lose interest from the users we've had testing stuff for us
[16:25] <ralsina> joshuahoover: I am about to do a build now
[16:25] <joshuahoover> ah, cool
[16:26] <ralsina> joshuahoover: hopefully I have the exact same building env we had for 2.0.3
[16:26] <briancurtin> joshuahoover: i'm trying to figure out why it's not starting up with the older Qt, but that won't figure out the real problem as far as i know
[16:43] <Fury1306> Hi!
[16:44] <Fury1306> It might sound a bit harsh, but there's one thing I really don't like.
[16:44] <Fury1306> Contact Sync isn't working since August 2011 (iirc).
[16:44] <Fury1306> The wikipage says "Syncing of contacts is not working for some users due to server issues. On 2011-10-10 we released an update to all clients to help ease the load on the server and improve reliability of syncing. "
[16:44] <Fury1306> I managed to get contact-sync working in 11.10, but in 12.04 there is not even a menu-entry in the U1 GUI to start.
[16:44] <Fury1306> Are you ever going to fix contact sync?
[16:44] <Fury1306> If not, will there be a webapplication so I can save my contacts from the website and restore them in Thunderbird or Evolution?
[16:59] <Chipaca> Fury1306: you can be harsher than that, if you want
[17:03] <Chipaca> Fury1306: can i give you a way to get all your contacts, but i'm not sure if it'll be enough for what you need
[17:03] <Chipaca> Fury1306: contacts service will come back over the next six months approximately, but i don't have a better date than that
[17:04] <mandel> dobey, using the latests pyqt fixes the issues in O
[17:05] <Chipaca> gah, i've got to run
[17:05] <Fury1306> That's fine. I do know where to look for my contacts on the website. And I think I can deal with it for some month. I just wanted to know if there is any hope.
[17:05] <Chipaca> Fury1306: i'll be back in a few hours
[17:06] <dobey> mandel: ok, if you can isolate the patch that fixed it, we can get it into an SRU probably
[17:06] <mandel> dobey, ok, I'll do my best to see the diffs etc..
[17:07] <Fury1306> I think it would be a good idea to make it more publik, that contact-sync comes back. I don't think I'm the only person missing it.
[17:07] <Chipaca> Fury1306: 1. get your couch root by doing: u1oauthrequest https://one.ubuntu.com/api/account/
[17:08] <Chipaca> Fury1306: 2. get all your couch contacts by doing, e.g.: u1oauthrequest --oauth-signature-method=PLAINTEXT https://couchdb.one.ubuntu.com/u/651/2bd/11/contacts/_all_docs?include_docs=true > contacts.out
[17:08] <Chipaca> now yes, i must run
[17:08] <Chipaca> Fury1306: announcements to happen after planning. Soon
[17:09] <Fury1306> thanks
[17:11] <mandel> ok, EOD for me catch you all tom!
[17:11] <ralsina> bye mandel!
[17:12] <daker> hi
[17:13] <daker> can anyone tell me why i am getting http://pastebin.com/D1CjWA9f using this code http://pastebin.com/EFz3Tj4R
[17:13] <daker> ?
[17:19] <daker> and i don't know why i am getting 'status': '401'
[17:25] <daker> S.O.S
[17:27] <dobey> if you're getting a 401, it means the auth was not valid
[17:27] <dobey> and probably means it didn't return any json
[17:35] <daker> dobey, ok
[17:37] <dobey> thisfred: ^^ maybe you can give some better advice to daker ?
[17:38]  * thisfred looks
[17:40] <thisfred> daker: why do you do answer[1]? w
[17:40] <thisfred> what does auth.request return?
[17:40] <thisfred> a tuple?
[17:41] <thisfred> webm0nk3y might know what is wrong with the authentication part
[17:41] <daker> thisfred, it was taken from a snippet
[17:42] <daker> thisfred, what i want is to login with my email/pwd
[17:42] <daker> and get a list the list of files i have on my account
[17:44] <thisfred> right. I see nothing obviously wrong, but I must admit I don't really know how the CredentialManagement code works. Where did you find the snippet?
[17:44] <thisfred> daker: so does it give you a pop up when you run it?
[17:45] <daker> thisfred, i think from duplicity
[17:45] <daker> thisfred, no
[17:49] <thisfred> daker: when I run it, it works, so there are two possibilities: 1. you ran into a temporary glitch with the system 2. somehow your system doesn't yet have the right credentials stored for u1db
[17:49] <thisfred> by 1. I mean maybe the server timed out
[17:49] <daker> #2 is possible
[17:50] <thisfred> daker: if you start the ubuntuone control panel
[17:50] <thisfred> it will tell you whether the machine is registered, and allow you to register it if not
[17:51] <thisfred> ignore u1db, I meant u1, u1db is something unrelated I'm working on, so it was a freudian slip
[17:51] <thisfred> daker, anyway, the code is correct
[17:52] <daker> thisfred, how to check if i have the right credentials ?
[17:52] <thisfred> though it could use better error reporting ;)
[17:52] <thisfred> daker: see above: just start the control panel
[17:55] <webm0nk3y> daker: is it a python snippet?
[17:55] <daker> yep
[17:55] <thisfred> webm0nk3y: sorry to disturb you, the code is correct
[17:56] <daker> thisfred, what do you think http://i.imgur.com/fMTi1.png ?
[17:57] <webm0nk3y> thisfred: k
[17:57] <thisfred> daker, my french is not great, but that looks like it's connected and fully synchronized
[17:57] <thisfred> daker, so the credentials must be there
[17:58] <daker> thisfred, yep
[17:58] <thisfred> daker: I would try to run it a few more times
[17:58] <thisfred> if it still does not work, I am not sure how to proceed
[17:59] <thisfred> maybe instead of taking answer[1] doing print repr(answer) so we can see what you're getting back
[17:59] <daker> thisfred,
[17:59] <daker> status': '401', 'content-length': '0', 'via': '1.1 one.ubuntu.com', '-content-encoding': 'gzip', 'transfer-encoding': 'chunked', 'vary': 'Cookie,Accept-Encoding', 'server': 'PasteWSGIServer/0.5 Python/2.6.5', 'etag': '"d41d8cd98f00b204e9800998ecf8427e"', 'x-bzr-revision-number': '5411', 'date': 'Wed, 02 May 2012 17:59:36 GMT', 'content-type': 'text/html; charset=utf-8'}, '')
[18:00] <thisfred> daker, right, so a 401 with empty content
[18:04] <dobey> oh gtk
[18:05] <thisfred> dobey: I'm all out of ideas here. Does running the script work for you?
[18:05] <dobey> i haven't tried it
[18:09] <thisfred> It does here, so other than server outage, or incorrect credentials, I can't think of a cause
[18:10] <daker> anyway thisfred if you find something that would be, if not just forgot about it
[18:10] <thisfred> daker, if you add a new file to your shared folder, does it in fact get synchronized ?
[18:10] <daker> lemme see
[18:10] <dobey> script works fine here
[18:12] <daker> thisfred, it does
[18:12] <thisfred> very strange
[18:12] <thisfred> so the credentials are there, but somehow the script doesn't get them correctly
[18:13] <daker> FYI 11.04
[18:13] <thisfred> daker, could be that that's the issue, I don't know if the script requires a newer version
[18:13] <dobey> oh
[18:14] <thisfred> dobey: if so, would nightlies on 11.04 fix it?
[18:14] <dobey> no
[18:14] <dobey> nightlies are disabled on 11.04 right now
[18:17] <daker> thisfred, just forgot about it
[18:17] <thisfred> daker, sorry I couldn;t be of more help
[18:17] <daker> no worris
[18:18] <daker> worries*
[18:18] <thisfred> daker, if you upgrade at some point it will probably just start working, is the good news ;)
[18:18] <daker> upgrade = nightmare
[18:18] <dobey> well 11.04 will be EOL in 6 months
[18:19] <daker> ツ
[19:06] <balloons> how can I move a file or folder inside of ubuntu one?
[19:07] <balloons> I added a new client, and I want to be able to choose which folders to sync. However, since everything is in my "ubuntu" one folder, I can't not sync all of my files to this client
[19:08] <dobey> balloons: move stuff out of the Ubuntu One folder on the other client where they exist, to other folders, and set those folders to be synchronized
[19:11] <balloons> thanks dobey.. a secondary question -- is there a way to turn off syncing the ubuntu-one folder?
[19:12] <dobey> no
[19:12] <balloons> I am on a machine with limited resources.. and currently I'd like to sync only one folder. However, all of my stuff is in ubuntu-one
[19:37] <ralsina> balloons: sorry, there is no way to not sync ~/Ubuntu One
[19:37] <ralsina> balloons: however, if you move all your thing into UDFs it should be pretty quick!
[19:59] <mmcc_lunch> gatox, are you around still?
[19:59] <gatox> mmcc, yes
[20:00] <mmcc> This line broke my tests when I merged with trunk just now - http://bazaar.launchpad.net/~diegosarmentero/ubuntuone-client/refactor-messaging/view/head:/tests/platform/messaging/test_linux.py#L115
[20:01] <mmcc> I really don't know why the autopilot didn't catch it before it merged it
[20:01] <mmcc> gatox, it has the wrong module name in a mocker.replace call
[20:02] <mmcc> gatox, s/wrong/old/
[20:03] <gatox> mmcc, can you show me the trace?
[20:03] <mmcc> sure, let me check my scrollback
[20:03] <gatox> mmcc, ahhhh yes.. i see it
[20:03] <gatox> mmcc, the weird thing is that it works here :S
[20:04] <gatox> mmcc, you were letting me know that you are fixing that in your branch..... or should i propose a branch for that?
[20:04] <mmcc> gatox, that is really weird. and apparently it works on the autopilot test too...
[20:05] <mmcc> gatox, whatever's the right way to do it… it'd be fast to fix in my branch but it kind of mixes up the purpose of the branch…
[20:06] <ralsina> do separate branches, let's try to keep merges clean
[20:06] <ralsina> unless it's a LOT of work
[20:06] <gatox> ralsina, ok, i'll create a branch for this
[20:07] <mmcc> gatox, for reference see http;//paste.ubuntu.com/963105
[20:08] <mmcc> so what's going on with the different test environments? the auto test (what's its name? ) should have caught this, right?
[20:08] <mmcc> gatox or http://paste.ubuntu.com/963105 even
[20:09] <gatox> yes
[20:17] <gatox> mmcc, ralsina if you can, please review this little branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/fix-messaging-tests/+merge/104444
[20:17] <ralsina> gatox: sure
[20:17] <mmcc> gatox: ok
[20:17] <gatox> ralsina, thx
[20:19] <mmcc> ralsina, gatox, any idea why the auto tests didn't catch that?
[20:19] <gatox> mmcc, nop
[20:19] <mmcc> gatox, approve - I made the same change in my branch and it passed the tests a minute ago
[20:19] <gatox> mmcc, great
[20:20] <ralsina> mmcc: no idea what "that" is. Should read backlog :-)
[20:20] <mmcc> ralsina, that = the bug that gatox's branch fixes
[20:21] <gatox> well...... eod for me.... too much fsevents in mac for me today :P
[20:21] <mmcc> it caused failing tests for me (and is wrong to look at) but passed the auto tests and works on gatox's system
[20:21] <ralsina> mmcc: don't know. Looks like it should hve
[20:21] <gatox> see you tomorrow people!
[20:21] <mmcc> bye gatox
[20:22] <ralsina> mmcc: well, the tests are only running on linux before merging. What OS was it failing on?
[20:22] <gatox> ralsina, linux
[20:22] <mmcc> ralsina: linux.
[20:22] <ralsina> no idea then
[20:22] <mmcc> in fact, linux is the only place I can run tests successfully right now
[20:23] <ralsina> mmcc that's the hole point of the refactor I guess ;-)
[20:23] <ralsina> whole*
[20:23] <mmcc> I have a todo to fix my windows setup, right now it passes many but not all of the tests
[20:23] <mmcc> ralsina: yep, true that
[20:23] <gatox> mmcc, did you try with the wiki instead of the buildout?
[20:24] <gatox> ok..... i'm really leaving now :P byeeeee
[20:24] <mmcc> gatox (for the record): I didn't try to start from scratch with the wiki instructions. the buildout does copy the specific files you said I needed, and
[20:24] <mmcc> gatox: for the record - I didn't try to start from scratch with the wiki instructions. the buildout does copy the specific files you said I needed, and
[20:25]  * mmcc has irc client problems
[20:25] <mmcc> gatox: for the record,  I didn't try to start from scratch with the wiki instructions. the buildout does copy the specific files you said I needed, and
[20:25] <ralsina> mmcc: hehe, lost it by this much
[20:25] <mmcc> oh well, I give up
[20:26] <ralsina> I had forgotten how fracking slow bzr is on windows
[20:26] <ralsina> it's *scary* slow
[20:27] <ralsina> dobey: bug #696568 is getting weird
[20:27] <ralsina> dobey: it involves LTSP, NM and probably small dwarves
[20:28] <mmcc> launchpad question - if I have a merge request pending and trunk moves on, does that request get re-checked for conflicts, or do I need to update my branch and redo the merge request if it's sufficiently old?
[20:28] <ralsina> mmcc: the proposal should show the conflicts
[20:29] <ralsina> mmcc: but you should merge trunk into your branch every now and then to keep things simple
[20:29] <ralsina> mmcc: idea about why it did not fail: did you run the tests on gatx branch, or you branched trunk and merged that branch?
[20:29] <mmcc> ralsina: so the proposal will show conflicts if they're introduced on trunk after the proposal was made?
[20:30] <dobey> ralsina: and lies
[20:30] <mmcc> ralsina: and yes to updating often. good hygiene, don'cha know
[20:30] <ralsina> mmcc: IIRC yes
[20:30] <mmcc> ralsina: ok cool
[20:30] <ralsina> dobey: trick is, maybe it's getting a dbus session from the client and the proces is running on the LTSP server?
[20:30] <mmcc> ralsina: I branched trunk and merged it, then made my own changes, then ran tests.
[20:31] <dobey> ralsina: no
[20:31] <ralsina> mmcc: ok, maybe it worked for gatox because he didn't do that :-)
[20:31] <ralsina> dobey: or viceversa
[20:31] <mmcc> ralsina: not sure I understand… my changes were unrelated.
[20:32] <ralsina> mmcc: ok, I don't know what the branches involved were, so I am making stuff up. It's kinda late :-)
[20:32] <dobey> ralsina: dbus doesn't work that way
[20:33] <mmcc> ralsina, no prob. I'm reasonably confident that the code is now right, so maybe we don't have to figure this out right now, but I'm definitely curious what happened
[20:34] <mmcc> if only so I'm sure I understand the tools
[20:34] <ralsina> dobey: is there a way for mmcc to see the tarmac test run logs?
[20:34] <dobey> how the heck does my memory usage go from 2.0GB to 7.9GB, when there is no single new process using 6.0GB
[20:34] <dobey> ralsina: for successful merges?
[20:35] <ralsina> dobey: yes
[20:35] <ralsina> dobey: we have a successful merge that seems like it should have failed
[20:36] <dobey> yeah i gave him the url last night for the http interface to browse the logs
[20:36] <dobey> url for the merge?
[20:36] <ralsina> dobey: ack
[20:36] <ralsina> mmcc: ^
[20:36] <dobey> although my system is about to hit swap thanks to firefox :-/
[20:37] <mmcc> dobey, the merge we're talking about is https://code.launchpad.net/~diegosarmentero/ubuntuone-client/refactor-messaging/+merge/104156
[20:37] <dobey> why should it have failed?
[20:38] <mmcc> dobey, this line: http://bazaar.launchpad.net/~diegosarmentero/ubuntuone-client/refactor-messaging/view/head:/tests/platform/messaging/test_linux.py#L115
[20:38] <mmcc> dobey causes this backtrace on the linux tests: http://paste.ubuntu.com/963105
[20:38] <ralsina> mmcc: the tests for gatox's latest branch pass on windows for me
[20:38] <mmcc> ralsina: it's a linux-specific test
[20:39] <ralsina> mmcc: yes, but you mentioned you could not run tests outside linux
[20:39] <dobey> huh
[20:39] <mmcc> ralsina: oh yeah, I think that's because my setup is still not set… up
[20:39] <ralsina> mmcc: oh :-)
[20:40] <dobey> hmm
[20:41] <dobey> well that's weird
[20:41] <ralsina> dobey: I see a sort of theme for the day
[20:41] <mmcc> hey, so I approved gatox's fix for this, since I made the same change and it fixed it for me. do we need a second review? I'm OK waiting until we have some clue for why it wasn't failing elsewhere…
[20:42] <mmcc> nevermind
[20:42] <dobey> oh
[20:42] <dobey> Makefile.am is broken
[20:43] <mmcc> dobey, do tell. I love makefile/automake bugs!
[20:43] <dobey> http://paste.ubuntu.com/963200/ is what you are seeing?
[20:43] <mmcc> dobey: yes, that's right
[20:44] <mmcc> dobey, that's from the tarmac logs?
[20:45] <dobey> no, that specifically is from the nightlies build that just finished successfully
[20:45] <dobey> but same thing would have happened in tarmac
[20:46] <dobey> the problem is that when the additional run of u1trial was added for the proxy tests, which use the qt reactor, it created a situation where failures in the normal set of tests would get ignored, and make would continue on
[20:47] <ralsina> oh, un
[20:47] <ralsina> fun*
[20:49] <mmcc> hooray make! 2012 and still causing pain
[20:50] <mmcc> hrm, so you're looking at Makefile.am:57 and 58?
[20:53] <dobey> maybe
[20:53] <dobey> yes
[20:54] <dobey> they need an "|| exit 1" or something
[20:56] <mmcc> dobey: yes, that would work. a quickie test makefile confirmed it
[20:59] <mmcc> dobey or just && instead of || exit 1. no big diff. who wants to make this change?
[21:00] <dobey> well i guess we need a bug for it
[21:01] <mmcc> dobey: if you're busy I can do it
[21:01] <dobey> i'll do it
[21:02] <mmcc> dobey: ok
[21:04] <ralsina> EOD for me.
[21:04] <ralsina> See you all tomorrow morning
[21:04] <mmcc> bye ralsina
[21:04] <briancurtin> see ya
[21:05] <ralsina> briancurtin: oh here you are, we made some progress on the SSL thing, am not sure you saw it
[21:05] <briancurtin> didn't see, i only got back on about 10 minutes ago
[21:05] <ralsina> briancurtin: basically: your 3.0.0 installer "fixed" it for a user, and we got an ubuntu user to try the extra SSL logging , and he has a totally bogus certificate
[21:06] <briancurtin> weird
[21:06] <ralsina> briancurtin: so, basically we have a mistery certificate (signed by "valicert")
[21:07] <ralsina> briancurtin: which is an old godaddy CA, and our certs are from godaddy
[21:07] <ralsina> briancurtin: so, may be bogus router/proxy in the middle, bogus validation, or bogus certificate in some bizarre place in our servers that we never see
[21:07] <ralsina> in increasing order of impossibility
[21:08] <briancurtin> interesting. is there a plan of attack?
[21:08] <ralsina> briancurtin: I asked for a little extra diagnostic: going with firefox to the URL and getting the cert details
[21:08] <dobey> ralsina: valicert isn't an old ca
[21:08] <ralsina> briancurtin: but we are rolling your release to the affected users for extra info
[21:09] <dobey> ralsina: godaddy has a bunch of different CAs, because they are a shady bunch of scheisters
[21:09] <ralsina> dobey: valicert.com doesn't even exist anymore
[21:09] <ralsina> dobey: godaddy has not sold valicert certificates in years AFAICS
[21:09] <dobey> what do you mean doesn't exist?
[21:10] <ralsina> dobey: http://www.valicert.com/ is now axway
[21:10] <ralsina> "Axway is the Business Interaction Networks company"
[21:10] <dobey> yes
[21:11] <ralsina> and talks about "monetizing cloud data transfer infrstructure"
[21:11] <ralsina> dobey: and https://www.valicert.com gives a certificate warning ;-)
[21:12] <dobey> ralsina: that doesn't mean godaddy doesn't still issue certificates with that name
[21:12] <dobey> ralsina: see bit about shady scheisters
[21:13] <briancurtin> a business interaction networks company that monetizes cloud data transfer infrastructure...sounds like a joke i once told
[21:13] <ralsina> dobey: that certificate is expired, invalid and for a different domain :-)
[21:13] <dobey> which one is?
[21:13] <ralsina> dobey: the one at https://www.valicert.com
[21:13] <ralsina> dobey: issued for localhost.localdomain, too
[21:13] <dobey> it's a self-signed cert, yes
[21:14] <dobey> but that has nothing to do with godaddy being a shady company
[21:14] <ralsina> dobey: tell me it's not funny in a "oh the world sucks" kinda way
[21:16] <ralsina> briancurtin: the quote is "what can our cloud-based community management solution do for your bottom line?"
[21:16]  * ralsina has to write snarky blog post. Maybe later.
[21:26] <ralsina> ok, EOD for real now, have to play ben10 power splash (don't try it at home)
[21:26] <dobey> i need to go too
[21:27] <dobey> meh
[21:30] <mmcc> bye dobey, ralsina
[21:56] <dobey> mmcc: https://code.launchpad.net/~dobey/ubuntuone-client/fix-test-fails/+merge/104463
[21:58] <mmcc> dobey: that looks good to me, are you asking for a review?
[22:03] <dobey> mmcc: yes
[22:04] <mmcc> dobey: approved.
[22:05] <dobey> mmcc: cool. diego's other fix is already landed, so this should land in about 10 minutes
[22:05] <dobey> and with that, i am really gone :)
[22:05] <dobey> later
[22:06] <mmcc> dobey, thanks - bye
[22:16]  * mmcc is leaving now