[01:25] <ralsina> AND THERE YOU ARE, STUPID TEST, YOU WILL PASS FROM NOW ON. And I go watch TV.
[09:03] <zyga> hi
[09:04] <zyga> is https://bugs.launchpad.net/ubuntu/precise/+source/ubuntuone-client/+bug/978903 still unfixed in precise?
[09:04] <zyga> I'm somewhat confused by the status
[10:53] <mandel> zyga, triaged means that we have assigned someone to take a look at it, confirm it and fix it
[10:54] <mandel> zyga, we have a bug queue and limited resources so we go as fast as we can trying to prioritize those bugs that have been reported by a  bigger amount of people, but since it is High I would not be surprise if we get fixed asap
[10:56] <zyga> mandel, it's 'fix released'
[10:56] <zyga> mandel, not triaged
[10:56] <zyga> mandel, that's why I'm asking, since it's not in precise-*
[10:57] <mandel> zyga, what bug 978903?
[10:57] <zyga> yes, look at the url
[10:57] <ralsina> zyga: we have a pending SRU on precise
[10:58] <ralsina> zyga: it's being reviewed. getting updates into releases is not a quick process.
[10:58] <mandel> zyga, look at the state [High, Triaged], so is not released in P
[10:58] <mandel> ralsina, morning ;)
[10:58] <ralsina> zyga: so, while we have fixed it, it's not in any precise packages yet.
[10:58] <zyga> mandel, right, I was trying to understand what that means exactly as the launchpad janitor marked it as fix released due to a package being published somewhere
[10:58] <ralsina> and good morning mandel :-)
[10:59] <mandel> zyga, yeah, is a little confusing indeed
[10:59] <ralsina> zyga: it's released for our branches which means we have uploaded a tarball, not a package. It will be marked as released for precise once a precise .deb is published
[11:00] <zyga> thanks for the explanation
[11:05] <gatox> good morning
[11:05] <mandel> gatox, mornign!
[11:05] <gatox> mandel, how are you?
[11:05] <mandel> gatox, well, with the g in the right place :)
[11:06] <mandel> gatox, I've started with the python code (mainly adding tests code was there) for the fsevents daemon, I wonder, how is your side going?
[11:06] <mandel> gatox, shall we mumble and see how we start integrating this things?
[11:07] <gatox> mandel, i'm kind of blocked with some tests..... because it seems that we are not receiving the events that filesystem_notifications expects..... i was going to start to do some tests about those specific cases using only the macfsevents lib as a stand alone application to see if the problem is there or in the integration of the lib and u1-client-fsevents
[11:08] <mandel> gatox, ok,, shall we chat after my lunch (like in 1:30 mins) and we take a look at those issues?
[11:09] <mandel> gatox, I'd like to start integrating the daemon code early next week to have the complete solution and see if we can manually run sync-daemon :)
[11:12] <gatox> mandel, yes.... i want to have this running as soon as possible....... these issues are driving me crazy! :P
[11:13] <mandel> gatox, I'm been there when I worked on the windows parts hehe
[12:12] <alecu> ehlo!
[12:12] <gatox> alecu, hi
[12:14] <ralsina> ehlo alecu.localdomain!
[12:17] <alecu> ralsina, 250 Hello alecu.localdomain
[12:18] <ralsina> 351 AUTH required
[12:18] <ralsina> OMG, I am port 587
[12:18] <ralsina> aaaaaanyway.... good morning ;)
[12:19] <ralsina> mandel: I don't know how to make that problem with the tests not happen again. OTOH, when it happened, we caught it :-)
[12:20] <gatox> alecu, i'm seeing here that when we have some events related to the same file, we lose some events, for example a tests does: open-write-close..... and expects: CREATE, MODIFY..... we only get CREATE in mac..... to get both.... we need to open-close, wait, open-write-close
[12:21] <mandel> ralsina, yes, which is good, but seems fragile that our tests depend on the env, dont you think?
[12:21] <ralsina> mandel: one way to "fix" it would be to replace the env when we run, but I don't like it much
[12:22] <mandel> ralsina, certainly not..
[12:22] <alecu> gatox, you are getting "modify" events with the FSEvents API?
[12:23] <alecu> gatox, how?
[12:23] <gatox> alecu, yes, the macfsevents api detect: modify, attrib, create, delete, moved_from, moved_to
[12:24] <alecu> gatox, and how is it "calculating" the moved_from moved_to?
[12:24] <gatox> alecu, how what?
[12:24] <alecu> gatox, by checksums?
[12:24] <gatox> alecu, with the cookie attribute, in the doc says: The cookie field is a unique identifier linking together two related but separate events. It is used to link together an IN_MOVED_FROM and an IN_MOVED_TO event.
[12:26] <gatox> alecu, what i did (but it doesn't sound really good for me), was to skip the original test in mac, and add a special versoin of that test for mac, where we wait a sec between the open-close and open-write-close operations
[12:31] <alecu> gatox, hmmm
[12:31] <alecu> gatox, I was asking about the macfsevents lib
[12:31] <gatox> alecu, yes, that was my answer
[12:31] <alecu> gatox, how is that library calculating the moves?
[12:32] <gatox> http://pypi.python.org/pypi/MacFSEvents/0.2.6
[12:32] <mandel> ok, lunch time for me!
[12:32] <mandel> :)
[12:32] <gatox> using the cookie
[12:32]  * mandel lunch
[12:32] <alecu> gatox, that's how the library presents things to you.
[12:32] <alecu> gatox, I'm wondering how it's calculating it *inside* the library.
[12:33] <gatox> alecu, ahhhhhh don't know.... i can take a look at the c code..... give me a couple of mins
[12:33]  * alecu checks the sourcecode.
[12:34] <alecu> gatox, the thing is that the library has no info to know if a file was "deleted" or if it was "moved" outside of a watched folder.
[12:34] <gatox> alecu, nop
[12:34] <alecu> gatox, anyway, for our case it should be the same.
[12:36] <gatox> alecu, basically, my remainng problems are related to timing issues... where we want 2 or more specific events for the same file/folder, we need to wait between events
[12:37] <Captain_Proton> what would cause this? http://pastebin.com/JP8xrvFE am I missing a package?
[12:38] <Captain_Proton> fyi I am using u1sdtool
[12:39] <ralsina> Captain_Proton: looks like syncdaemon is not running
[12:39] <ralsina> Captain_Proton: but you should not see that error anyway
[12:40] <Captain_Proton> hmm I run u1sdtool --start that command run fine as in on errors
[12:41] <Captain_Proton> when I try to quit I get: sysadmin@5xdata-01:~$ u1sdtool --quit    ubuntuone-syncdaemon still running.
[12:41] <Captain_Proton> This is even after a reboot
[12:47] <ralsina> Captain_Proton: don't know, really
[12:47] <ralsina> Captain_Proton: have not seen it before
[12:47] <Captain_Proton> just my luck :)
[12:47] <Captain_Proton>  I will try to force reinstall ubuntuone see if that helps
[12:52] <gatox> alecu, in the .py group the moved_from, moveed_to
[12:57] <gatox> alecu, it tracks the file that are modify, deleted, etc in a dict that represents a map of the directory stat
[13:10] <dobey> meh. soft freeze days are different now
[13:27] <ceed^> Can't get U1 to work for me. U1 hangs on a synced folder containing a backup with 30,000 small files. The directory is updated every night with rsync. This works with Dropbox.
[13:28] <gatox> mandel, let me know when you are back from lunch..... maybe we can talk a little about fsevents
[13:29] <ralsina> ceed^: if you have thousands of small files, u1 will eat CPU and disk for several minutes before it starts moving
[13:30] <dobey> s/minutes/potentially a very long time/
[13:31] <ceed^> ralsina: I know that part, but when I get to my computer in the morning, the rsync backup happens during the night, U1 hangs and there are files not getting synced. I have top stop and start U1 again every morning. I have now tried for a week. Same thing happens.
[13:31] <dobey> ceed^: do you have a ~/.cache/ubuntuone/log/syncdaemon-exceptions.log file? and does it have many "KeyError" errors in it?
[13:31] <ralsina> ceed^: are you sycing the same folderusing dropbox?
[13:32] <ceed^> dobey: I will check. I think it has to do with the way rsync works which only updates changed and new files. Seems to confuse U1
[13:33] <ceed^> ralsina: Yes, but not at the same time.
[13:34] <ceed^> dobey: I have a large ~/.cache/ubuntuone/log/syncdaemon-exceptions.log file.
[13:34] <dobey> ceed^: it's more likely that syncing the same folder with dropbox and u1 is confusing things. or possibly you're just hitting the bug i'm asking about :)
[13:36] <ceed^> dobey: I have not had Dropbox installed for the last week. I did install Dropbox again this morning though. But the problem has been there while Dropbox wasn't even installed
[13:36] <dobey> ceed^: what is in the syncdaemon-exceptions.log file? lots of KeyErrors?
[13:40] <ceed^> dobey: Lots of this kind of errors: http://hastebin.com/kepolaxusa.vbs
[13:40]  * mandel back
[13:42] <ceed^> It's always the same backup folder causing the problem. I can see the files not synced using Magicicada.
[13:42] <dobey> oh
[13:42] <dobey> you are hitting the same bug that popey was hitting yesterday, it seems
[13:43] <dobey> i wonder what is causing that
[13:43] <ceed^> and why only that particular folder. The only unique thing with that folder is the it gets updated at the same time every night using rsync.
[13:44] <alecu> ceed^, is this on an up-to-date 12.04? Do you have the u1 nightlies ppa enabled?
[13:45] <ceed^> alecu, fully updated 12.04. No I do not have the PPA.
[13:45] <dobey> alecu: is this fixed already in nightlies or 3.0.1?
[13:46] <alecu> dobey, I'm not sure. I'm just trying to reproduce the issue in my vm.
[13:47] <dobey> ah ok
[13:50] <ceed^> Should I enable the PPA and see if that helps?
[13:51] <dobey> i need to get 3.0.1 into the stable ppa
[13:51] <dobey> but also need to rush and get 3.99.0 released/uploaded for quantal
[13:53] <ralsina> dobey: 3.0.1 is way more important
[13:54] <ceed^> or should I maybe try the beta ppa?
[13:55] <dobey> beta doesn't have newer than what's in precise right now either
[13:56] <ceed^> Okay, I'll run with nighties and see if it helps
[13:58] <dobey> well, add ppa:ubuntuone/stable and wait ~40 minutes or so
[13:59] <ralsina> alecu: how's the status on the windows proxy bug?
[14:00] <ceed^> okay, thanks
[14:00] <alecu> ralsina, I'm testing the twisted patch, and will submit it in a little while.
[14:00] <ralsina> alecu: awesome
[14:01] <ralsina> alecu: it's a patch on twisted, not on our code?
[14:01] <ralsina> alecu: because if it needs patching for us, dobey needs to hold 3.0.1 for it
[14:01] <dobey> ?
[14:02] <alecu> ralsina, I think it will be just a twisted patch, yes.
[14:02] <ralsina> dobey: we have a critical bug with proxy support on windows which we need to fix ASAP
[14:03] <ralsina> dobey: OTOH, you already did tarballs for 3.0.1 it's just not in precise, right?
[14:03] <dobey> ralsina: right
[14:03] <ralsina> ok, 3.0.2 then :-)
[14:03] <dobey> ralsina: i figured "release next week" meant "make 3.0.2"
[14:03] <ralsina> dobey: correct, I got off-by-one there
[14:07]  * ceed^ is counting down 40 minutes :)
[14:09] <ceed^> brb
[14:18] <ceed^> So it's 3.0.1 I'm waiting for?
[14:19] <karni> ceed^: By any chance, weren't you the person I shared an apk with recently? :) (1.1.4-sth or 1.2.0-sth)
[14:20] <ceed^> yes, that's me.
[14:21] <karni> ceed^: How did you like it? Any unexpected behaviour? All good?
[14:22] <ceed^> All good! I wish for shortcut to specific folders to put on the home screen, but the client works great so far
[14:22] <karni> ceed^: Great! We should have that some time soon, thanks :)
[14:22] <karni> ceed^: FYI, I published 1.2.0 last night on Google Play.
[14:23] <ceed^> I hate bringing up features from Dropbox all the time, but that's where I come from so to speak :)
[14:23] <ceed^> karni: Yes, I now see there was an update of the U1 client on my phone.
[14:24] <karni> ceed^: There's nothing to hate. I'd say it's quite a basic functionality any file-browsing/based app should have :)
[14:24] <karni> ceed^: Also, Dropboxes app is pretty solid, we've been there first with auto upload, they've been there first with other candy :)
[14:25] <ceed^> good to hear, karni. But in the end reliabilty over time is the most important feature. I would hate to find a broken backup in the cloud
[14:25] <dobey> ceed^: yes, ubuntuone-client 3.0.1 and ubuntuone-control-panel 3.0.1; publishing to the ppa archive seems to be going a bit slow though
[14:26] <karni> ceed^: Agreed. This is why we haven't introduced *sync* quickly and without thinking much about it. It's a delicate thing, so I will make sure we get it right when it's released.
[14:29] <ceed^> The one thing I've noticed is that the daemon consumes 500 MB ram at all times even when idle. That's about twice of what Dropbox uses. I wonder if that is my system?
[14:32] <mmcc> morning folks.
[14:34] <dobey> ceed^: it's probably because of extremely large number of small files
[14:40] <dobey> ralsina, thisfred, mmcc: anyone care to give a quick look over https://code.launchpad.net/~dobey/ubuntu-sso-client/update-stable-4-0/+merge/108754 ?
[14:40] <mmcc> dobey, sure
[14:40]  * thisfred looks
[14:43] <thisfred> dobey: as a wider philosophical point : why do we review these backports, and how are we supposed to? I assume the code has been reviewed and tested on trunk, so I'm not sure what I'm supposed to look for. Clashes with the stable branch? I don't think code inspection is going to reveal those...
[14:43]  * mmcc was going to ask...
[14:44] <ralsina> thisfred, mmcc: you are supposed to run the tests and look at the code just in case dobey missed something when doing the proposal. It has happened.
[14:45] <thisfred> ralsina: don't we have tarmac to run the tests? And I don't know how I'm gonna spot anything that's been missed when it's a merge of 5 different branches
[14:45] <thisfred> it feels like busywork to me, that has a very low chance of actually catching a problem
[14:46] <dobey> we have tarmac to run the tests. mostly you just approve it and move on :)
[14:46] <ralsina> also, dobey doesn't run the tests on windows, so that's something to check
[14:47] <dobey> thisfred: basically, a quick scan to make sure the changes look reasonable. and running the tests on win/osx is always nice
[14:48] <mmcc> thisfred: you want to split running the tests on win/osx? you pick
[14:48] <thisfred> I'm happy to rubberstamp these. I don't have a windows setup that I can easily switch to, and no mac env at all
[14:48] <mmcc> one question - the merge is to stable-4-0, but line 158 sets the version to 3.99?
[14:49] <dobey> mmcc: yes. 4.0.0 will be released in october
[14:50] <dobey> mmcc: we make the stable branch at the beginning of the cycle
[14:50] <mmcc> ok, so you just don't bump the number until the release?
[14:52] <dobey> right
[14:52] <dobey> so we have a more clear commit to tag to say "this is in fact, this exact version"
[14:52] <mmcc> dobey: ok, makes sense
[14:53] <dobey> ceed^: ok. so publishing to a PPA is apparently going very slowly today, but 3.0.1 is currently in the process of publishing itself into the stable ppa
[14:53] <dobey> ppa:ubuntuone/stable that is
[15:00] <gatox> me
[15:00] <dobey> thisfred: for the record, i do not like the multiple-commits-in-one for doing the release to stable-4-0 either. but i got overruled on that
[15:01] <mandel> me
[15:01] <dobey> meh
[15:01] <mmcc> does anyone else get intermittent test hangs on osx for ubuntu_sso.utils.tests.test_tcpactivation test_is_already_running?
[15:01] <mmcc> me
[15:01] <briancurtin> me
[15:01] <dobey> ralsina, thisfred, alecu, briancurtin: it's that time apparently
[15:01] <ralsina> Oh
[15:01] <ralsina> me
[15:02] <alecu> me
[15:02] <mmcc> ok, more than intermittent
[15:03] <mandel> mmcc, no, I have seen no issues
[15:03] <mandel> mmcc, after standup let me take a look
[15:03] <gatox> DONE:
[15:03] <gatox> Fight and fight with the fsevents tests for mac os. Find some weird problems about the latency and the events received from the system. Talking with mandel about this.
[15:03] <gatox> TODO:
[15:03] <gatox> Try to find a workaround for these problems.
[15:03] <gatox> BLOCKED:
[15:03] <gatox> Trying to fix the macfsevents lib limitations.
[15:03] <gatox> mandel, go
[15:04] <mandel> DONE: Improve python api for fsevents. We now have better erro handling and tests.
[15:04] <mandel> TODO: Debugging with gatox. Reduce the size of the fsevents json used for ipc. Ask about a license concern (I do have a BSD alternative).
[15:04] <mandel> BLOCKED: no
[15:04] <mandel> dobey, please
[15:04] <dobey> λ DONE: releases, uploads, SRU poking, client/cp 3.0.1 in stable PPA
[15:04] <dobey> λ TODO: finish releases/uploads, SRU poking, tarmac tweakery
[15:04] <dobey> λ BLCK: None.
[15:04] <dobey> mmcc
[15:04] <mmcc> DONE: monday lost to Qt plugin support
[15:04] <mmcc> TODO: maybe just build a static Qt?
[15:04] <mmcc> BLCK: NO
[15:04] <mmcc> NEXT: briancurtin
[15:04] <briancurtin> DONE: testability is killing me, yet another day of basically nothing. building with mingw doesn't seem to work. will check with elopio if he ever got it built himself and what on earth he did...
[15:04] <briancurtin> TODO: finalizing lazr removal, moving on to finfixing the C2C page. will come back to testability
[15:04] <briancurtin> NEXT: ralsina
[15:04] <ralsina> DONE: bug triaging, askubuntu, canonicaladmin, other sites that drain my will to live on a daily basis, proposed a branch (yay) for sorting shares in u1cp which showed some test weaknesses, so it wasuseful. TODO: merge that, drain will to live further, start on another bug blOCKED: no, NEXT alecu
[15:04] <thisfred> me
[15:04] <alecu> DONE: worked on patch for bug #1006899
[15:04] <alecu> TODO: reproduce on a vm the bug that popey found
[15:04] <alecu> BLOCKED: no
[15:05] <alecu> NEXT: thisfred
[15:05] <thisfred> DONE: https://code.launchpad.net/~thisfred/u1db/home-on-the-range/+merge/108606
[15:05] <thisfred> TODO: wildcards in ranges, planning piker
[15:05] <thisfred> poker
[15:05] <thisfred> BLOCKED: no
[15:06] <alecu> thisfred, planning poker?
[15:06] <ralsina> briancurtin, mmcc: want to talk about the problems you are experiencing?
[15:07] <ralsina> I am all ears
[15:07] <alecu> thisfred, did I miss a memo?
[15:07] <ralsina> alecu: it's just for u1db
[15:07] <thisfred> alecu: sry: u1db
[15:07] <alecu> thisfred, ralsina: oh, ok :-)
[15:07] <thisfred> welcome to join though :D
[15:07] <ralsina> alecu: I suspect they just play poker
[15:07] <thisfred> very high stakes
[15:07] <ralsina> and they are planning to play MORE poker
[15:08] <ralsina> "This is a high stakes game, you know." -- Vlad the impaler
[15:08] <dobey> ceed^: looks like it is published, so you should be able to apt-get update/upgrade with ppa:ubuntuone/stable added, and get ubuntuone-client 3.0.1
[15:09] <mmcc> ralsina: groan
[15:09] <alecu> lolz
[15:09] <ralsina> mmcc: I am trying to fill in for mandel
[15:09] <ralsina> and BTW, where's mandel?
[15:09] <ceed^> apt-getting now. Thanks
[15:10] <mmcc> ralsina: so I summarized my status when I left last night: it looks like something inside qtnetwork or twisted is forking a new Qt process and that process doesn't have the right qt.conf. So I can put the image plugins in the app pkg and they will be found correctly but the code loading the network plugins doesn't look in the same path
[15:10] <ralsina> Oh, up there, missed it. Ok
[15:10] <briancurtin> ralsina: poor/out of date docs. since i've gotten the furthest with mingw build i will give it one more shot before moving on, but the docs seem to pick up in the middle and assume too much. this really shouldn't be that hard...
[15:10] <mandel> ralsina, I'm here
[15:10] <mandel> ralsina, I did the standup etc..
[15:10] <ralsina> mandel: ok, I missed your standup, sorry
[15:10] <mandel> ralsina, nad told mmcc  to tell me what was wrong with his problem :P
[15:10] <ceed^> yep, installed. Going to restart U1 and see if it works right now.
[15:10] <ralsina> briancurtin: you may want to ask for help in the qt forums, for example
[15:11] <briancurtin> ralsina: yeah that's next on the list, ask for help there and move on to other stuff while keeping tabs on the forums
[15:11] <ralsina> mmcc: interesting, if you said "thread" that would sound familiar
[15:11] <ralsina> mmcc: but I have not seen qt spawn processes, so it may be something osx-speciic
[15:12] <ceed^> wow it's much faster at the initial check!
[15:12] <mmcc> mandel, it's simple: I did bzr branch lp:~dobey/ubuntu-sso-client/update-stable-4-0 sso-client-stable-4-0 , then I run-mac-tests, then it stalls on test_is_already_running in tcpactivation
[15:12] <mandel> mmcc, can you try with trunk please?
[15:12] <ralsina> mmcc: OTOH we are spawning processes for a lot of things (like proxy detection) but you should notice those
[15:12] <mmcc> mandel: it's done that before and restarting the tests makes it stop but now it's doing it every time
[15:12] <ralsina> mmcc: check you don't have a sso already running
[15:13] <mandel> mmcc, I wonder if the domain sockets changes are not in stable 4,  are they dobey?
[15:13] <dobey> ceed^: cool
[15:13] <dobey> mandel: huh?
[15:13] <mmcc> ralsina: this only happens when I don't already have an sso running. just a sec and I'll find the code that spawns it
[15:13] <mandel> dobey, no worries, I'll take a look :)
[15:14] <mmcc> my mac shut itself down overnight, so all my emacs buffers went away :(
[15:14] <ralsina> mmcc: so maybe our spawner code is Doing It Wrong ®
[15:16] <mmcc> ralsina: the plugin load happens after ubuntu_sso/utils/webclient/qtnetwork.py:121 prints "Querying OS for proxy"
[15:16] <dobey> mandel: the merge proposal is "merge what is in trunk, into stable-4-0"
[15:16] <ralsina> mmcc: ok, that's expected, we have a separate process for obtaining proxy info
[15:17] <mandel> dobey, ok, I also did a grep to look for enpoints and is there :(
[15:17] <ralsina> mmcc: you can run it yourself, it's ubuntuone-proxy-tunnel
[15:17] <dobey> mandel: so unless you're using an old dev tools or something, i don't see how that would be the problem. unless of course that code is just broken :)
[15:17] <mmcc> ralsina: aha, ok, so I need to package that one too I guess
[15:17] <gatox> ralsina, do you have time for mumble??
[15:17] <ralsina> mmcc: yep :-)
[15:17] <ralsina> gatox: sure
[15:17] <gatox> ralsina, we are with mandel talking about some fsevents issues
[15:17] <ralsina> gatox: let me find my headset
[15:18] <mandel> mmcc, check what dobey said, you might be using and old dev-tools that is not closing the twisted services correctly
[15:18] <mandel> mmcc, can you do a pull from lp:ubuntuone-dev-tools install it and run the tests in you mac?
[15:18] <mmcc> mandel: argh, devtools is on rev 73 now, isn't it?
[15:19] <mandel> mmcc, yes, is a little higher that it used to be
[15:19] <ralsina> mandel, gatox: I can hear you
[15:19] <gatox> ralsina, we can't hear you
[15:19] <ralsina> mandel:, gatox: NO HOLIDAYS ;-)
[15:19] <gatox> jajaja
[15:19] <mandel> :(
[15:20] <ralsina> Looks like I can't mumble today
[15:21] <dobey> maybe you should stop mumbling, and SPEAK UP!
[15:22] <mmcc> mandel, I'm pretty sure that was my problem. thanks.
[15:22]  * mmcc can't seem to keep his deps up to date
[15:22] <mandel> mmcc, hurray!!!
[15:22]  * mmcc spoke too soon
[15:22] <mandel> not working?
[15:23] <mmcc> mandel: same hang on test_is_already_running
[15:23] <mmcc> do we use tcpactivation on darwin?
[15:23] <mmcc> I'm a little confused with the naming, since we don't use TCP, right?
[15:23] <mandel> mmcc, yes we do, the funny thing is that it works on linux which is the same code
[15:24] <mmcc> not funny ha-ha
[15:24] <mandel> mmcc, yes, naming is bad (I also think we should do launchd activities)
[15:25] <mmcc> mandel: yeah, launchd is the preferred way, but unless you need it, like for the root daemon, it's a pain. I think it'd make testing annoying
[15:27] <dobey> alright, am off to get lunch. bbiab
[15:28] <mandel> mmcc, so on linux everything goes ok, let me try on the mac
[15:29] <mmcc> mandel: ok
[15:33] <mandel> mmcc, I can confirm that I have the same issue.. I wonder what is the diff from os x and linux2
[15:34] <mandel> mmcc, can you add a bug in sso and assign it to me so that I fix this asap
[15:35] <mmcc> mandel: sure
[15:36] <mandel> mmcc, thx!
[15:37] <gatox> alecu, ping....... whenever you are (before mandel eod if possible :P)
[15:38] <mmcc> mandel: https://bugs.launchpad.net/ubuntu-sso-client/+bug/1009071
[15:39] <mandel> mmcc, superb, thx, I'll try to fix it asap
[15:40] <mmcc> ok thanks mandel
[15:42] <alecu> gatox: pong
[15:42] <alecu> gatox: mumble?
[15:42] <gatox> alecu, ack
[15:42] <gatox> mandel, mumble again?
[15:45] <mandel> gatox, sure, sorry
[15:45] <mandel> gatox, joining now
[16:27] <dobey> mmcc: did you fix your tests issue?
[16:27] <dobey> oh i guess not
[16:27] <mmcc> dobey: no, mandel reproduced it and is going to look at it
[16:28] <dobey> i guess they are passing on windows though
[16:28] <mmcc> dobey. however, all the tests do pass on windows
[16:29] <mmcc> yeah, and I recognized all the changes as passing on macos and linux earlier, so I'm going to +1
[16:29] <dobey> hmm
[16:29] <mmcc> hmm?
[16:30] <dobey> oh, the ubuntu-sso-client-windows-test jenkins job hadn't run in almost 6 days
[16:30] <mmcc> hmm.
[16:30] <dobey> but neither have any branches landed in it since then, and it's set to run when the branch changes
[16:30]  * gatox lunch!!!!
[16:33]  * ralsina goes do something trivial on sso so it doesn't feel left out
[16:34] <dobey> weird
[16:41] <mmcc> how do I sign out of SSO on linux? I want to look at the login screen
[16:41] <beuno> mmcc, I think, you delete the token from the keyring
[16:42] <dobey> you open ubuntuone-control-panel-qt
[16:42] <dobey> and on devices tab click the [Remove] button at the top
[16:43] <mmcc> dobey, that just logs me out, and nothing else?
[16:43] <dobey> mmcc: it disconnects the computer you do it on, from u1, also removing that token from the server (since it's now useless)
[16:44] <mmcc> dobey: ok, thanks.
[16:44] <dobey> mmcc: you probably want to click [Disconnect] in the control panel first, to have syncdaemon stop talking to the server, since i don't think it auto-disconnects when the token is removed
[16:45] <mmcc> dobey: ok
[16:45] <dobey> hrmm. what in the royal hell is going on with this branch
[16:46] <mmcc> hrm, press the [Remove] button and you get a confirmation dialog that asks you if you're sure without giving any info as to why you might not be sure... and the title of the dialog is 'ubuntuone-installer'
[16:46]  * dobey did not design or implement it
[16:46]  * mmcc is not blaming dobey
[16:47] <dobey> but why the heck would ./run-tests fail to run, without giving any output
[16:47] <mmcc> hey, yeah I saw that just now. wtf
[16:48] <dobey> looks like xvfb was not installed tehre
[16:48] <dobey> there
[16:48] <dobey> but i would expect some error about that
[16:48] <mmcc> oh yeah I ran into that when I was setting up the tests. it fails silently
[16:49] <dobey> weird
[16:49] <dobey> because the script has set -e
[16:49] <dobey> maybe we need to stick a -x in there too
[16:50] <dobey> but oddly, i am also getting this, running the tests locally in sso trunk (and in the backport branch): http://pastebin.ubuntu.com/1025328/
[16:50] <mmcc> hrm - SSO client shows a "?" icon in the launcher on linux?
[16:50] <dobey> oh
[16:50] <dobey> nevermind
[16:50] <dobey> of course i'm getting that
[16:51] <dobey> mmcc: it doesn't have an icon, so yes
[16:51] <dobey> well, it does that in unity anyway
[16:52] <dobey> or maybe that's the human theme
[16:52] <dobey> need to poke at barry's changes
[16:52] <mmcc> so on osx I'm hiding SSO client from the dock, but if we got an icon for it, we could show it in both places...
[16:52] <mmcc> the '?' icon isn't great
[16:53] <dobey> yeah, there's an open bug about the icon on sso dialogs
[16:53] <mmcc> ah, ok
[16:54] <mmcc> dobey: not bug 992595?
[16:54] <dobey> oh i guess i also need to release ubuntuone-dev-tools first
[16:54] <dobey> mmcc: the backend is a different issue i guess
[16:55] <mmcc> dobey: yes, but I couldn't find any other icon-related sso bugs.
[16:56] <dobey> https://bugs.launchpad.net/ubuntu/+source/ubuntu-sso-client/+bug/840045
[16:56] <dobey> https://bugs.launchpad.net/ubuntu-sso-client/+bug/966996
[16:56] <mmcc> arg, why didn't searching for 'icon' in ubuntu-sso-client find 840045?
[16:57] <dobey> because it's filed against the ubuntu package, not the upstream project
[16:57] <mmcc> hey, that sounds familiar
[16:57] <dobey> and now it's a duplicate of the other one anyway
[16:57] <mmcc> thanks
[16:58] <dobey> of 966996 that is
[16:58] <dobey> dear UPS, where is my RAM
[17:05] <dobey> ugh and i think my coke went flat. and i don't have a CO2 bubbler
[17:07] <mmcc> ralsina you wanted something trivial for sso? bug 1009107
[17:07] <ralsina> haha
[17:07] <ralsina> mmcc: sounds trivial enough, since the start pageis not supposed to have a back button
[17:07] <mmcc> yep
[17:08] <ralsina> I should check every bug reported on package and attach it to projects. The idea makes me want to gouge my eyes with a spoon.
[17:09] <mmcc> be nice if lp was easily scriptable. I wouldn't mind a 'newbug' script to file a bug in one line
[17:10] <ralsina> anyone know WTF is a "likewise user" in a linux+windows domain context?
[17:10]  * ralsina hates projects that use common words 
[17:10] <ralsina> mmcc: lp is scriptable, don't know about the "easily" part
[17:11] <mmcc> ralsina: any more context on that 'likewise user'?
[17:11] <ralsina> mmcc: bug #983756
[17:11] <Koukourikos> Hello to everybody!! I have a QUESTION: Ihave uploaded a video(mp4) in ubuntuone and I made it public but when I use the link for download I dont get the file but the video starts to play in the browser. How can I download the video ?
[17:12] <beuno> Koukourikos, there's a secret thing you can use
[17:12] <ralsina> likewise-open - Authentication services for Active Directory domains
[17:12] <dobey> ralsina: http://likewise.com/ i guess?
[17:12] <beuno> add this to the url:  ?attachment=true
[17:12] <ralsina> ok, lunch
[17:12] <ralsina> see you all in a bit
[17:13] <dobey> ralsina: also, found this: https://help.ubuntu.com/8.04/serverguide/likewise-open.html
[17:14] <Koukourikos> bueno Thank you so much!!! but why is this secret ?
[17:19] <mandel> beuno, ^
[17:19] <beuno> Koukourikos, well, it was originally added as a workaround
[17:19] <beuno> and we haven't get figured out a good way to expose this in the interface
[17:19] <beuno> it's hard to communicate since sometimes we can show the file and sometimes we can't
[17:19] <mandel> dobey, I have found that the run-tests in ubuntuone-dev-tools cannot be used on darwin, shall I add a run-mac-tests script or try and do some magic and detect if we are on mac in the run-tests one?
[17:19] <beuno> so it;s confusing
[17:20] <dobey> mandel: why can't it be used?
[17:20] <mmcc> mandel I vote for magic in the existing script
[17:20] <mandel> dobey, very simple, wrong twisted reactor (glib)
[17:21] <mandel> mmcc, I'm not very picky, so whatever you guys prefer
[17:21] <dobey> mandel: i think i'd be fine with having u1trial use a different default reactor based on sys.platform or whatever
[17:22] <mandel> dobey, maybe the select one on windows and mac, what do you think?
[17:22] <mandel> dobey, I need to go, but let me know here and I'll fix it tom morning, I think mmcc bug is related to u1-devtools (tcpactivation)
[17:23] <dobey> mandel: whatever is appropriate, is fine by me. do all the tests pass if you just run the tests with the right reactor?
[17:23] <dobey> ok, got get your ass kicked some more in rugby :)
[17:23] <mandel> dobey, they do pass, so it a super trivial fix :)
[17:24] <dobey> ok
[17:24] <mandel> dobey, ok, I'll deal with that tom morning then
[17:24] <mandel> all, have a great evening!!!
[17:24] <briancurtin> elopio: so i think i have testability built on my machine. do you have any info on how to use U1 with testability, so i can see that it actually functions before i go further?
[17:27] <briancurtin> ralsina: 1-1?
[17:27] <dobey> i hate having to triage bugs when i make the release
[17:27] <ralsina> briancurtin: lunch. In 30?
[17:27] <briancurtin> ralsina: sounds good
[17:28] <dobey> especially when lp is being a pain
[17:29] <ralsina> dobey: which is on dys ending with "y"
[17:30] <dobey> dys is a good analogy
[17:32] <Koukourikos> bueno: xmm... ok I think I can understand. Anyway thank you VERY MUCH you just saved a video-present-surprise for a friend... thank you and see you around :)
[17:36] <mmcc> briancurtin: if you end up with notes on building testability, please share them so I can do that for darwin while I still understand the plugin issues here
[17:37] <mmcc> briancurtin: at your leisure - I can wait until you're done
[17:37] <briancurtin> mmcc: open up visual studio command prompt, type "nmake"
[17:37] <briancurtin> not documented anywhere, just mentioned in passing without saying anything
[17:37] <briancurtin> i went around in circles for days trying to get mingw and all sorts of ruby crap to work together
[17:38] <briancurtin> i have no idea what you'll need to do to actually compile it, but i can point you to the source
[17:38] <mmcc> briancurtin: hmm, VS cmd prompt, eh? sounds like I'll have some work to do to use your advice :)
[17:38] <mmcc> briancurtin: but I'll make a note to remember nmake
[17:39] <briancurtin> mmcc: well you'll probably just use regular old make and it might be able to read that makefile, i'm not sure. this doc had me installing all kinds of BS and none of it does anything
[17:39] <briancurtin> i still dont even know if i have the right thing
[17:39] <elopio> briancurtin: did you build the visualizer?
[17:39] <briancurtin> i did get a testability folder which contains testability.dll
[17:39] <briancurtin> elopio: not yet, just the "agent_qt" project so far. i can try the visualizer next
[17:40] <elopio> ok. If you start the qttasserver, and then the sample calculator with the -testability option it's visible on the visualizer.
[17:40] <elopio> you should be able to start the control panel with the -testability option and see it there too. If it appears, testability is working on the app.
[17:51] <briancurtin> elopio: thanks, i'll try to get that stuff working
[17:52] <mmcc> will you look at the time
[17:52]  * mmcc lunch
[17:53] <ralsina> briancurtin: 1-1, prefer mumble or IRC?
[17:53] <elopio> briancurtin: thanks. And you can also take a look at how it works on ubuntu: http://pad.ubuntu.com/u1-testability
[18:25] <alecu> briancurtin, when you can, please re-review this branch: https://code.launchpad.net/~alecu/ubuntuone-windows-installer/tunnel-use-console/+merge/108279
[18:25] <alecu> briancurtin, but using the twisted patch mentioned.
[18:25] <briancurtin> alecu: will do
[18:25] <alecu> thanks!
[18:25]  * alecu will have some lunch now.
[18:26] <alecu> gatox_ultra, why "ultra"?????
[18:27] <gatox_ultra> alecu, because now i'm in the ultrabook :P
[18:27] <alecu> gatox_ultra, .... ultrabook? what did you buy now?????
[18:27] <gatox_ultra> alecu, using the mac with vnc from the couch
[18:27] <gatox_ultra> :p
[18:28] <gatox_ultra> alecu, i bought the asus ux31... and i think is the best thing i bought!! i like it even more than the alienware... and it works like a charm with ubuntu (maybe that's why i like it more than the alienware jje)
[18:29] <alecu> gatox_ultra, awesome, congrats!
[18:29] <ralsina> gatox: use quassel so you have the same IRC from everywhere intead of rubbing it in our faces ;-)
[18:29] <alecu> gatox_ultra, what will you be doing with the alienware?
[18:29] <gatox_ultra> alecu, thx! :D......... i really recommend this lapto
[18:30] <gatox_ultra> ralsina, jejej i could do that too
[18:30] <alecu> gatox_ultra, but the ux31 has no colored leds!
[18:31] <dobey> haha
[18:31] <gatox_ultra> alecu, i'm using the alienware as a desktop pc now... but i haven't lose my faith... i'm going to make the videocard work on that laptop and ubuntu is goiing to stop crashing
[18:40] <briancurtin> asus ux31
[18:40] <briancurtin> ha, meant to paste that in google
[18:41] <dobey> irc is pretty much the same
[18:41] <dobey> similar signal:kittens ratio
[19:15] <ceed^> Magicicada doesn't work right after U1 got updated to 3.0.1 today.
[19:17] <dobey> how so?
[19:18] <ceed^> I know it runs since I can tail the log, but it doesn't show the files in the window. It's blank. If I open and close it a few times it starts showing the file activity again for a while, but then it goes blank again.
[19:18] <dobey> ceed^: do you have the same issue as earlier showing up in ~/.cache/ubuntuone/log/syncdaemon-exceptions.log ?
[19:20] <ceed^> dobey. not one entry in that log so far.
[19:21] <dobey> interesting
[19:22] <ceed^> good or bad interesting? :)
[19:22] <dobey> ceed^: does "u1sdtool -s" exhibit similar problems when magicicada stops working right for you?
[19:22] <dobey> unknown interesting :)
[19:22] <ceed^> lemme check
[19:24] <ceed^> dobey: no, it's only magicicada. Seems to be a display problem since everyting else in the application works normal. It's just the running file listing in the window that disappears.
[19:25] <dobey> ceed^: odd.
[19:25] <dobey> ceed^: maybe they're just transferring faster now, and disappearing from the list? :)
[19:25] <ceed^> And when I click in the window where the file sync listing normally is it turns pinkish
[19:25] <dobey> weird
[19:25] <dobey> no idea what that would be
[19:25] <ceed^> dobey: I wish! :)
[19:26] <dobey> facundobatista: ^^ any idea about magicicada?
[19:26] <ceed^> Now it went blank again...
[19:26] <dobey> ceed^: does u1sdtool --waiting show what you'd expect to see in the UI?
[19:27] <ceed^> yes
[19:29] <dobey> ok
[19:29] <dobey> maybe you should open a bug against magicicada
[19:32] <ceed^> I will
[20:00] <dobey> eep. way too many changes in sso client
[20:01] <dobey> i need a break or something
[20:08] <gatox> ok, eod for me.... and i need to stop reading this code for my own sanity :P
[20:10] <gatox> see you tomorrow people!
[20:17] <dobey> mmcc: can you test https://code.launchpad.net/~dobey/ubuntuone-dev-tools/default-reactor/+merge/108823 on osx/windows please? :)
[20:17] <mmcc> dobey, sure
[20:22] <mmcc> tests pass on macos, runs much faster too
[20:22] <mmcc> at least until it gets to linting
[20:23] <dobey> faster than when it didn't work?
[20:23] <mmcc> well, it worked occasionally before :) maybe it was timing out?
[20:24] <dobey> on macos i'd guess they weren't working at all, due to the linked bug :)
[20:24] <mmcc> no, I meant on macos it worked occasionally.
[20:25] <dobey> shouldn't have
[20:25] <mmcc> I tested a few merges with that broken code ... I have no idea why it worked sometimes and then stopped today
[20:25] <dobey> unless you had glib and pygobject/pygtk installed as well
[20:25] <mmcc> yeah, nope
[20:26] <dobey> you mean you tested branches of ubuntuone-dev-tools, or branches of other things using ubuntuone-dev-tools via their run-mac-tests scripts?
[20:26] <mmcc> I tested branches of sso using devtools
[20:27] <dobey> right. sso's run-mac-tests probably specifies which reactor to use
[20:27] <dobey> but ./run-tests in ubuntuone-dev-tools should have failed previously
[20:28] <mmcc> AFAIK I never ran run-tests in dev-tools.
[20:29] <mmcc> sso's run-mac-tests does specify reactor=qt4, but that's where I was seeing my issues
[20:29] <mmcc> so, hmm
[20:29] <dobey> ah. hrmm
[20:30] <dobey> probably unrelated to this change then :)
[20:31]  * dobey makes the run-tests.bat there use the default reactor as a means to prevent regressions, as well
[20:36] <mmcc> dobey, sorry, which run-tests.bat are you changing right now?
[20:37] <dobey> mmcc: in ubuntuone-dev-tools
[20:37] <dobey> i just pushed my change to that branch
[20:38] <dobey> without changing that, it would definitely continue working on windows anyway
[20:39] <mmcc> ah ok. I merged on windows just after you pushed I guess. glad Iw asn't seeing things
[20:39] <mmcc> anyway yeah, on windows it works just fine before and after your change
[20:39] <mmcc> although it complains about pep8 - known problem or my setup?
[20:40] <alecu> mmcc, how does it complain?
[20:41] <alecu> mmcc, perhaps it's not finding the pep8 executable?
[20:41] <mmcc> trying to copy the error...
[20:42] <mmcc> the dev-tools/run-tests.bat output from after the coverage is printed: http://paste.ubuntu.com/1025704
[20:43] <dobey> hrmm
[20:50] <dobey> weird
[20:50] <dobey> that one doesn't make any sense
[20:50] <mmcc> also, while we're here - u1lint doesn't find dirspec when running run-tests in dev-tools. It isn't using the buildout-generated python, I guess
[20:50] <dobey> unless for some reason sys.platform != 'win32' wher eyou ran that
[20:51] <mmcc> dobey: just tested, sys.platform is 'win32' in the shell where I got that output
[20:52] <mmcc> sorry, the comment above is about u1lint and dirspec on *macos*
[20:52] <mmcc> s/above/two lines above/
[20:55] <mmcc> so now on windows and macos, dev-tools' run-mac-tests runs the same tests twice...
[20:56] <dobey> mmcc: can you update from my branch on windows and try run-tests.bat again? r78 should fix all those issues you pasted
[20:57] <mmcc> dobey, ok
[20:57] <mmcc> btw on macos, adding 'python' to the beginning of line 22 of devtools' run-mac-tests solves my other issue
[20:57] <mmcc> but we probably want to turn off line 20 on mac/win
[20:58]  * mmcc goes to test r78 on windows
[20:58] <dobey> mmcc: devtools dosen't have a run-mac-tests script
[20:58] <mmcc> my bad, s/mac//
[20:59] <dobey> hrmm, that's weird, re: prefixing with python
[21:01] <mmcc> it's because u1lint has a hashbang that says /usr/bin/python and we need the python in PATH
[21:02] <dobey> by why doesn't u1trial fail?
[21:02] <dobey> oh because it's doing env
[21:02] <dobey> which is wrong for other reasons
[21:03]  * dobey looks longingly at his compiler book
[21:14] <dobey> mmcc: does it work cleanly on win now?
[21:19] <mmcc> dobey: sorry, afk for a sec. running run-tests.bat now
[21:19] <mmcc> no. says u1lint: error: -i option requires an argument
[21:21] <dobey> oh
[21:22] <dobey> oops. forgot to put the variable in there
[21:22] <dobey> mmcc: just pushed r79 to fix that
[21:23] <dobey> also, usb1.0 is horribly slow
[21:25] <mmcc> dobey yes with that change it's clean
[21:25] <mmcc> on windows
[21:26] <dobey> ok, now let's fix osx :)
[21:27] <mmcc> how 'bout just the dev-tools tests and not the whole os?
[21:29] <dobey> heh
[21:29] <dobey> someone has to do it, eh
[21:31] <mmcc> we're gonna need a bigger boat
[21:33] <mmcc> I want to file a "need .icns file for mac os packaging" bug but the right place to do it is still called "ubuntuone-windows-installer... I guess that won't stop me
[21:35] <dobey> not sure that is the right place actually
[21:35] <dobey> but we can always move it
[21:35] <mmcc> hrm, what would the right place be?
[21:35] <dobey> mmcc: just pushed up to r81 in my branch. *should* solve your issues on osx if you could try it please
[21:35] <mmcc> dobey my pleasure
[21:36] <dobey> mmcc: well, the right place doesn't exist as a project yet. hopefully i can make it exist this week or next, though
[21:36] <mmcc> dobey: go on... I thought we were going to rename windows-installer to 'packaging tools' or something and dump the osx setup.py and misc in there?
[21:37] <dobey> mmcc: yes. but we need a common place for data. not sure a project that's all packaging stuff is the right place for that
[21:38] <mmcc> you mean data like icon files?
[21:38] <dobey> or maybe, the project i'm thinking of is more generically named, and it's the right place to put packaging data/scripts next to other stuff, which we'll need on all platforms
[21:39] <dobey> yeah, icons, graphics, fonts, etc
[21:39] <mmcc> ahok
[21:42] <mmcc> dobey, run-tests runs lint OK, but still runs the same set of tests twice (since the default reactor is now twisted on osx) and now it tells me I don't have pep8 installed (it's right)
[21:42] <mmcc> also, why $PYTHON and not just python?
[21:43] <dobey> so we only have to change it once in the future if we need to (if we switch to python3 for example)
[21:44] <dobey> yeah, the tests running twice is mildly confusing but i guess not a real problem
[21:44] <mmcc> ah. color me ignorant of python3 issues. thanks
[21:45] <dobey> well i don't know about on osx, but on ubuntu/debian, installing python3 doesn't replace the /usr/bin/python symlink; for good reason :)
[21:46] <mmcc> the tests are fast, but a quick test of uname -s might be nice? I won't argue too hard for this thoguh
[21:47] <mmcc> you don't want to know about osx. it all depends on how you install it. homebrew? python.org installer? from source? a million ways to screw it up. py3 isn't on 10.6 by default, but probably apple did something reasonably smart in 10.7 or .8, wherever they first put it in
[21:47] <dobey> heh
[21:50] <dobey> ok well i am off
[21:50] <dobey> have a good evening :)
[21:50] <mmcc> ok, bye dobey
[22:32] <mmcc> mmcc out