[08:21] <JamesTait> G'mornin' all! :D
[08:29] <mandel> ralsina, ok, I'll do that now so its ready for when you get here
[08:30] <mandel> oh, morning!
[10:52] <rye> mandel: ping, is currently released windows version safe for user profiles with UTf-8 names?
[10:52] <rye> i mean whether it works
[10:52] <mandel> rye, AFAIK it should because diego tests everything with a username that is a single utf8 char
[10:53] <mandel> rye, unless you found something we don't know about it should be more than ok
[10:55] <rye> mandel: bug #979759
[10:56] <rye> mandel: btw, what's the latest release?
[10:57] <rye> mandel, ah 19-Dec-2011
[10:58] <mandel> rye, is old.. very old, mainly due to the work we have been doing on linux, we plan to have dainly build set for windows by the end of this week
[10:58] <rye> mandel: i suppose ubuntuone-2.99.3-windows-installer.exe was pulled off at some point
[10:58] <mandel> rye, a number of bug were present, like no autoupdate O_o
[10:58] <mandel> rye, so it was a 'useless' release :(
[10:59] <rye> mandel: oh, ok
[10:59] <mandel> rye, there is a new package ready from yesterday to be used by QA before the release, do you know how to get those?
[11:00] <mandel> rye, I know the RT has been done and we are waiting to get the binaries sign
[11:03] <rye> mandel: no, no logs with release or ubuntuone.com in lastlog here
[11:04] <mandel> rye, should happen soon, maybe the .exe was sent manually via mail.. I'm trying to fix that so that you can find this as a jenkins artifact, can you wait 'til ralsina or briancurtin are back? they should help you better
[11:04] <rye> mandel: sure
[11:05] <mandel> rye, they should be here in a few mins
[12:11] <alecu> hola mandel!
[12:11] <mandel> alecu, buenos dias!
[12:11] <mandel> alecu, first one to appear!
[12:11] <mandel> alecu, this is sooooo boring in the mornings
[12:12] <alecu> mandel, sorry to hear that :-)
[12:13] <mandel> alecu, he, I got used to it, being the only eu one in desktop+ :)
[12:13] <mandel> alecu, what shall we do with bug 939671 bug 955339 and bug 955831
[12:14] <mandel> alecu, invalid since we are not going to allow the used of pinned certs?
[12:15] <alecu> mandel, what about "wishlist", since we would like to have this in order to have a more generic http library, but we don't need it right now?
[12:16] <alecu> mandel, if it's ok with you, I'm setting them to wishlist
[12:16] <mandel> alecu, sure, lets do that
[12:16] <mandel> alecu, I'm just fixing small bugs little by little until I can get that windows thing done :)
[12:17] <mandel> alecu, I fixed bug 904842 is a very easy branch, can you take a look?
[12:17] <mandel> MP https://code.launchpad.net/~mandel/ubuntu-sso-client/remove-duplication
[12:18] <alecu> mandel, I also added the label "webclient-ssl" to all of those, in case we need to get back to all of them together.
[12:18] <alecu> mandel, awesome. But I really need to review your two branches from yesterday before I can review this one :-)
[12:19] <mandel> alecu, sure! actually, do those first so that we can get automatic .exe from jenkins, that is going to make josh and QA very very happy
[12:20] <mandel> I'll be working on bug 961357 so that the errors make more sense
[12:30] <ralsina> good morning!
[12:36] <ralsina> mandel: thanks for doing the keyring thing. Is there any reason why we are not asking upstream to take that patch?
[12:37] <mandel> ralsina, hm.. I tried, then I got dragged to something else then there was a huge diff, I'll try again this afternoon :)
[12:37] <ralsina> mandel: cool, thanks!
[12:37] <mandel> ralsina, the updated version is there for us to be used right now so at least we are not blocked
[12:39] <mandel> ralsina, do you know which is the dialog used to show the exceptions in control panel/sso?
[12:43] <ralsina> mandel: not from memory
[12:43] <mandel> ralsina, ack, I'll have lunch and find it myself
[12:44] <mandel> ralsina, do you have a windows machine?
[12:44] <ralsina> mandel: no VM, I would have to reboot
[12:45] <mandel> ralsina, ok, no worries, I'll ask brian to run the tests on his and get the +1 for the sso branches
[12:45] <mandel> off to lunch now
[12:45]  * mandel lunch
[12:47] <ralsina> mandel: buen provecho!
[13:02]  * alecu reboots
[13:22] <dobey> oi
[13:25] <ralsina> good morning dobey
[13:26] <ralsina> dobey: how's releases?
[13:30] <zeroEvidence> Does anyone know if the Ubuntu One team are able to recover files from ones online drive?
[13:30] <dobey> ralsina: done
[13:30] <ralsina> dobey: awesome
[13:31] <dobey> ralsina: well, i don't know if you guys finished the windows build, but i think i saw you sent it to be signed :)
[13:31] <ralsina> zeroEvidence: yes, there is a possibility. duanedesign, rye, can you help zeroEvidence?
[13:31] <dobey> zeroEvidence: you need to open a support ticket I think
[13:31] <zeroEvidence> Thank you Ralsina.
[13:31] <duanedesign> hello zeroEvidence
[13:31] <zeroEvidence> Ok, lovely. Where do I go to fill one out?
[13:31] <briancurtin2> ralsina: speaking of releases, i'm signing python27.dll right now. they also want the new update.xml - does this look right https://pastebin.canonical.com/64166/
[13:32] <duanedesign> zeroEvidence: I will send you a PM so I can get more info from you
[13:32] <dobey> ralsina: i also got the peer reviews done that i had left, last night
[13:32] <zeroEvidence> Ok, Thanks duanedesign
[13:36]  * mandel back
[13:37] <mandel> briancurtin2, I'll take a look :)
[13:37] <mandel> briancurtin2, I don't know the numbers, but from where I stand it looks perfect
[13:39] <briancurtin2> mandel: the numbers match what's in the autoupdate.xml and ubuntuone.xml on the stable 3.0 branch and release-3_0_0 tag
[13:39] <mandel> briancurtin2, then it shoudl be ok, when I said 'I don't know about the numbers' is that I don't have a clue which version we are releasing, I'm terrible with that :)
[13:40] <mandel> briancurtin2, by the way, if you have time, I'd love to have some reviews from you about the fixes for windows on sso and control panel
[13:40] <briancurtin2> mandel: send me a list and i'll check them all out. we should also have that mumble about them - after the team call?
[13:41] <mandel> briancurtin2, whenever is better for you I already had lunch :)
[13:41]  * mandel gets list
[13:41] <ralsina> dobey: awesome. I screwed up a bit on the windows release process (told brian the wrong DLL to sign) [sigh]
[13:42] <dobey> ah
[13:42] <ralsina> briancurtin2: usually I give them the XML but tell them not to deploy it for a couple of days
[13:42] <ralsina> briancurtin2: to get some testing so we don't push a broken version to users
[13:43] <ralsina> briancurtin2: but yes, that xm looks fine to me
[13:43] <dobey> bah. just push it :)
[13:43] <briancurtin2> ralsina: ok, i'll sign that one as well and post it to the RT
[13:43] <ralsina> briancurtin2: thanks, remember to explain all this to the people in the other side of the RT
[13:44] <ralsina> briancurtin2: we need to be very explicit to avoid confusions
[13:44] <mandel> briancurtin2, here you go: http://paste.ubuntu.com/926427/
[13:45] <briancurtin2> mandel: ack, thanks
[13:45] <briancurtin2> ralsina: do you have a preference for when the update.xml should go live? maybe Tuesday?
[13:45] <mandel> ralsina, briancurtin2 at some point, it would be nice to get the xml and the .exe from jenkins and for them to always follow the same steps so that we don't always have to tell them, right?
[13:46] <ralsina> mandel: right
[13:46] <ralsina> mandel: but never going to happe
[13:46] <ralsina> mandel: because they want a dev gpg-signing the binaries so we take responsibility for what we ship
[13:47] <mandel> ralsina, really, pussies!
[13:47] <mandel> cats I mean, cats..
[13:48] <briancurtin2> could tie in pexpect to type in the password to sign the bins :)
[13:49] <ralsina> briancurtin2: and upload your private key to ec2? You are brave :-)
[13:50] <dobey> ok, now what order to review these branches in
[13:51] <briancurtin2> oh windows update, thanks for rebooting as i'm uploading this zip file...
[13:51] <mandel> dobey, first lp:~mandel/ubuntuone-dev-tools/tcp-testcases and then lp:~mandel/ubuntuone-dev-tools/mocked-webserver :)
[13:53] <ralsina> briancurtin2: ugh
[13:57] <ralsina> alecu: we did our 1-1 the other day, so skipping
[13:58] <alecu> ralsina, right
[13:59] <briancurtin2> ralsina: do you have a preferred date for update.xml to go live, or do we wait to tell them that until we have the full fledged installer built?
[14:00] <ralsina> let's say two days after we push the binaries to the site, so we get some user feedback
[14:03] <briancurtin2> ok. i'll wait to give them the update.xml until we have the installer at the same time
[14:04] <joshuahoover> ralsina: so the new installer is out there? :)
[14:05] <ralsina> joshuahoover: it's tryig to escape
[14:05] <joshuahoover> heh
[14:07] <dobey> mandel: several needs fixing/comments on the tcp-testcases which need clarified
[14:07] <dobey> joshuahoover: no, but the truth is
[14:08] <mandel> dobey, sure, on it!
[14:14] <mandel> dobey, so to clarify, there is a Pb test case which is indeed specific to our tests but the SaveServiceRunner (which should be called Safe) is a general way to execute a service that listens to a a tcp port and connect a client to use such a service
[14:15] <mandel> dobey, the class will clean all resources when the clean_up calls is adde via a tearDown method or a addCleanup call
[14:16] <mandel> dobey, I used pb as an example in the tests but any other example listening to a tcp would have worked, regarding the header, added and confusion between the verb save and the adj safe will be fixed
[14:17] <dobey> so calling it anything like "ServiceRunner" i think is a problem
[14:17] <mandel> dobey, ok, what other name does it sound valid?
[14:17] <dobey> because we already have this concept of a service runner in devtools
[14:17] <dobey> i don't know
[14:18] <mandel> dobey, the class can run more than one, and I'm terrible with names.. my dog is called Iron because he was wrinkles.. :P
[14:18] <dobey> i'm just concerned about the naming conflict :)
[14:18] <mandel> dobey, mkes perfect sense to worry about it, I completely forgot about that..
[14:19] <mandel> dobey, TcpFactoryManager maybe?
[14:20] <dobey> maybe SafeTCPServer then? and naming the module tcpserver or something? and change TwistedServiceTestCase or whatever you called it to TCPServerTestCase?
[14:20] <dobey> and similar related renamings throughout?
[14:23] <mandel> dobey, yep, sounds good, I'll leave the Pb one just becaue is used everywhere, naming it tx.py is ok then? I added the twisted Webserver there because of that naming convention
[14:27] <dobey> mandel: i'd prefer things were named specifically to their purpose. having generic modules like "tx" and such leads to wanting to put everything remotely related, in the same file, which means we ned up with all test cases in a single file, which kind of sucks; which is why we refactored a bit when adding the squid test case :)
[14:29] <mandel> dobey, ok, I'll do a tx_tcp_server.py and a tx_webserver.py ones
[14:30] <mandel> dobey, I want to be very explicit that the are tx related of course
[14:30] <dobey> s/_//g please :)
[14:31] <dobey> gatox donde es?
[14:31] <mandel> dobey, al joined, sure I have no problems
[14:31] <mandel> dobey, he is playing halo or something like that
[14:31] <mandel> aka holidays
[14:32] <dobey> doh
[14:32] <dobey> https://bugs.launchpad.net/ubuntu/+source/ubuntuone-client/+bug/979759
[14:33] <mandel> dobey, yep, rye told me about that, I think is based on the very old windows release, should be ok in trunk
[14:33] <mandel> dobey, but do not take my word for it
[14:33] <dobey> ok
[14:34] <dobey> well, i'll assign it to diego :)
[14:34] <mandel> dobey, best thing to do
[14:34] <mandel> some bastard added my canonical mail to http://www.lelong.com.my/
[14:34] <mandel> wtf!
[14:35] <dobey> haha
[14:36] <ralsina> dobey: that's an old bug, has been fixed on trunk for a while, but we never got to do a release of it
[14:36] <ralsina> dobey: until today!
[14:36] <dobey> yay
[14:55] <ralsina> dobey, alecu, thisfred, mandel, briancurtin2 (and Harry!), team call in 5'
[14:55] <ralsina> urbanape and gatox are off today
[14:56] <alecu> ralsina, has Harry already started?
[14:56] <ralsina> alecu: no, Harry is imaginary.
[14:57] <ralsina> alecu: because when I asked if I forgot someone, thisfred said I forgot Harry
[14:57] <ralsina> So he's like Harryj
[14:57] <alecu> ralsina, but he has already started, right?
[14:57] <ralsina> alecu: in my imagination, yes :-)
[14:57] <alecu> awesome!
[15:00] <thisfred> :)
[15:00] <ralsina> so mumble time!
[15:00] <briancurtin2> logging in
[15:00] <alecu> mandel, probably someone that thinks you might enjoy buying fancy imitation lingerie
[15:01] <mandel> alecu, someone in the internet really know me, I get fake watches, russian wifes and facy imitation lingerie :P
[15:07] <thisfred> you'll look really fancy imitation sexy!
[15:28] <mterry> I tried the "Send to Ubuntu One" thing and got a 403 Forbidden message.  Is that a known issue?
[15:29] <beuno> aquarius, ^
[15:30] <aquarius> mterry, no, it isn't. Does it happen for all send buttons or just one, and does it happen every time or just some times? (Refresh the page to get the button back...)
[15:31] <mterry> aquarius, first time I went to the page, I've tried two and it happened on both.  Will try refreshing
[15:31] <mterry> mmm
[15:31] <mterry> just tried a third before refreshing and it worked...
[15:32] <mandel> dobey, updated: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/tcp-testcases/+merge/99759
[15:32] <mterry> bradbury, heinlein, and vonnegut failed.  let me refresh
[15:32] <mterry> after a refresh, all three worked
[15:32] <mterry> aquarius, ^
[15:32] <aquarius> ok
[15:32] <aquarius> so, that's annoying, then
[15:32] <mterry> (but the others worked the first time, besides those three)
[15:33] <aquarius> I can't decide whether it's better to have an intermittent failure or better to have it always fail :)
[15:33] <aquarius> mterry, thanks. I'll try and replicate the problem...
[15:33] <mterry> Specifically, what happened, was the box got replaced by a text field, that said "Forbidden (403)". The text was too large for the space and there was a horizontal scrollbar to see the rest
[15:33] <mandel> dobey, AFAIK there is no use of openssl in devtools so the headers should be ok, right?
[15:33] <mterry> Not a text field, but a label
[15:33] <aquarius> heh. they all work for me. :(
[15:33] <mterry> yay....  :-/
[15:34] <aquarius> mterry, this is useful feedback, though!
[15:34] <dobey> mandel: it needs to have the exception though
[15:34] <mandel> dobey, is there an example header I can use?
[15:35] <dobey> mandel: as i said in the comment. all the other modules in trunk have it now, and there is the LICENSE file. :)
[15:36] <mandel> dobey, all the others? there are a couple missing, or at least the one I exactly opened :P
[15:38] <dobey> eh?
[15:38] <dobey> ok, well. i need to go get lunch
[15:40] <dobey> so, bbiab
[15:45] <mandel> briancurtin2, https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-webclient-tests/+merge/101410
[15:50] <duanedesign> On XP is the syncdaemon metadata folder in C:\Users\yourusername\AppData\Local\xdg\ubuntuone ?
[15:51] <briancurtin2> duanedesign: it'll be different on XP, i'll boot up my VM and see
[15:51] <ralsina> duanedesign: %localappdata% should get you close to it
[15:51] <duanedesign> thank you very much brian
[16:00] <briancurtin2> duanedesign: that same folder on XP would be C:\Documents and Settings\brian\Local Settings\Application Data\xdg\ubuntuone
[16:01] <duanedesign> thank you!
[16:30] <mandel> briancurtin2, FYI I had to update all the branches to use the new paths in  ubuntuone-dev-tools, all the changes are already in the lp MP
[16:30] <mandel> briancurtin2, sorry for any PITA :)
[16:30] <briancurtin2> mandel: no problem, i'll take a look
[16:41] <dobey> back
[16:45] <mandel> dobey, I need to go, I have update all the branches, let me know of any issues and I'll fix it after the rugby training
[16:46] <dobey> mandel: ok. thanks
[16:49] <mandel> ok, rugby time after taking pain killers, see you tom!
[17:21] <nemo> aquarius: So, I showed my mom your site.  She's grateful
[17:21] <nemo> aquarius: but she's decided to wean herself off of ubuntu one and move more of her notes to Google Docs
[17:21] <nemo> I did emphasise if she wanted to keep using tomboy, I'd be ok w/ working to make it more functional
[17:21] <nemo> aquarius: but anyway, it is a good interim measure. so I appreciate it being there. thanks.
[17:34] <aquarius> nemo, thanks for letting me know
[17:40] <ralsina> Looks like I am finally coming down with the cold that has been hovering around me for weeks
[17:40] <ralsina> I have 39C fever, so am going to lie down and call a doc
[17:40] <briancurtin2> jeez, get some rest
[17:40] <ralsina> if there is an emergency that can be fixed by a guy who is half delirious, phone me ;-)
[18:23] <dobey> hrmm, i need to get some pilldws for the new office futon
[19:21] <dobey> alecu: ping
[19:22] <dobey> alecu: were you looking at bug #943046 ? it has a *lot* of duplicates :-/
[19:26] <briancurtin2> ah crap...for some reason, when quitting the u1cp from the tray the log file can't be opened...
[19:28] <briancurtin2> wait, ughhh. why is it trying to write a log file at """C:\Program Files (x86)\ubuntuone\dist\ubuntuone-control-panel-qt.exe.log"""
[19:33] <ralsina> briancurtin2: you don't have the patched py2exe!
[19:33] <ralsina> briancurtin2: you need to change a file, let me find the way to do it
[19:34] <ralsina> briancurtin2: in Python27/Lib/site-packages/py2exe/boot_common.py
[19:34] <ralsina> you need to add sys.stderr = backhole (or close to that, just like for stdout)
[19:35] <ralsina> briancurtin2: bad news, that may require a re-signing of binaries, check if it changes the .exes :-/
[19:35] <briancurtin2> shit.
[19:35] <ralsina> briancurtin2: yep. Sorry, should have remembered to tell you
[19:35] <ralsina> briancurtin2: or if I did, should have remembered to remind you
[19:35] <briancurtin2> nah its my fault
[19:36] <briancurtin2> this happened because re-setting up the buildout which was messing with, i should have done it on another checkout, not the live one i'm actually using. when i rebuilt the buildout, it re-ordered the sys.path so my local C:\python27\lib\site-packages comes before the patched py2exe
[19:37] <briancurtin2> i cant imagine why buildout would ever do that, but now i know...
[19:42] <alecu> dobey, pong.
[19:43] <alecu> dobey, ugh, yes, I saw that bug. I was unable to reproduce it in the past. Will try again in a clean vm.
[19:44] <dobey> alecu: thanks. the last duplicate seems to have been filed on tuesday.
[19:45] <alecu> ralsina, that "sys.stderr = backhole" line seems to lead to a lot of misinterpretations.
[19:46] <ralsina> alecu: there is also a bug about never ever ever writing to stderr wich is a better fix.
[19:46] <alecu> ralsina, I meant because of the "*back*hole"
[19:46] <ralsina> alecu: ooooooook
[19:48] <dobey> lol
[19:48]  * dobey wasn't going to say anything
[19:48] <dobey> freud said enough on the subject already
[20:00] <joshuahoover> ralsina, briancurtin2: still working on the windows release i gather? :)
[20:01] <briancurtin2> joshuahoover: yeah, i screwed up the one i sent to be signed, so i'm making another right now
[20:01] <joshuahoover> briancurtin2: ralsina is sharing all his "tricks" with you i see ;)
[20:13] <dobey> ralsina: speaking of winodws installer tricks… is there any trick to say, have it download/install a thing from MS, on XP, but not on Vista/7?
[20:13] <dobey> (or rather, if said thing from MS isn't already installed)
[20:15] <ralsina> dobey: not that I know of. briancurtin2 probably knows better
[20:15] <ralsina> dobey: I mean, installbuilder has support to detect windows version and to download stuff (I think) so maybe it can do it manually
[20:16] <briancurtin2> dobey: i dont know if installbuilder does downloads, but we could always download what's needed and conditionally install it, like we do for those firewall exceptions
[20:27] <dobey> hrmm
[20:28] <dobey> briancurtin2: so "we could always download" means download from within the installer, or we distribute it with the installer?
[20:29] <briancurtin2> dobey: i meant manually or as a part of creating the installer. what we ship to customers would already have it in the installer, and only install if needed
[20:30] <dobey> ah. i suppose that is possible, assuming it's legally allowed from the license
[20:32] <dobey> ralsina: also, will we be shipping u1db for windows as part of the same installer we currently have (when we ship it), or as a separate install? or both?
[20:33] <ralsina> dobey: have not thought about it at all
[20:33] <ralsina> dobey: u1db is a library, so normally all apps that use it on windows would include them
[20:34] <dobey> aside from the 2 scripts it also installs, but i suppose so
[20:34] <dobey> hmm
[20:38] <dobey> i'm thinking about what the cross-platform build system for it will be.
[20:38] <dobey> the best, that is
[20:44] <briancurtin2> ralsina: so i got the log error to go away from py2exe...but now i'm seeing the SSO cred login screen pop under the signup page. is there yet another thing i screwed up that causes this?
[20:44] <ralsina> briancurtin2: I don't know
[20:44] <ralsina> briancurtin2: I don't think so
[20:45] <ralsina> briancurtin2: I think this is because we removed syncdaemon from autostart
[20:45] <briancurtin2> i dont think so either, but it never used to do that
[20:45] <ralsina> or something related
[20:45] <ralsina> briancurtin2: check the process tree, I bet sso is not started by u1cp
[20:45] <ralsina> briancurtin2: if you had sso running from a previous install, that explains it
[20:46] <ralsina> briancurtin2: that happened to elopio on testing a while ago
[20:46] <briancurtin2> ralsina: yep, that's it. i had everything running, removed creds, SSO remained running, went to login again and now it's behind
[20:46] <ralsina> briancurtin2: cool, kill it and try again, should be in front :-)
[20:51] <briancurtin2> ralsina: ok, now it works. RT updated
[20:52] <ralsina> briancurtin2: awesome. We'll get this out yet!
[20:52] <ralsina> briancurtin2: also, it's 12% easier the 2nd time
[20:53] <briancurtin2> ralsina: it's also easier if i don't mess with the buildout that produced working packages beforehand :) that is now circled in my notes
[21:42] <dobey> later all!