| === tgall__ is now known as Dr_Who | ||
| JamesTait | Happy release day, folks! | 08:32 |
|---|---|---|
| mandel | mornig! | 08:37 |
| mandel | JamesTait, is the day where I follow all the #ubuntu tags ;) | 08:37 |
| JamesTait | mandel: You mean you don't normally? ;) | 08:39 |
| mandel | JamesTait, waaaay too much noise, I follow just #ubuntuone :P | 08:40 |
| JamesTait | mandel: Pah, some people just have no commitment. :-P | 08:40 |
| mandel | je | 08:40 |
| mandel | or he in eng | 08:41 |
| gatox | good morning! | 11:04 |
| mandel | gatox, morning! | 11:07 |
| mandel | gatox, I tried to do the review, but I found an issue on my mac | 11:07 |
| gatox | mandel, how are you? | 11:08 |
| gatox | mandel, crap | 11:08 |
| mandel | gatox, fine thx! shit weather here :P | 11:08 |
| gatox | mandel, here is really cold.... so i'm happy :D | 11:08 |
| mandel | gatox, also, I finished the following: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/domain-sockets/+merge/103660 we know will be able to run all the tests over tcp for windows and over domain sockets on mac \o/ | 11:09 |
| gatox | mandel, awesome!! | 11:09 |
| mandel | gatox, I'm ok with cold, not that much with rain :( | 11:09 |
| mandel | gatox, can you do a review of that branch running it on your evil vms? | 11:09 |
| gatox | mandel, of course | 11:09 |
| mandel | gatox, if that is ok, I'll have to change all our project because twisted.internet.endpoints is a little diff that reactor.connectTCP and I had to change the place of the deferreds | 11:10 |
| mandel | gatox, not a lot of work for what it will bring, free domain sockets for all! | 11:10 |
| mandel | hehe | 11:10 |
| gatox | mandel, bad news: http://paste.ubuntu.com/947157/ | 11:18 |
| mandel | gatox, is that on windows? | 11:19 |
| gatox | yes | 11:19 |
| gatox | mandel, and in linux i get this lint issues: | 11:19 |
| gatox | bin/u1lint: | 11:19 |
| gatox | 258: [W0511] XXX Testing that W0511 does not cause a failure | 11:19 |
| gatox | 258: [W0511] XXX Testing that W0511 does not cause a failure | 11:19 |
| mandel | gatox, that is correct, is a test to see that it does not fail with TODO :) | 11:19 |
| gatox | ok | 11:20 |
| mandel | gatox, and the error on windows seem to be because I forgot to add a skipIfOS(win32) and is trying to use domain sockets on windows, which breaks :) | 11:20 |
| mandel | gatox, easy fix, on it! | 11:20 |
| gatox | mandel, great | 11:20 |
| mandel | gatox, feel free to pull it should be fixed by now | 11:24 |
| * gatox testing.... | 11:25 | |
| mandel | gatox, I also renamed the module to txsocketserver since it supports more than tcp servers | 11:26 |
| gatox | mandel, ALL GREEN! | 11:27 |
| mandel | gatox, hurray! :) | 11:28 |
| mandel | gatox, can you test lp:~mandel/ubuntu-sso-client/use-new-dev-tools in your vm adding the new dev tools to the path? | 11:36 |
| gatox | ok | 11:39 |
| mandel | gatox, I'm noticed that we use mocker for the windows ipc tests on windows, I think we should add as a TODO to remove mocker and the tests I added to devtools and test it over tcp on windows and over domain sockets on mac, what do you think? | 11:40 |
| * mandel needs to update the bugs list | 11:41 | |
| gatox | mandel, yes....+1 to remove mocker | 11:41 |
| mandel | gatox, also, because although they will stop being unit tests perse, we need to test ipc better | 11:42 |
| gatox | @ping | 11:54 |
| gatox | mandel, ping | 11:55 |
| gatox | mandel, do you get my messages? | 11:55 |
| gatox | mandel, some failures on windows: http://paste.ubuntu.com/947192/ | 11:56 |
| mandel | gatox, I was not, now I do | 11:56 |
| mandel | gatox, let me see | 11:56 |
| mandel | gatox, oh, true i did not update the tcpactivation tests.. sorry I cannot run the tests on my vm, I have some small problems with it.. | 11:57 |
| mandel | :( | 11:57 |
| ralsina | good morning | 11:58 |
| gatox | ralsina, good morning | 11:58 |
| mandel | ralsina, morning! | 12:05 |
| ralsina | good morning mandel | 12:05 |
| === yofel_ is now known as yofel | ||
| gatox | brb... need to restart the router | 12:06 |
| mandel | gatox, hm.. funny there is no reason why the tcpactivation tests should fail.. | 12:06 |
| * mandel hates dirty reactors.. | 12:06 | |
| mandel | gatox_mac, looks like a timepit error, can you run it again, if it happens, can you go to the testcase and change timeout to, lets say 5 | 12:09 |
| mandel | gatox, looks like a timeout error, can you run it again, if it happens, can you go to the testcase and change timeout to, lets say 5 | 12:10 |
| gatox | back | 12:10 |
| gatox | mandel, ack | 12:10 |
| mandel | gatox, thx, if it continues failing, please try it without the new dev-tools | 12:10 |
| gatox | mandel, where do i change to 5? | 12:13 |
| ralsina | Argh, have to run an errand, will be back before team call | 12:14 |
| mandel | gatox, in the testcase of the error, just add a class level attribute which whould be timeout=5 | 12:15 |
| alecu | buenos días #ubuntuone! | 12:16 |
| gatox | alecu, buensa | 12:16 |
| mandel | alecu, hola! I;m interested in a review from you here: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/domain-sockets/+merge/103660 | 12:17 |
| alecu | a very good day to be just that: a developer making software. | 12:17 |
| mandel | alecu, more twisted :) | 12:17 |
| alecu | mandel, sure | 12:17 |
| mandel | alecu, thx! | 12:17 |
| gatox | mandel, with 5 it pass | 12:18 |
| mandel | gatox, hm.. you machine is slow! | 12:18 |
| mandel | gatox, can you try running trunk? | 12:18 |
| gatox | ok | 12:18 |
| mandel | gatox, no new devtool in the path, I hope I did not make the test slower | 12:18 |
| gatox | mandel, for trunk is all green | 12:20 |
| mandel | hm.. that is annoying, I did not touch tcpactivation tests | 12:20 |
| mandel | gatox, can you run them a couple of times, worst case we increase the timeout a few secs | 12:20 |
| gatox | mandel, run a couple of times your branch or trunk? | 12:21 |
| mandel | gatox, mine, please | 12:21 |
| gatox | happy release day everyone! :P | 12:21 |
| gatox | mandel, with or without the timeout=5? | 12:21 |
| mandel | gatox, without | 12:22 |
| mandel | gatox, and today it should be holidays ;) | 12:22 |
| gatox | mandel, i already run it without that and it fails always | 12:22 |
| gatox | a couple of times i mean | 12:22 |
| mandel | gatox, then add 5 and see if it always passes | 12:22 |
| gatox | ok | 12:22 |
| alecu | gatox, mandel: which branch needs a timeout=5? | 12:25 |
| alecu | mandel, the one I'm reviewing? | 12:25 |
| gatox | mandel, with timeout=5 it's always ok | 12:25 |
| mandel | alecu, no, sso + the one you are looking at on tcpactivation | 12:25 |
| alecu | mandel, great. | 12:25 |
| gatox | alecu, use-new-dev-tools | 12:25 |
| alecu | gatox: do you need the timeout=5 in osx? | 12:28 |
| gatox | alecu, no, windows | 12:28 |
| alecu | gatox, I should start calling you gatOSX | 12:28 |
| gatox | alecu, jejejejje | 12:28 |
| mandel | alecu, gatox, the deal, I did not need to change the test from tcpactivation.. | 12:29 |
| mandel | I'm off to have lunch, will be back in a few mins and will continue with tcp over domain sockets using that dev tools branch | 12:31 |
| mandel | alecu, my idea is to have domain sockets and tcp tests running both on linux and mac os x and just change the type of server used | 12:32 |
| mandel | alecu, which just means diff transport | 12:32 |
| alecu | mandel, great. | 12:32 |
| gatox | ahh f$%&ing internet provider! | 12:32 |
| alecu | mandel, ping | 12:37 |
| alecu | mandel, client_endpoint_patter = 'tcp:host=127.0.0.1:port=%s' | 12:37 |
| alecu | mandel, why %s there? | 12:37 |
| * alecu realizes that print("%s" % 1) works just fine. | 12:39 | |
| === gatox_ is now known as gatox_mac | ||
| alecu | it's ugly though :) | 12:39 |
| alecu | works even on py3k | 12:39 |
| alecu | mandel, btw: https://bugreports.qt-project.org/browse/QTBUG-23625 | 12:51 |
| alecu | mandel, that might be the bug we are experiencing... | 12:52 |
| alecu | well... not completely sure. | 12:52 |
| thisfred | alecu: did you hear Noel Fielding has a new comedy show? https://en.wikipedia.org/wiki/Noel_Fielding%27s_Luxury_Comedy | 12:57 |
| alecu | thisfred, yes! I've even watched a few episodes already, but I miss Barratt's music :-) | 13:03 |
| thisfred | I only saw a clip, but it looked very similar to Boosh, which can't be all bad. There was also a movie they were involved in, which looked great but wasn't very funny. | 13:05 |
| thisfred | I forget the name | 13:05 |
| alecu | thisfred, Bunny and the Bull ? I kinda liked that. | 13:06 |
| thisfred | I'm gonna rewatch all of Mighty Boosh this weekend I think. It's pretty cheap on amazon | 13:06 |
| thisfred | alecu: yeah that one. I quit half way through. There was some funny stuff in there, but somehow I was missing something | 13:07 |
| alecu | thisfred, btw: I've forced my brother to watch the first episode of "Garth Marenghi's Darkplace" and he loves it. He still won't watch a whole Boosh episode. | 13:08 |
| thisfred | I've never actually seen that either | 13:08 |
| * thisfred puts on list | 13:08 | |
| alecu | thisfred, you can sample it on youtube; most episodes are there. | 13:09 |
| thisfred | cool | 13:09 |
| alecu | thisfred, hope I'm not ruining u1db's future: http://www.youtube.com/watch?v=nNfQ0ORwSDM | 13:10 |
| thisfred | alecu: hehe | 13:11 |
| thisfred | I'll save it for tonight | 13:11 |
| === m_conley_away is now known as m_conley | ||
| mandel | alecu, let me take a look | 13:41 |
| alecu | mandel, btw: I've reviewed your branch. | 13:41 |
| mandel | alecu, ok, thx! | 13:41 |
| mandel | alecu, Why are the UNIX client and server endpoint strings different? -> bloody twisted! same happens with the tcp ones, client and server endpoints use diff descriptions | 13:43 |
| alecu | mandel, just awesome :-( | 13:43 |
| mandel | alecu, yeah.. I had to read the code to find out because is not documented O_o | 13:43 |
| mandel | alecu, will fix the other comments in a few mins (reading the bug description) | 13:44 |
| mandel | dobey, may I have a re-review for: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/fix-squid-tests/+merge/103325 | 13:44 |
| mandel | alecu, in that qt bug reported, should we mention we are seeing similar issues? | 13:45 |
| mandel | gatox, timeout=5 fixes the tests on windows, right? | 13:45 |
| gatox | mandel, yes | 13:46 |
| mandel | gatox, which testcase was it? | 13:46 |
| dobey | mandel: but why can't we do the "if sys.platform" check inside the functions, instead of creating more imports? | 13:46 |
| gatox | mandel, test_timestamp | 13:47 |
| mandel | dobey, I find it dirtier, if we can isolate platform bits I prefer to do it so | 13:47 |
| mandel | dobey, if sys.platform is like ifdef in a way.. but I can merge them if you give me a good reason :) | 13:47 |
| mandel | gatox, thx! | 13:48 |
| mandel | gatox, can you push the branch, that way I'll just have to pull from it :) | 13:49 |
| mandel | gatox, I don't want to screw it up hehe | 13:49 |
| gatox | mandel, ok | 13:49 |
| dobey | mandel: the *only* platform-specific code there really though, is the kill_squid() function. The \\ escaping is probably something we also need on linux (in the event the path has a \ in it) | 13:51 |
| mandel | dobey, true that, it might be the case that the path has a \ on linux.. (stupid squid) | 13:51 |
| mandel | dobey, and the auth program location? that is platform specific | 13:52 |
| gatox | mandel, lp:~diegosarmentero/+junk/sso-timestamp | 13:53 |
| mandel | gatox, thx | 13:53 |
| dobey | true, as are the squid args, but i don't think we should have separate platform-specific imports for it | 13:53 |
| mandel | dobey, hm.. lets put them together but I will leave the tests like they are | 13:54 |
| dobey | mandel: also, the diff still shows the same spelling/grammar errors :) | 13:55 |
| mandel | dobey, really? let me double check | 13:55 |
| dobey | +# Do use doble \ because squids config needs \ to be escaped | 13:55 |
| dobey | is one example | 13:56 |
| dobey | 330+# -X make it very verbose which makes things crash when on windows and | 13:56 |
| dobey | 331+# capturing stdout :( | 13:56 |
| gatox | brb...... need to relogin | 14:02 |
| mandel | dobey, ok, got it | 14:03 |
| mandel | alecu, updated: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/domain-sockets/+merge/103660 | 14:06 |
| alecu | mandel, ack | 14:06 |
| mandel | gatox, updates: https://code.launchpad.net/~mandel/ubuntu-sso-client/use-new-dev-tools/+merge/103673 can you please add your review stating that you always failing machine passes all the tests? | 14:08 |
| gatox | mandel, looking..... | 14:08 |
| mandel | gatox, thx | 14:09 |
| gatox | mandel, we'll definitely need to refactor platform before runnning the tests for u1-client..... because it keeps trying to take the mmodules from the wrong pplaces in mac | 14:10 |
| mandel | gatox, sounds like a reasonable thing to work on | 14:11 |
| mandel | gatox, what about you, mmcc and me chat have a chat after the team meeting and start adding bugs to the diff projects? | 14:12 |
| mandel | gatox, I can jot it down with you guys and I'll deal with the paperwork (launchpad) | 14:12 |
| gatox | mandel, ok | 14:12 |
| gatox | mandel, i think the three of us could make a sprint in spain :P (/me wants to travel jejee) | 14:13 |
| mandel | gatox, pleople is more than welcome, ask ralsina ;) | 14:13 |
| gatox | mandel, mmmmmm but maybe they won't let me in to spain right now jeje | 14:14 |
| gatox | mandel, me being argentinian and all that :P | 14:14 |
| mandel | gatox, we don't hate argentinias, we really don't care about YPF | 14:16 |
| mandel | gatox, but apparently argentinias do ;) | 14:16 |
| dobey | briancurtin, ralsina: care to do a second review of https://code.launchpad.net/~dobey/ubuntuone-dev-tools/gi-glib-fallback/+merge/103561 ? | 14:27 |
| mandel | dobey, what about the kill_squid method, what shall I do with that one? | 14:28 |
| briancurtin | dobey: sure thing, looking | 14:28 |
| gatox | brb....... need to run some tests without internet connection | 14:29 |
| dobey | mandel: i'd put the win-specific bits inside an if sys.platform == 'win32': block, and do the normal thing inside the else:, or outside the if and just return at end of if block for win32 | 14:29 |
| ralsina | dobey: will do | 14:29 |
| mandel | dobey, sure | 14:29 |
| ralsina | mandel, gatox: you know, sprints need a purpose that is not 'tapas!' | 14:29 |
| gatox_mac | not need to brb… i also has this machine :P | 14:30 |
| mandel | ralsina, is 'tapas with manuel' which is a diff experience | 14:30 |
| dobey | oh man, i want some tapas | 14:33 |
| mandel | dobey, updated https://code.launchpad.net/~mandel/ubuntuone-dev-tools/fix-squid-tests/+merge/103325 | 14:55 |
| ralsina | mandel, mmcc, gatox_mac, dobey,alecu, thisfred, joshuahoover: mumble in 4' | 14:56 |
| thisfred | ack | 14:56 |
| dobey | mandel: ok. i think at this point i'll have to look at it after meeting and lunch | 14:56 |
| ralsina | urbanape: welcome to join for a last call if you have something to say | 14:56 |
| ralsina | urbanape: or just to say hi, of course ;-) | 14:56 |
| mmcc | ralsina, ack. | 14:56 |
| urbanape | howdy, folks! | 14:56 |
| urbanape | omw | 14:56 |
| dobey | hi urbanape | 14:57 |
| mandel | dobey, sure, no problem :) | 14:57 |
| dobey | gatox: btw, did you see my comment on your proposal? i thought we had agreed it was an environment problem and not something we should hack around in u1trial | 14:57 |
| dobey | err, in u1lint even | 14:57 |
| gatox | dobey, no | 14:57 |
| mandel | dobey, is that the mac-port branch? I had issues with that in my system.. | 14:58 |
| gatox | mandel, with what? | 14:58 |
| mandel | gatox, with python_u1: command not found | 14:58 |
| mandel | gatox, which has no meaning for me by the way.. | 14:58 |
| gatox | mandel, did you execute env-mac script? | 14:59 |
| dobey | mandel: the python_u1 yeah | 14:59 |
| mandel | gatox, no, where is that? | 14:59 |
| gatox | mandel, in u1-windows-installer/scripts/devsetup (is the other branch i propose) | 15:00 |
| gatox | dobey, which is the suggestion?? to not run u1lint? | 15:01 |
| alecu | gatox, ralsina: mumble? | 15:01 |
| mandel | gatox, but what does python_u1 do? | 15:02 |
| dobey | gatox: i don't think we run u1lint on windows. but i think the problem is an environment issue. having a special python_u1 seems totally broken to me | 15:02 |
| dobey | gatox: the buildout should really be as close to the actual end-user version as possible | 15:03 |
| gatox | mandel, executes the python from the buildout, not the one from the system.... and i'm adding the bin folder to the PATH.... so i change the name of the buildout python to avoid name crashes, and don't affect the user system | 15:03 |
| gatox | dobey, ^ | 15:03 |
| mandel | gatox, and why not calling python using the full path of python? | 15:05 |
| gatox | mandel, what? | 15:05 |
| mandel | gatox, so, if we have python in ./bin/python, just call ./bin/python, right? | 15:06 |
| gatox | mandel, we have a different python inside the bin foldeer in the buildout | 15:06 |
| gatox | not in the system | 15:07 |
| mandel | gatox, je ne compre pa | 15:07 |
| mandel | gatox, what do you mean? | 15:07 |
| gatox | mandel, mumble after this one | 15:07 |
| gatox | :P | 15:07 |
| mandel | gatox, sure | 15:07 |
| === caravone is now known as caravone123 | ||
| dobey | alecu: when do you arrive to UDS? | 15:22 |
| alecu | dobey, I'm staying in SF on Sat night, to walk around town. | 15:23 |
| alecu | dobey, when are you arriving? | 15:23 |
| dobey | alecu: late Sat night | 15:24 |
| alecu | dobey, do you staying in Oakland that night? | 15:25 |
| alecu | *stay | 15:25 |
| dobey | alecu: yeah, that's the plan. my plane lands at like 23:00 | 15:26 |
| dobey | alecu: so i'll be at the hotel :) | 15:26 |
| * mmcc is reconnecting to mumble | 15:32 | |
| mmcc | gatox, mandel - will you be mumbling about mac port stuff next? I'd like to join… | 15:42 |
| mandel | mmcc, we are.. ralsina. gatox switch to eng! | 15:42 |
| mandel | mmcc, feel free to join please! | 15:42 |
| mmcc | mandel, ok. be right there. is it a different room/channel? | 15:43 |
| mmcc | need ~2min break first, but don't wait for me | 15:43 |
| ralsina | mmcc same place! | 15:43 |
| dobey | ok, am off to lunch here. bbiab | 15:46 |
| gatox | mmcc, so, i'll have lunch now, and after that i'll review the platform modules so we can split the tasks | 15:56 |
| roasted | hello! | 15:56 |
| roasted | Question - Does Ubuntu One delete files at all? Reason I ask is, let's say I have Ubuntu One set up on my Android to sync my pictures to Ub1. If I delete pictures on my phone, would that "delete" change filter upt hrough Ubuntu One and to my computers? Or does it not delete anything? | 15:56 |
| mmcc | gatox: ok, have a good lunch. I did some poking around there yesterday - I'll send you an email with a summary | 15:57 |
| gatox | mmcc, ack | 15:58 |
| mandel | ralsina, briancurtin, will this be old enough: http://www.lfd.uci.edu/~gohlke/pythonlibs/ | 15:58 |
| mandel | ralsina, briancurtin, under pyqt, they have version 4.8.6-1 which I think it has qt 4.7 | 15:59 |
| briancurtin | mandel: i'll see if i can verify that. if it's a good version to use, i can create an installer fairly easily | 16:00 |
| mandel | briancurtin, sweet, let me know, I found the link here: http://python.6.n6.nabble.com/Couldn-t-find-PyQt-4-8-x-with-Qt-4-7-to-download-td4681517.html and the are asking for exactly what we need :) | 16:00 |
| === salgado is now known as salgado-lunch | ||
| === gatox is now known as gatox_lunch | ||
| mmcc | gatox, mandel - assuming I typed your email addrs right, look for an update when you get back. | 16:31 |
| === mmcc is now known as mmcc_lunch | ||
| briancurtin | mandel: so for this qt 4.7 build, what exactly do i need to package. is this just stable 3.0 with qt 4.7, or do i need to do include any branches with it (maybe your logging one?) | 16:38 |
| === salgado-lunch is now known as salgado | ||
| dobey | hmm | 16:59 |
| === gatox_lunch is now known as gatox | ||
| gatox | mandel, we can have the mumble whenever you want | 17:10 |
| alecu | briancurtin, I think mandel's branch got some fixes, and then got merged into trunk | 17:16 |
| alecu | briancurtin, but you are releasing from stable-3, right? | 17:16 |
| briancurtin | ah, i didn't see that | 17:16 |
| briancurtin | alecu: yes, stable-3_0_0 tag | 17:16 |
| alecu | briancurtin, then I think including the branch by mandel with the fixes is a good idea. | 17:17 |
| === mmcc_lunch is now known as mmcc | ||
| alecu | briancurtin, and we probably want to backport that branch to stable-3 | 17:17 |
| dobey | alecu: hi. for the proxy support that reads gsettings, do we have any plans to allow it to work without gsettings? | 17:18 |
| dobey | alecu: for example, on lucid or natty | 17:19 |
| alecu | dobey, I have no plans for that, yet. | 17:26 |
| alecu | dobey, I suspect just replacing the call to gsettings with whatever worked then. | 17:27 |
| * mandel back | 17:27 | |
| gatox | mmcc, do you have time for a quick mumble? | 17:27 |
| mmcc | gatox: yes | 17:27 |
| dobey | alecu: well the settings in gconf are a bit different | 17:27 |
| mandel | briancurtin, alecu, the bracnh with the logging is already present in trnk | 17:27 |
| dobey | alecu: is the code modular enough to just have a separate parsing module for gconf? | 17:27 |
| * gatox connecting mumble | 17:27 | |
| alecu | mandel, right. But briancurtin releases from stable-3 | 17:28 |
| briancurtin | mandel: i need to release from release-3_0_0 tag so i'll need to include your branch | 17:28 |
| alecu | dobey, probably not. Looking. | 17:28 |
| * mmcc is having mumble problems | 17:28 | |
| mandel | briancurtin, ok, let me find that for you | 17:28 |
| gatox | mmcc, well....... we can do it here | 17:28 |
| mmcc | gatox: let me try one last time… mumble app keeps crashing | 17:29 |
| mandel | briancurtin, alecu, the branch with the ssl fixes is: lp:~mandel/ubuntu-sso-client/fix-ssl | 17:29 |
| gatox | mmcc, move to manou-ah-manou | 17:29 |
| mandel | gatox, mumble? | 17:29 |
| gatox | mandel, you too? | 17:30 |
| mandel | gatox, launching it | 17:30 |
| gatox | mandel, wait, i'm going to have one with mmcc firsst | 17:30 |
| mandel | gatox, mmcc, do you mind if we don mine first, its 7:30 pm and I have rugby training, please, please :) | 17:30 |
| === JamesMR1 is now known as JamesMR | ||
| alecu | dobey, the code seems modular enough. The entrypoint is ubuntu_sso.utils.getclient.gsettings.get_proxy_settings() | 17:33 |
| gatox | mandel, jump into mumble | 17:33 |
| alecu | dobey, and it returns a python dict of proxy settings for the different protocols (http, https, socks) | 17:33 |
| dobey | alecu: ok, maybe we can do something, though it'll be difficult to pick the right one | 17:33 |
| dobey | at least, will be difficult on natty | 17:33 |
| dobey | where "gsettings" is available, but the proxy settings aren't stored there | 17:34 |
| alecu | dobey, why would it be difficult to pick the right one? | 17:34 |
| alecu | dobey, oh, right. | 17:34 |
| dobey | hrmm. though we have bigger problems on natty right now it seems, becuase of GI | 17:35 |
| ralsina | dobey: we could just support the environment variables on natty as a last resource | 17:35 |
| alecu | dobey, perhaps we can select the python module that should be used on the natty packaging diffs? | 17:35 |
| alecu | dobey, right: GI was pretty broken in natty, IIRC | 17:36 |
| dobey | alecu: well i'm just dealing with nightlies right now, so no patches, and one debian/ directory for all versions | 17:36 |
| dobey | AttributeError: 'gi.repository.GLib' object has no attribute 'GError' | 17:37 |
| dobey | yeah, fun times :-/ | 17:37 |
| dobey | guess i should just stick to getting oneiric working for now | 17:37 |
| dobey | then try for natty once that's done | 17:38 |
| dobey | though i am sure it will be painful :-/ | 17:39 |
| ralsina | dobey: baby steps | 17:41 |
| dobey | baby has long legs | 17:42 |
| dobey | ralsina: https://code.launchpad.net/~diegosarmentero/ubuntuone-dev-tools/u1lint-mac-support/+merge/103524 <- didn't we agree the solution to this problem, was to fix the environment, and not have this workaround in u1lint? | 17:44 |
| ralsina | looking | 17:44 |
| ralsina | I would have to read the backlog, really. I think we argued and noone decided anything. | 17:45 |
| ralsina | gatox, dobey: let's clear this definitively | 17:45 |
| dobey | ralsina: maybe you and i agreed it was an env problem, but gatox didn't agree :) | 17:46 |
| ralsina | gatox: I suspect the problem is that pylint is not being installed using the "good" python, which makes it useless. Why can't we install pylint using the python_u1 interpreter? | 17:46 |
| gatox | dobey, ralsina we just have a mumble with mandel about this | 17:47 |
| dobey | ralsina: well, i think we need to not have a special python_u1 | 17:47 |
| ralsina | gatox: and/or if that doesn't work, why not change the pylint .bat to make it work right as part of setting up the development environment instead of fouling our code | 17:47 |
| gatox | dobey, ralsina i0m going to implement something different and that branch is not going to be necessary | 17:47 |
| ralsina | gatox: ok, then | 17:47 |
| dobey | gatox: ok | 17:47 |
| * dobey rejects it then | 17:48 | |
| gatox | mmcc, free to mumble again | 17:48 |
| ralsina | dobey: that's the difference between 4 hour setups and 15 minutes setups | 17:48 |
| mandel | dobey, we sorted it out PATH=/path/to/python:$PATH | 17:48 |
| mandel | dobey, for runtests with the buildout that way we do not have to touch a thing | 17:48 |
| dobey | ralsina: hold the screw loosely, while you bludgeon it with that sledgehammer :) | 17:49 |
| ralsina | Maybe we could switch from buildout to a combination virtualenv+pip freeze | 17:49 |
| dobey | mandel: ok | 17:49 |
| dobey | ralsina: for virtualenv i think we definitely have to fix all the #! line no? | 17:50 |
| ralsina | dobey: yep | 17:50 |
| ralsina | dobey: we are using a tool (buildout) for its specified purpose (setting up environments), and it's working on 3 OSs except for a silly shebang-like-thing. I'd say this sledgehammer is a mighty nice screwdriver :-) | 17:51 |
| dobey | heh | 17:51 |
| mmcc | gatox: ok. see you there in a sec | 17:51 |
| gatox | mmcc, ack | 17:51 |
| ralsina | dobey: OR keep fixing PATH even with virtualenv. My main thing against going with virtualenv, is that WE WOULD HAVE TO DO A LOT OF STUFF AGAIN. | 17:52 |
| dobey | ralsina: i wasn't saying buildout was the sledgehammer :) | 17:52 |
| ralsina | Sorry for screaming, there ;-) | 17:52 |
| dobey | TURN DOWN THE GAIN ON YOUR MIC! | 17:52 |
| dobey | :) | 17:52 |
| ralsina | dobey: we should not need a u1lint anyway | 17:52 |
| mandel | this is when we says pyflakes... | 17:52 |
| mandel | :P | 17:52 |
| * briancurtin reboot | 17:53 | |
| ralsina | and I say "misses missing docstrings" | 17:53 |
| dobey | i'd say we should keep u1lint with pyflakes anyway | 17:53 |
| ralsina | and who knows what else. | 17:53 |
| dobey | ralsina: that's a pep8 issue, not pyflakes | 17:53 |
| dobey | so hah! :) | 17:53 |
| ralsina | dobey: you lawyerly squirrel you | 17:53 |
| mandel | briancurtin, ralsina: I need an extra review for: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/domain-sockets/+merge/103660 | 18:02 |
| briancurtin | mandel: i have my review hat on today, ill take a look in a min | 18:02 |
| mandel | briancurtin, thx! | 18:02 |
| * ralsina goes buy/mail actual review hats | 18:02 | |
| mandel | ralsina, briancurtin, also, with that branch, this one also needs a +1: https://code.launchpad.net/~mandel/ubuntu-sso-client/use-new-dev-tools/+merge/103673 | 18:02 |
| ralsina | So, not this one, google! http://www.newyorker.com/arts/critics/theatre/2011/04/25/110425crth_theatre_als?currentPage=all | 18:03 |
| briancurtin | the review hats could be like frank rossitano's hats on "30 Rock" | 18:03 |
| joshuahoover | ralsina: in regards to the 3.0.0 issues...what are we going to try next? | 18:05 |
| mmcc | briancurtin: +1 | 18:05 |
| ralsina | joshuahoover: we are doing a build with an older Qt to see if the bug is a new Qt bug | 18:05 |
| briancurtin | mandel: i already +1'ed the domain-socket branch, just didn't set to approve. doing that now | 18:05 |
| joshuahoover | ralsina: k, that's what i thought but wasn't sure if anything changed after the meeting or not | 18:06 |
| ralsina | joshuahoover: nope, just doing that AFAIK | 18:06 |
| mandel | joshuahoover, ralsina, so the deal is, all bug reports related to ssl issues are in Qt applications that moved from Qt 4.7 to Qt 4.8, so we found and old pyqt version in a university dir to test with | 18:06 |
| gatox | mmcc, maybe elopio is having lunch..... wait for him to response, he'll be able to tell you how to download that | 18:06 |
| mandel | ralsina, which brings to mind, is pyside better in this respect, as in, finding old releases etc.. | 18:07 |
| mandel | briancurtin, sweet, thx! | 18:07 |
| mmcc | gatox: ok, no problem. I'm getting precise up now | 18:07 |
| gatox | mmcc, great..... let me know if you have any problem | 18:07 |
| ralsina | mandel: well, they do have a public repo | 18:07 |
| briancurtin | oh wtf. i can't even uninstall this PyQT4 and i just rebooted...how are the files already in use? | 18:07 |
| mandel | briancurtin, is bzr is running, maybe.. it has some qt in it on the windows version (I think qlog etc..) | 18:08 |
| mandel | ralsina, it could be something to consider for 20.04 or around that | 18:08 |
| ralsina | mandel: ha | 18:09 |
| ralsina | mandel: sigh | 18:09 |
| dobey | i suppose i should have looked at that domain-socket branch | 18:10 |
| mandel | dobey, if it is not merged, we can stop it from landing! | 18:10 |
| mandel | dobey, there you go: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/domain-sockets/+merge/103660 | 18:11 |
| mandel | dobey, I think tarmac did not pick it up | 18:11 |
| mmcc | launchpad question: should I be joining some teams on there? ubuntuone-hackers, for example? | 18:11 |
| ralsina | mmcc: you should have been added automatically | 18:12 |
| mandel | dobey, ok, shit it did :( | 18:12 |
| ralsina | mmcc: or manually but before now ;-) | 18:12 |
| dobey | mandel: i can look at it later | 18:12 |
| mmcc | ralsina: nope… no teams for me | 18:12 |
| dobey | ralsina: he should be added to ubuntuone-desktop+ | 18:12 |
| ralsina | dobey: do you have the right superpowers? | 18:13 |
| dobey | mmcc: what's your lp username? | 18:13 |
| mmcc | mikemc | 18:13 |
| mmcc | dobey --^ | 18:13 |
| ralsina | pfffft I am a transitive member of "“Turkey Pop-up Team” team" | 18:13 |
| * ralsina wonders what people drink sometimes | 18:14 | |
| dobey | ralsina: you are an admin of ubuntuone-desktop+ team | 18:14 |
| ralsina | dobey: ack, want me to add him? | 18:14 |
| dobey | ralsina: yes | 18:14 |
| ralsina | dobey: doing | 18:14 |
| dobey | probably need to move some people around in our teams | 18:15 |
| ralsina | added | 18:15 |
| mmcc | ralsina: thanks | 18:15 |
| dobey | half the people in that team, aren't on our team any more. and a bunch of people aren't in that team | 18:15 |
| ralsina | dobey: yeah, I'll do some cleaning now | 18:16 |
| ralsina | dobey: but I have to ask because I don't want people locked out of other projects just to be neat | 18:16 |
| dobey | true. but would be a good joke :) | 18:16 |
| mmcc | so making me a member of ubuntuone-desktop+ makes me a member of the teams that it's a member of? | 18:17 |
| dobey | mmcc: yes | 18:17 |
| mmcc | ok, cool | 18:17 |
| mandel | dobey, can you do me a favour and keep track of this guy: https://code.launchpad.net/~mandel/ubuntu-sso-client/use-new-dev-tools/+merge/103673 it needs to be merge after the nightlies of u1-dev-tools have landed, otherwise we will have sso blocked with tarmac errors | 18:18 |
| mandel | dobey, the api changed a little because twisted is not consistent between internet.endpoints and reactor.connectTCP :( | 18:19 |
| dobey | fun | 18:20 |
| mandel | dobey, yeah, I know.. | 18:21 |
| dobey | almost as fun as python-gi | 18:21 |
| dobey | ralsina: so, what if i do stick twisted 11.1 in nightlies? think the world will end? | 18:22 |
| ralsina | dobey: yes | 18:23 |
| ralsina | dobey: nightlies for what release? | 18:23 |
| dobey | oneiric | 18:24 |
| mmcc | roadmap question: what versions of OS X do we want to target (aka, test on?) | 18:24 |
| briancurtin | mandel: last check before i do this new installer: PyQt 4.8.6 with Qt 4.7.4 is what we want, right? | 18:24 |
| dobey | ralsina: the patch for the gireactor/gtk3reactor doesn't work on 11.0 it seems :( | 18:25 |
| ralsina | holy carp. | 18:25 |
| ralsina | dobey: doesn't work as in "doesn't fix it" or as in "has conflicts/needs tweaks"? | 18:26 |
| dobey | ralsina: as in "uses API which isn't in 11.0 but is in 11.1" | 18:26 |
| ralsina | dobey: ugh. | 18:26 |
| mandel | briancurtin, yes, qt 4.7 is what we want :) | 18:27 |
| ralsina | dobey, alecu: twisted is supposed to be very backwards compatible so that putting 11.1 in nightlies will not break every other twisted-usig thing. Right? | 18:27 |
| dobey | ralsina: i /presume/ that things will only break if they had the same bug we did when 11.1 landed in precise, which we had to fix | 18:28 |
| mandel | in theory... | 18:28 |
| ralsina | grmbl | 18:28 |
| dobey | we can go cry in a corner maybe :) | 18:30 |
| dobey | don't think it will help though | 18:30 |
| ralsina | that was my current plan. Damn. | 18:30 |
| ralsina | dobey: how about asking for a twisted 11.1 SRU? ;-) | 18:30 |
| dobey | ralsina: well we could probably get the precise version with patches into backports, don't know if we could get it into SRU yet | 18:31 |
| ralsina | ok, backports looks like a good idea. | 18:31 |
| dobey | but neither of those helps me to get nightlies working :) | 18:31 |
| ralsina | dobey: yes. So, we do need to get it into nightlies anyway. | 18:32 |
| ralsina | So go ahead. | 18:32 |
| dobey | at least, not before UDS anyway | 18:32 |
| gatox | mmcc, here you can see the list of the issues assigned to you in U1: http://is.gd/Is9R2X | 18:33 |
| Scunizi | Is this a developers channel or also a generic help channel for Ubutu one? | 18:34 |
| mmcc | gatox: ack | 18:34 |
| dobey | Scunizi: it's basically the only channel for ubuntu one. though there's also #u1db for questions about that | 18:35 |
| Scunizi | dobey: thanks.. I had a question about syncing Tomboy.. It keeps failing to syncronize after prompting me to rename a note.. then shows several notes that are updated.. None of which I can see when logging into my account online. | 18:36 |
| dobey | Scunizi: yes, the notes web ui was removed. there was an announcement about it | 18:37 |
| Scunizi | dobey: must have missed that.. will it be coming back? | 18:37 |
| dobey | i don't know | 18:37 |
| Scunizi | any ideas about the syncronization? | 18:37 |
| dobey | http://voices.canonical.com/ubuntuone/2012/02/05/an-important-note-about-notes/ | 18:38 |
| dobey | i don't have any ideas about your sync issue, no. | 18:38 |
| dobey | rye: ^^ can you help Scunizi ? | 18:38 |
| ralsina | dobey: may be a bit late for rye. Maybe duanedesign? | 18:39 |
| dobey | Scunizi: i /think/ there's a read-only interface to see your notes, at http://u1.to/ | 18:39 |
| Scunizi | cool.. I'll look at it.. | 18:39 |
| Scunizi | Interesting.. got my list of notes. | 18:41 |
| Scunizi | The frustrating thing about syncing multiple machines is after updating from one machine and attempting on another, the second machine is prompted to rename the changed note to (old) so you end up with 2 copies. | 18:43 |
| Scunizi | sorry.. must afk for a while. | 18:44 |
| dobey | Scunizi: that may be a problem with tomboy itself. i'm not sure. we just implement tomboy's "snowy" sync protocol | 18:45 |
| Scunizi | Nice to know.. I'll see if I can chat with them as well. thanks. | 18:46 |
| dobey | gatox: hmm. why all these bugs that are just "Refactor foo" ? | 18:49 |
| gatox | dobey, we talk with mandel ralsina and mmcc ...... we are refactoring u1-client/platform to enhance that, and be able to run the tests for each module better.... and start identifying which is needed for mac, etc | 18:50 |
| duanedesign | ralsina: whats up | 18:50 |
| ralsina | duanedesign: Scunizi here has a notes issue | 18:50 |
| gatox | dobey, we had a couple of mumbles about this | 18:50 |
| duanedesign | Scunizi: I think I can help | 18:51 |
| ralsina | dobey: and you even was there in one of those :-) | 18:51 |
| ralsina | were | 18:51 |
| * ralsina feels abandoned by his english | 18:51 | |
| dobey | gatox: right, i understand the task. but why are the bugs not describing the problem? | 18:51 |
| ralsina | dobey: we seem to be having an epidemic of those | 18:52 |
| dobey | ralsina: right, i'm not asking about why that is being done. bugs that don't describe problems are not good though :) | 18:52 |
| duanedesign | Scunizi: https://bugs.launchpad.net/ubuntu/+source/tomboy/+bug/848250/comments/10 | 18:52 |
| ubot5 | Ubuntu bug 848250 in tomboy "Tomboy sync fails on Oneiric, New Note Template already exists" [Medium,In progress] | 18:52 |
| duanedesign | Scunizi: rename the note locallly using the preferences. Sync and rename the note | 18:53 |
| gatox | dobey, ok....... i'll update the description of the bugs in a while | 18:53 |
| ralsina | gatox: you need to word them like "The current code layout makes it impossible/hard to separate the tests per-platform" or somesuch. Describe the problem, then you can describe the proposed solution in the description. | 18:54 |
| ralsina | or even in the 1st comment | 18:54 |
| dobey | well, i'd like us to avoid splitting tests per-platform where possible | 18:54 |
| gatox | ralsina, roger that | 18:54 |
| ralsina | like "1st!!!!oneeleven! we shoul refactor like this blah blah" | 18:54 |
| dobey | i think it's better to have skips than all the crazy logic to ignore specific tsts on specific platforms | 18:55 |
| gatox | dobey, it would be just: -i "test_windows.py" | 18:55 |
| dobey | gatox: right, though i'd also really like if we could get rid of that as well :) | 18:56 |
| ralsina | ok, now we are getting bugs because apt failed to replace files from protobuffer. Sigh. | 18:56 |
| gatox | dobey, only where it's necessary...... sometimes we can check just os_helper (for example)...... but sometimes we need to check some stuff os_helper/windows.py | 18:56 |
| duanedesign | Scunizi: If that does not work for soe reason, you can also try: http://paste.ubuntu.com/947947/ | 18:57 |
| gatox | odso, i think that in most cases we are going to have both | 18:57 |
| gatox | dobey, ^ | 18:57 |
| dobey | gatox: right, but i'd also hope that we can work to get rid of those platform imports as well :) | 18:57 |
| gatox | dobey, what do you suggest? | 18:57 |
| dobey | gatox: well, it's something that has to be dealt with on a case-by-case basis. i don't think there's a general solution to it | 18:58 |
| ralsina | dobey: we are not putting "if sys.platform" everywhere. Just in case ;-) | 18:58 |
| gatox | dobey, http://youtu.be/NFQCYpIHLNQ ? jeje | 18:59 |
| dobey | heh | 18:59 |
| gatox | dobey, we are going to need in some places if sys.platform...... and in some other places not | 18:59 |
| dobey | gatox: yes of course. it's not that which i don't like, but the imports that are platform-specific (linux.py, windows.py, etc) | 19:00 |
| dobey | anyway, u1client is a long way off from having a single import where we have those now | 19:01 |
| === JanC_ is now known as JanC | ||
| briancurtin | mandel, alecu, ralsina: here's 3.0.0 + mandel's fix-ssl branch, built with Qt 4.7.4 - http://u1.to/~brian.curtin/Y/ubuntuone-3.0.0-windows-installer.exe | 19:10 |
| alecu | briancurtin, awesome. | 19:10 |
| ralsina | joshuahoover, rye ^ | 19:10 |
| briancurtin | i installed it, then i synched the installer with it, so it works for that basic definition of "works" | 19:10 |
| joshuahoover | briancurtin, ralsina: excellent, let me see if i can get some users to test this out and get us logs/feedback | 19:11 |
| ralsina | joshuahoover: awesome, thanks | 19:11 |
| dobey | man, there is like no infrastructure for python 3.x | 19:11 |
| briancurtin | if there's anything special i should do to exercise whatever Qt differences, let me know and i'll do any tests | 19:11 |
| ralsina | dobey: which is why all the python elite wants us to use it so we feel the burn and fix it. | 19:11 |
| ralsina | dobey: and I agree with them, actually. I just don't want to *work* on it. | 19:12 |
| dobey | 0o777 | 19:14 |
| dobey | wtf is that | 19:14 |
| dobey | yay for breaking compat | 19:15 |
| roasted | Question, does Ubuntu One delete any files? Like if i have 10 files on Ubuntu One, and I delete 5 of those items, would those corresponding 5 files delete off my computers linked to that Ubuntu One account? | 19:16 |
| roasted | Or does Ubuntu One retain everything ever synced until the user manually deletes it off each system? | 19:16 |
| alecu | dobey, you can also use things like: chr(0b11111010010101001) | 19:17 |
| alecu | (or unichr, in 2.7) | 19:18 |
| ralsina | roasted: they will be deleted from all your computers, because Ubuntu One syncs the deletions (or rather, they will be moved to the trash on each computer) | 19:20 |
| dobey | ralsina: i'm not entirely sure that's entirely true, for the mobile apps case | 19:21 |
| ralsina | dobey: the mobile apps don't sync, that is true | 19:21 |
| dobey | oh | 19:21 |
| dobey | he didn't mention mobile apps this time | 19:21 |
| dobey | was expecting the same question to be asked :) | 19:21 |
| ralsina | dobey: yep. Computers :-) | 19:21 |
| dobey | roasted: yes anything that syncs your data on u1 also syncs deletions | 19:22 |
| dobey | roasted: it's a sync service, not a backup service :) | 19:22 |
| mmcc | ralsina, I was just looking at move_to_trash; it looks like if there's an error moving to trash, we just delete... I was wondering how often that happens. | 19:23 |
| ralsina | mmcc: really? | 19:23 |
| ralsina | mmcc: I must confess I hardly ever delete aything | 19:23 |
| ralsina | mmcc: oh, sorry, I misread you | 19:24 |
| mmcc | I expect that the move-to-trash code basically always works | 19:24 |
| ralsina | well, moving to trash should never fail. And yes, delete is the only valid action | 19:24 |
| mmcc | right | 19:24 |
| ralsina | mmcc: it may fail if disk is vry full | 19:24 |
| roasted | dobey: that's exactly what I want. I just wasn't sure how multidirectional it worked. I use it for syncing all of my pictures on my android to U1 so my computers all get those pictures. I just wanted to make sure if I delete a picture on my phone and U1 kicks in to sync, it also deletes it on the computers as well. | 19:24 |
| ralsina | mmcc: and the trash is not on the same disk as the file you are deleting, or something. But just guessing. | 19:24 |
| mmcc | ralsina, yes - is trash on another disk in ubuntu? | 19:25 |
| ralsina | roasted: no, the phone client doesn't do sync, it just does uploads, I think | 19:25 |
| dobey | roasted: i am not sure it will in that case | 19:25 |
| ralsina | mmcc: not usually :-) | 19:25 |
| mmcc | mmcc: on macosx, trash is per-volume | 19:25 |
| dobey | roasted: the photos stuff on android doesn't sync. it only auto-uploads, as i understand | 19:25 |
| dobey | karni: ^^ is that true? | 19:25 |
| roasted | ahhh okay | 19:25 |
| mmcc | er ralsina ^ | 19:25 |
| ralsina | dobey: it's like 10PM in the ukrayne :-) | 19:25 |
| roasted | dobey: but if I have 5 computers, and I drop an item in my U1 folder on computer A, would B C D E all receive it? Likewise, if A deletes that ifle, would it be removed from B C D E? | 19:26 |
| dobey | ralsina: karni is in poland :) | 19:26 |
| dobey | so 9:30, but it's karni | 19:26 |
| ralsina | dobey: right. It's like 10PM in poland ;-) | 19:26 |
| dobey | roasted: yes | 19:26 |
| ralsina | mmcc: well, if it never fails, cool. If it fails, it probably is because of something very bad happening to yur disk and all yur data will die anyway ;-) | 19:27 |
| roasted | But on my phone... it'd just be one way... phone pushes up, computers receive it... however if A deletes those U1 items the phone just uploaded, the B C D E all lose it too.... | 19:27 |
| roasted | I think I'm getting it. :P | 19:27 |
| ralsina | roasted: right | 19:27 |
| ralsina | roasted: but it will still be there in the phone | 19:27 |
| roasted | gotcha, gotcha | 19:27 |
| mmcc | ralsina: exactly. if this happens, "you have other things to worry about™" | 19:27 |
| dobey | roasted: yeah. until the phone client gets sync support (which it may have now, but i am not 100% sure) | 19:27 |
| roasted | the phone is the only exception, right? | 19:27 |
| ralsina | roasted: you have to manually delete in the phone to lose the last copy | 19:27 |
| ralsina | roasted: yes, only the phone clients are not sync-based | 19:28 |
| dobey | roasted: also, on the computers it doesn't entirely delete it | 19:28 |
| roasted | I'm not sure I'd want my phone to sync. I like the upload only feature. | 19:28 |
| dobey | roasted: it will move it to trash on each of the computers | 19:28 |
| dobey | roasted: so if it was an accident, you can easily restore it | 19:28 |
| roasted | I'd hate to accidentally delete the U1 folder on my desktop and BAM its gone on my phone | 19:28 |
| roasted | ah that's good at least | 19:28 |
| ralsina | mmcc: right. Never got a complaint, specially because most users assume it just deletes anyway. | 19:28 |
| roasted | I wonder if the android app gets sync support if we can still set it to be upload only. I use U1 for the convenience of not having to plug it in to the computer to receive my pictures. | 19:29 |
| roasted | Fire up computer, U1 starts, and automagically the camera pics come down. | 19:29 |
| ralsina | roasted: sync on android is probably going to be optional/manual because of battery concerns | 19:31 |
| dobey | roasted: possibly. no idea. there are battery and storage concerns for mobile, certainly | 19:31 |
| joshuahoover | ralsina: k, sent out the latest installer to the 4 users who got back to us with logs from the last debug build | 19:37 |
| * rye missed the whole fun here | 19:38 | |
| ralsina | joshuahoover: awesome | 19:40 |
| joshuahoover | ralsina, alecu: ok, got one response from a user testing briancurtin's latest...i updated bug #985816 with logs (and the user included a screenshot in the archive) | 19:49 |
| ubot5 | Error: Launchpad bug 985816 could not be found | 19:49 |
| ralsina | joshuahoover: thanks, will look in 1' | 19:49 |
| roasted | hm | 19:51 |
| roasted | I'm uploading ~300 pictures but they're not showing up on Ubuntu One web.. yet. I wonder if the entire sync (150 to go) has to complete before they populate on the site under my account. | 19:52 |
| rye | roasted: are you uploading from the phone? | 19:52 |
| roasted | yes | 19:52 |
| rye | roasted: then no, once one is uploaded it should be already available | 19:53 |
| roasted | well, dang | 19:53 |
| roasted | wonder why it's not populating | 19:53 |
| rye | roasted: check what is the location for the uploaded files | 19:53 |
| roasted | from the phone or on the site? | 19:53 |
| dobey | ah crap. now sso fails because of the test case changes | 19:53 |
| rye | roasted: Auto Upload settings on the phone itself | 19:54 |
| roasted | rye: well, I had a bunch of stuff uploading from my phone I didn't want. So I went on U1's site, deleted everything, then from the phone under auto upload settings, I selected "upload all photos now" | 19:55 |
| roasted | rye: I had 300 something and I'm down to 130, but nothing is showing up on the U1 site yet. | 19:55 |
| rye | roasted: what's the location set on the phone, Pictures - "name of the phone" ? | 19:57 |
| dobey | sigh | 19:58 |
| roasted | rye: yes | 19:58 |
| dobey | mandel: why the heck did you rename the module?! | 20:00 |
| roasted | allright, we got itnow | 20:03 |
| roasted | I had to go to the folder on the phone, hit menu, then auto upload here. | 20:03 |
| roasted | then went back into auto upload settings, hit upload all now, and I see items in the web populating. | 20:03 |
| roasted | so flippin nice. | 20:04 |
| rye | interesting | 20:05 |
| rye | then where did it was uploading the files before that | 20:06 |
| rye | roasted: is that an upgrade from earlier version or a fresh install of ubuntuone files? | 20:06 |
| roasted | rye: it was uploading to the same place. I think I hit unsubscribe to the folder on the U1 web interface on accident when I was trying to figure out how to delete everything on U1 | 20:07 |
| roasted | rye: I had nothing on U1 at all, but I didn't realize I had so many ridiculous pictures on my phone, so when U1 uploaded them I was like nahhh get that off there... so I went through my phone (after finding out the app isn't an active sync, but just an upload) and deleted my old stuff off. | 20:07 |
| alecu | joshuahoover, I'm looking at the new debug log. It seems that the problem is still present with an older qt (4.7 instead of 4.8). | 20:08 |
| roasted | rye: so I wanted to nuke my U1 database to start fresh... now I'm reuploading from the phone and it works since I designated WHERE to put the files. durp durp. | 20:08 |
| joshuahoover | alecu: yeah, appears that way...still waiting to see some others | 20:08 |
| alecu | joshuahoover, I wonder if we are missing something else, like the user having some firewall installed, or some kind of antivirus that's troubling us. | 20:08 |
| rye | roasted: let me reiterate that, you unsubscribed a folder on the web (effectively deleting it) but the phone kept uploading? | 20:08 |
| roasted | rye: the phone said it was uploading, yes. but I was F5'ing the site and it kept saying 0 KB. | 20:09 |
| joshuahoover | alecu: we've asked the impacted users about this and there is no pattern...some have firewall turned on, some not...some have norton, some have microsoft security essentials, others have it turned off | 20:09 |
| roasted | rye: 150 pictures uploaded into the ~300 total, I canceled it and looked further. | 20:09 |
| joshuahoover | alecu: i've tested with norton and ms security essentials (separately) and wasn't able to reproduce, just in case that was an issue | 20:09 |
| rye | alecu: i tried avast, norton and avg - nothing resembling any ssl errors | 20:09 |
| alecu | :/ | 20:09 |
| roasted | rye: once I went into the Pictures - LG 500whatever folder, I hit the menu button and saw the auto upload here option. So I hit that... tried to upload 1 item... came up fine when I F5'd the U1 Web. | 20:10 |
| rye | alecu: one user had avast | 20:10 |
| roasted | rye: so I just went back, upload all, and now it's almost done. :) | 20:10 |
| rye | roasted: ok, i guess we will need to re-test this, i suspect something is wrong there but I can't pinpoint at what stage, but I am glad you were able to fix it | 20:10 |
| roasted | rye: I think part of the issue was me unsubscribing and then trying to upload as if nothing ever happened. So if you wnat to duplicate what I did... android phone with U1 installed... unsubscribe online (when data is already uploaded) to the folder the phone is linked to, then upload all photos ot U1 from phone... Then F5 the U1 web and see if anythign comes up. | 20:11 |
| roasted | rye: for me, nadda... once I told U1 on the phone where to upload (I assume that effectively re-establishes the subscription since it now has a destination folder) it worked fine during uploads. | 20:12 |
| === m_conley is now known as m_conley_away | ||
| gatox | EOD for me! see you tomorrow! :D | 20:31 |
| ralsina | bye gatox! | 20:31 |
| ralsina | EOD for me now | 21:04 |
| ralsina | bye! | 21:04 |
| dobey | same here. have to bug mandel in the am | 21:10 |
| dobey | later all | 21:10 |
| mmcc | bye guys | 21:13 |
| mandel | dobey, because it is not longer just tcp, it has domain sockets too, it does no make sense to keep the name | 21:17 |
| alecu | ralsina, mandel, briancurtin: do any of you run win 7 x64? | 21:32 |
| briancurtin | alecu: I do | 21:32 |
| alecu | hmmm | 21:33 |
| mandel | I do | 21:33 |
| mandel | on a vm | 21:33 |
| alecu | there goes my theory. | 21:33 |
| alecu | anyway, here's the exact same error we found: http://stackoverflow.com/questions/3435024/qt-qsslerror-being-signaled-with-the-error-code-set-to-noerror | 21:33 |
| alecu | mandel, briancurtin ^ | 21:33 |
| alecu | "everytime I try to download an https page, it will signal two QSslError, each with contents NoError" | 21:33 |
| alecu | I've tracked the "NoError" to X509_V_OK here: http://www.openssl.org/docs/apps/verify.html#DIAGNOSTICS | 21:34 |
| mandel | briancurtin, oh, and hi from reddit ;) | 21:34 |
| mandel | alecu, so, wait, it is sending an error saying is ok, wtf? | 21:34 |
| alecu | mandel, that error means: "no verification error occurred, it does however not indicate success" http://www.openssl.org/docs/ssl/SSL_get_verify_result.html#BUGS | 21:35 |
| alecu | mandel, that means that OpenSSL is trying to verify a certificate, but got no certificate in response. | 21:35 |
| briancurtin | mandel: haha. if twisted was done and ready, we could port in no time. at the pycon language summit we talked about giving someone a grant to finish the port (and grants for people reviewing it) | 21:36 |
| mandel | alecu, hm.. when can that be the case? | 21:36 |
| mandel | briancurtin, I know facundobatista did help a little in pycon with twisted towards python 3... it is going to be interesting :) | 21:36 |
| alecu | mandel, hacked network? MitM? don't know! | 21:37 |
| mandel | alecu, verterok mentioned time out issues in the connection, then we get disconnected before the ssl cert gets there and we fail in the handsacke | 21:37 |
| alecu | mandel, I don't think the problem is with timeouts... :-( | 21:38 |
| mandel | alecu, if there was only I way to know which machine is taking the hit.. | 21:39 |
| mandel | alecu, that would make it simpler to check since we would have the logs from the other side | 21:39 |
| alecu | mandel, so, we should try to reproduce a case were we get an "empty certificate" from a fake server | 21:40 |
| alecu | mandel, and see if that produces that error. | 21:40 |
| alecu | mandel, my current wild guess is that transparent proxies are messing with us. | 21:40 |
| mandel | alecu, yep, looks like the 'cause since the cert returned by the ssl error has no information in it which probably means that the ssl erro signal is using the default values | 21:40 |
| mandel | alecu, I can do a test case for ubuntu-sso tom morning based on the ssl test cases we already have | 21:41 |
| alecu | mandel, right: there must be no cert in that case. | 21:41 |
| alecu | mandel, anyway: we still have the problem where one guy had this issue with one computer in his lan, but he had no problem with the other computers in that same lan... :/ | 21:41 |
| mandel | alecu, we can ask him which version he is using, also, he can be that unlucky that he hits a server that is missing a cert.. and just in one of the calls | 21:43 |
| alecu | mandel, I still don't believe we have a problem in *our* servers. | 21:43 |
| mandel | alecu, since all the diff calls we make are async it means that he should be getting more than one erro dialog due to the fact that the webclients don't know the existance of each other | 21:43 |
| mandel | alecu, I don't say is THE reason, but is one thing to check | 21:43 |
| alecu | mandel, we have a finite number of servers, so we should be seeing this problem much more if that was the case. | 21:43 |
| mandel | alecu, I'll write the test and we will have an answer, do we get the error or not? | 21:44 |
| mandel | alecu, then we can move.. | 21:44 |
| mandel | if we could only reproduce it.. | 21:44 |
| alecu | mandel, awesome. Now get to bed :-) | 21:44 |
| alecu | mandel, oh, it's not so late there :-) | 21:45 |
| mandel | alecu, is ok, just close to 00:00 | 21:45 |
| mandel | alecu, I want to play with pb and see if I can change bannana.py for a json version :) | 21:45 |
| alecu | mandel, so you don't like the banana? I thought you did :-) | 21:46 |
| mandel | alecu, is a PITA when interacting from other langs, so, or I do a C verion of bannana or I change the thing | 21:47 |
| mandel | alecu, I did get the pun ;) | 21:47 |
| briancurtin | BitRock is an example of a project that has thousands of pages of documentation which contain almost nothing in them. Hundreds of examples with zero explanation. arghhhh | 21:51 |
| mmcc | hey guys, anyone want to look at my first pushed branch and tell me if I did the right stuff to push it to lp, link to bug etc? | 21:51 |
| mandel | briancurtin, yes, is terrible, and some of them do not even make sense.. | 21:52 |
| mmcc | it's bug 989119 | 21:52 |
| ubot5 | Launchpad bug 989119 in Ubuntu One Client "The current code layout makes it hard to separate the tests per-platform (constants)" [High,New] https://launchpad.net/bugs/989119 | 21:52 |
| mandel | mmcc, lets take a look :) | 21:52 |
| mmcc | hey, stupid windows question here: I'm grabbing a windows image now, what version should I test with (7, I guess) and do I need visual studio to run our tests (yes I assume?) | 21:56 |
| mandel | mmcc, no need of visual studio, 7 is the best one to pick | 21:57 |
| mmcc | thx mandel. windows is ... not my forte :) | 21:57 |
| alecu | mmcc, the branch is uploaded ok, and linked to the bug ok. | 21:57 |
| briancurtin | i guess XP is "supported" but it's not high on the radar, or something like that | 21:57 |
| mandel | mmcc, there you linked the bug, yet you did not propose the branch for merging, in case you want to do that, you can go to https://code.launchpad.net/~mikemc/ubuntuone-client/refactor-platform-constants | 21:57 |
| mandel | mmcc, and then say propose for merging :) | 21:58 |
| alecu | mmcc, we usually do reviews on "merge proposals" which are created as mandel says | 21:58 |
| mmcc | mandel, alecu aha ok | 21:58 |
| mandel | mmcc, there you have a description (some info to give about the branch) and a commit one (expand the extra options | 21:58 |
| alecu | mmcc, you should do that when your branch has the tests and you've run them in linux and windows... and now osx too :-) | 21:58 |
| mandel | mmcc, the commit is what will be added in the merge to trunk and should be of the format: | 21:59 |
| alecu | mmcc, I guess you are reviewing tomorrow, so you'll get more familiar with that procedure. | 21:59 |
| mandel | - Sentence explaining things (LP: #211212) :) | 21:59 |
| mandel | mmcc, but yeah, run tests on windows and linux and we will do the entire review cycle :) | 22:00 |
| briancurtin | ooooh this is going to be good. i think i need to write some VBScript for the installer... | 22:00 |
| mandel | briancurtin, wtf?! | 22:00 |
| mmcc | mandel: so that # after your sentence is the bug my proposed branch will fix? | 22:01 |
| mandel | mmcc, yep :) | 22:01 |
| mandel | mmcc, let me show you an example, one sec | 22:01 |
| alecu | mmcc, as mandel says, the format of the "commit message" is important, because it's used for the ubuntu packaging. The description of the merge is not that much important, but we usually put there some info on how the reviewers should test the branch, if it's different than what they usually do. | 22:01 |
| mmcc | and I think I'll hold off on proposing this until I can run the tests on win & linux... | 22:01 |
| mandel | mmcc, here you have an example: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/fix-squid-tests/+merge/103325 | 22:01 |
| briancurtin | mandel: upgrading from 2.0.3 to 3.0.0, we want the old version to stop and to do the uninstaller, then install 3.0.0. i made it do the uninstaller, but we have to be able to stop the existing U1 before proceeding. i wrote a python script and py2exe'ed it, but we need to do the process killing before we have access to "installdir" stuff | 22:02 |
| briancurtin | but apparently i can embed VBScript and it'll work, or something | 22:02 |
| mandel | briancurtin, ou can embed anything, right? | 22:02 |
| === salgado is now known as salgado-afk | ||
| mandel | mmcc, from my point of view, failing tests on mac are ok at this stage, nothing will land if tests dont pass on linux so no worries | 22:03 |
| mmcc | mandel: is it pretty common for the commit message to be a subset of the description? | 22:03 |
| mandel | mmcc, no, that is my style, the description can be anything | 22:03 |
| mandel | mmcc, braking things or windows are possible, that is why we have: https://jenkins.errormessaging.com/view/Windows/ | 22:03 |
| briancurtin | mandel: well it has support for passing stuff to VBScript since it's built into windows, supposedly. the python script i wrote is simple and just uses WMI to find any process names that start with "ubuntu" then it kills them. have to try porting it to VBScript i guess | 22:03 |
| mandel | briancurtin, ouch, well at least we are not using msi directly | 22:04 |
| alecu | mmcc, I usually put the same text in both the description and the commit message. But the important one is the latter. | 22:04 |
| briancurtin | this whole thing is filled with "i guess" and "supposedly"...not good, thanks to their great docs | 22:04 |
| mmcc | mandel: ok, so the proposed merge's commit message has a specific format, but the description is my discretion, and includes testing info if required? | 22:05 |
| mandel | briancurtin, lol I remember we had something similar the first time with the autoupdater, it failed because it does not support https, wtf?! I discovered via trial an error | 22:05 |
| mandel | mmcc, exactly | 22:06 |
| mmcc | mandel, alecu ok good, got it | 22:06 |
| mmcc | mandel: I get 404 from that jenkins link ;/ | 22:06 |
| mmcc | did I need to let it know my group membership from SSO? I left that unchecked | 22:07 |
| mandel | mmcc, yes, you do :) | 22:07 |
| * mmcc is deleting cookies (and getting hungry) | 22:08 | |
| mmcc | oh hey neat, that looks useful | 22:09 |
| mandel | mmcc, we are working on getting things more stable :) | 22:10 |
| mmcc | mandel: I SEe | 22:10 |
| mandel | mmcc, there are some unexpected errors when running windows tests on a user with no home et.. | 22:11 |
| mandel | mmcc, my bad I suppose :) | 22:11 |
| mmcc | so i'm inclined to use the rest of my time to get this branch tested on my super new ubuntu VM image, before I propose it for merging. sound like a good idea? | 22:12 |
| mmcc | mandel: poor homeless users | 22:12 |
| mmcc | Hmmm, which windows 7 do I get? there are three - ultimate, enterprise, or professional... | 22:13 |
| mandel | mmcc, yes, +1000 | 22:13 |
| mandel | mmcc, if it is free, get the expensive one ;) | 22:13 |
| mandel | mmcc, the diff is minumun, no support for AD etc.. | 22:13 |
| mmcc | mandel, no price shown... but seriously, which one is more complete? | 22:14 |
| mandel | mmcc, one trick for the linux vm apt-get build-dep ubuntuone-client (will install all devl deps) | 22:14 |
| mandel | mmcc, ultimate | 22:14 |
| briancurtin | mmcc: professional is fine. there's no difference that would matter | 22:14 |
| briancurtin | do you have MSDN? | 22:14 |
| briancurtin | (i need to ask about MSDN myself) | 22:14 |
| mmcc | briancurtin, go search for elopio and my nick on u1-internal... | 22:15 |
| mandel | briancurtin, I have it too :) | 22:15 |
| alecu | briancurtin, do you have the 2.0.3 link handy? | 22:15 |
| mmcc | and thanks, briancurtin - getting pro now. I feel very professional | 22:15 |
| briancurtin | alecu: it's in http://one.ubuntu.com/windows/ -- http://one.ubuntu.com/windows/ubuntuone-2.0.3-windows-installer.exe | 22:16 |
| alecu | briancurtin, right, thanks. | 22:16 |
| mandel | ok, all see you tom! | 22:17 |
| mmcc | bye mandel, thanks | 22:17 |
| mmcc | hey briancurtin, more questions - I assume I want "with service pack one", but do I want "win 7 professional N with service pack 1", "win 7 professional with service pack 1, VL build", or just "win 7 professional with service pack 1 - DVD" | 22:18 |
| briancurtin | mmcc: let me login and see the list, i always forget which one you want. 1 min | 22:19 |
| mmcc | briancurtin: ok thanks | 22:19 |
| mandel | mmcc, do not pick N, N is bad, very bad I tried it and never managed to install it | 22:20 |
| briancurtin | Windows 7 Professional with Service Pack 1 | 22:20 |
| * mandel really goes now | 22:20 | |
| mmcc | briancurtin: excellent, thanks. do you know what the others are? | 22:21 |
| mmcc | briancurtin: it's not really important, just curious | 22:22 |
| briancurtin | the N builds don't come with Windows Media Player, and VL is for volume licensing | 22:23 |
| mmcc | briancurtin: aha! now I can sleep tonight :) thanks again | 22:23 |
| briancurtin | no problem, and with that, i'm out of here | 22:23 |
| mmcc | my time is up as well. I'm going to try leaving this thing on tonight though. | 22:24 |
| === mmcc is now known as mmcc_absent | ||
| alecu | briancurtin, already gone? | 22:29 |
| alecu | briancurtin, I installed this: http://u1.to/~brian.curtin/Y/ubuntuone-3.0.0-windows-installer.exe | 22:29 |
| briancurtin | alecu: i went from my office to my couch | 22:30 |
| alecu | briancurtin, in a clean vm. But the qt .dll says "4.8.0.0" | 22:30 |
| alecu | briancurtin, ok, enjoy your couch then. :-) | 22:30 |
| alecu | briancurtin, don't worry about this, we can worry about that tomorrow. | 22:30 |
| briancurtin | alecu: hmm, ah i think i know why...the py2exe found the existing build and didn't bother to pick up what's on the system | 22:30 |
| briancurtin | alecu: i'll do a clean build tomorrow morning, that should have the right version | 22:31 |
| alecu | briancurtin, great. | 22:31 |
| alecu | briancurtin, another thing I noticed: SSLEAY32.dll in 2.0.3 is 0.9.8.18; and in 3.0.0 is older: 0.9.8.12 | 22:32 |
| briancurtin | weird...maybe it picked up part of the older version (the one i installed) and part of the newer version (which was in the previous build and dist folders) | 22:32 |
| alecu | briancurtin, I'm not sure any of this will fix the issue, but I've reverted to comparing version numbers since we don't have any other clear clue :P | 22:33 |
| briancurtin | alecu: in any case i'll be sure to clean out the environment and get a proper build. i can also add something to the installer builder to output versions, perhaps for a checklist | 22:34 |
| alecu | briancurtin, perhaps we can send that to the installer log? | 22:35 |
| alecu | what I just said makes no sense. | 22:36 |
| * alecu should get some rest. | 22:36 | |
| alecu | briancurtin, right: adding it to the installer builder and checking it before release makes sense. | 22:36 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!