[00:00] nessita, hi, are you there? [00:25] DiegoSarmentero: I am! [00:26] nessita, Hi, I having some problems trying to test windows installer on linux [00:26] the dependency that i show you today...... [00:26] DiegoSarmentero: why? [00:27] nessita, that it says that there isn't a module named qt.gui [00:27] in test_gui.py [00:27] DiegoSarmentero: you trying on linux or windows? [00:27] linux [00:28] DiegoSarmentero: installer will not work on Linux :-( [00:28] nessita, neither the tests? [00:28] nessita, i'm only want to run the tests [00:29] DiegoSarmentero: with what we have in trunk, not even the test. I just build a branch that I'm about to propose that allows you to: [00:29] 1- build the QT ui modules in sso (the thing that is failing for you) [00:30] 2- run the tests in the installer [00:30] DiegoSarmentero: wanna try it? [00:30] sure! [00:30] nessita, ^ [00:31] DiegoSarmentero: so, for building SSO in linux please branch the following: [00:31] lp:~nataliabidart/ubuntu-sso-client/build-on-linux [00:31] DiegoSarmentero: enter that dir and run ./setup.py build in it [00:31] ok [00:31] doing it :P [00:32] then, go to the installer code, and set PYTHONPATH like this: [00:32] PYTHONPATH=../../../ussoc/build-on-linux/:../../controlpanel/trunk/:. ./run-tests [00:33] DiegoSarmentero: that will fail with [00:33] exceptions.AttributeError: 'module' object has no attribute 'windll' [00:33] i'm building a branch to fix it, but basically grab ubuntuone_installer/gui/qt/local_folders.py and make default_folders return a fixed list of one folder [00:35] nessita, ok... i'll try it! [00:35] nessita, because in windows i couldn't run the test ever [00:35] nessita, i'm setting up some other environments in different vms, with win7 too [00:42] DiegoSarmentero: the needed changes in the windows-installer branch are lp:~nataliabidart/ubuntuone-windows-installer/run-test-in-linux [00:43] nessita, ok, downloading that now....... [00:43] nessita, question..... [00:43] DiegoSarmentero: shoot [00:43] nessita, where do i have to set the pythonpath in the installer? [00:44] nessita, it is loaded in a file property anywhere? [00:44] DiegoSarmentero: when running the tests, for example: [00:44] or just at the os level? [00:44] PYTHONPATH=../../../ussoc/build-on-linux/:../../controlpanel/trunk/:. ./run-tests [00:44] nessita, ahhhhhhhh okok [00:47] DiegoSarmentero: if that works for you, would you review https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/build-on-linux/+merge/68311 and https://code.launchpad.net/~nataliabidart/ubuntuone-windows-installer/run-test-in-linux/+merge/68310 ? [00:47] nessita, ok! [00:47] ralsina: if you have time, I would reviews from you for those above ^ :-) [00:50] nessita, YES!! IT WORKS!!! [00:50] nessita, that was... LEGEN.. wait for it... DARY! jeejjejej [00:51] DiegoSarmentero: lolol [00:51] DiegoSarmentero: all HIMYM is great until season 6, when the sitcom becomes SO boring [00:52] nessita, wait wait!! i'm in season 2!!!!! [00:52] ejejjeje [00:52] DiegoSarmentero: ooooooh enjoy then, just up to season 5 is all way up [00:52] nessita, jeje ok...... i'll do that :P [00:52] nessita, now i'll review the branch [00:53] both [00:53] great [00:53] DiegoSarmentero: most of the changes are to match up the controlpanel's build class [00:53] Night everyone! [01:56] ralsina: estás? [02:30] nessita, are you still there? [02:31] nessita, i have submitted more code in my windows installer branch....... now the tests are ok!! :D [02:34] DiegoSarmentero: ok, I'll re-review tomorrow [02:34] I m very tired now, I'm about to go to sleep [02:34] nessita, great! :P [02:34] DiegoSarmentero: see ya tomorrow! [02:34] nessita, bye [02:34] bye all! [03:57] ralsina, ping? [05:04] y-2-kids [06:53] hello folks [06:53] I have a question - is it possible to unsync a folder within a synced folder? [08:40] morning all!!! [09:42] Morning! [09:43] morning karni [09:43] \o [10:38] *yawn* morning all [10:38] morning duanedesign [10:45] * mandel away, bbl === mandel is now known as mandel_afk [10:48] hello fagan [11:18] rye: good morning! [11:20] rye: updating some of my canned responses. When a user needs to remove their Ubuntu One Token is it still necessary for them to not only remove the token in Seahorse but also visit https://one.ubuntu.com/account/machines/ and remove the computer? [11:23] duanedesign, when a user is removing the machine on Natty the control panel will notice this and ask for reauthorization, when machine is removed from earlier releases from ubuntuone-preferences this is not happening and the token needs to be removed from the server too [11:36] rye: thank you that answers it === mandel_afk is now known as mandel [11:40] * mandel back [11:40] mandel: hi there! [11:40] nessita: hello [11:41] oh hey nessita [11:41] nessita: shall we take a look at the brach? [11:41] mandel: have some minutes to talk about the listdir implementation? [11:41] yeap [11:41] nessita: we think alike :) [11:41] :-) [11:41] nessita: primate chat in spanish? [11:41] rather than making noise here :) [11:41] lol [11:41] sure [11:41] :) [11:55] well really im the only regular in here at the moment that doesnt speak spanish at all [11:56] so it would only be noise to me [11:56] :) [12:36] rye: here you go: http://ubuntuone.com/p/10qw/ [12:37] rye: ups, wrong channel, but is no a big problem :) [12:37] wow holy crap I responded to a bug report and a guy's auto responder is flooding emails back and forth [12:38] Oh lp deletes posts that are spamming didnt know that [12:38] thats pretty awesome [12:39] fagan, it may not be deleting them but we need lp people to unsubscribe him [12:39] fagan, because lp is sending notification back to his mailbox and it autoresponds [12:39] and we send notification again === 77CAAV1DS is now known as karni [12:41] fagan, i asked on #launchpad [12:41] rye: well I went to the bug and it was gone so I just presumed lp deletes them [12:41] ill go ask === m_conley_away is now known as m_conley [12:56] Standup in 4 [12:56] I think only mandel, nessita and I are actually here though [12:59] lazy people…. [12:59] vagos, que sois unos vagos! [12:59] :P [12:59] mandel: we would all prefer to be sleeping I know it [13:00] moi [13:00] me [13:00] oh alecu nice timing [13:00] me [13:00] ralsina: you feeling better? [13:00] alecu: say me [13:01] dobey, ralsina: stand up? [13:01] oh dobey is back forgot about that [13:02] me [13:02] fagan: go [13:02] we might as well go and let the others catch up [13:02] DONE [13:02] * Did some bug reports (mainly just changing statuses of bugs that affected ubuntuone-client and ubuntuone-client in ubuntu so they can time out) [13:02] * Phone call [13:02] TODO [13:02] * More of the 3 months unassigned bugs (if nothing else is needed) [13:02] mandel: go [13:02] BLOCKED [13:02] * Nope [13:02] DONE: Fix fix-os-helper based on the comments form nessita. Played around with the idea of using FlushFileBuffers to ensure that the file system notification work on the tests... failed miserable at it. [13:02] TODO: Finish changes in fix-os-helper after a long chat with nessita about ? :P [13:02] BLOCKED: no [13:02] alecu, te toca! [13:03] * alecu is writing notes [13:03] or nessita, don't know... [13:03] mandel: my turn was before you :-) [13:03] DONE: bug #810656, syncdaemon-analysis mumble, daily mumble, lots of reviews, proposed a couple of branches for having SSO building on linux and installer tests passing on linux [13:03] TODO: bug #810662, push forward mandel's branches, meetings [13:03] BLOCKED: nopes [13:03] NOTE: I'll leaving right after mumble to go to the Uni to evaluate some students [13:03] NEXT: mandel^W alecu [13:03] Launchpad bug 810656 in ubuntuone-control-panel "QT UI: Preferences tab do not show the loading overlay (affects: 1) (heat: 6)" [Medium,Fix committed] https://launchpad.net/bugs/810656 [13:03] Launchpad bug 810662 in ubuntuone-control-panel "QT UI: "Delete" button in the devices tab does not work (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/810662 [13:03] alecu: tu [13:04] DONE: a branch to fix #811307; helped __lucio__; reviews [13:04] TODO: probably work on one of bug #803672 and bug #803669 [13:04] BLOCKED: no [13:04] Launchpad bug 803672 in ubuntuone-client (Ubuntu) (and 2 other projects) "Do not start two instances of this process (affects: 1) (heat: 6)" [Undecided,Invalid] https://launchpad.net/bugs/803672 [13:04] Launchpad bug 803669 in ubuntuone-client (and 1 other project) "Get the port numbers from the users' registry (affects: 1) (heat: 6)" [Low,Triaged] https://launchpad.net/bugs/803669 [13:04] alecu: I think you fixed bug #811007 with one of your latests branches, did you? [13:04] Launchpad bug 811007 in ubuntuone-client "Syncdaemon is not exposing the "connect" method on windows (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/811007 [13:04] alecu: and I reviewed the unbreak-connect, I asked some info there [13:05] nessita, then I wrote the wrong number in my report [13:05] nessita, nice [13:05] alecu: I think you got the numbers right, that bug report may have been fixed as a side effect [13:05] nessita, mandel: I need to take amelia to kinder right now; I'll be back for the mumble meeting [13:05] alecu: cool, I'll be at lunch :) [13:09] morning [13:09] nessita, you were right: the bug you mention was fixed in a previous u1-client branch. [13:09] sorry about being late, I had a bad case of gastritis last night, standup coming up in 1' [13:09] holas al todos [13:10] ok, I'm off: see you guys in half an hour or so. [13:10] alecu: ack [13:11] ralsina: feeling better? [13:11] nessita: yeah [13:11] λ DONE: holiday, finished bug #809561, started reorg plan [13:11] λ TODO: reorg plan, shim code [13:11] λ BLCK: None. [13:11] Launchpad bug 809561 in ubuntuone-dev-tools "Support ignoring paths in u1lint (affects: 1) (heat: 6)" [High,Fix committed] https://launchpad.net/bugs/809561 [13:11] nessita: after I get the medicine it goes away in 2 hours or so, the problem is I was hurting from 2 to 6 AM :-( [13:11] ralsina: you don't have some medicine at home? [13:12] nessita: I do, but sometimes I need a shot [13:13] nessita: usually, a pill is enough, but last night I had the pill and was feeling worse and worse so I ended going to the doctor at 4-something AM [13:13] nessita: Got the shot, was well in 5 minutes [13:14] DONE: sprint planning, worked in bug #800376 (making some progress), calls [13:14] Launchpad bug 800376 in ubuntuone-windows-installer "Add "syncing your computer to the cloud" page to the installer (affects: 1) (heat: 8)" [High,In progress] https://launchpad.net/bugs/800376 [13:14] TODO: fix what needs fixing on my pending branches, finish 800376 [13:14] blOCKED: no [13:18] Ok, work. Nessita, do you still need reviews on the branches you mentioned last night? [13:20] ralsina: yessir [13:25] nessita: ok, I have the mgmt call in 5' will check them right after === zyga is now known as zyga-food [13:26] ralsina: okis [13:31] * nessita -> panadería, brb === mbarnett` is now known as mbarnett [13:41] * nessita is back [13:41] facundobatista, hi, do you recall the bug # for syncdaemon saturating the link due to bw throttling code logic? [13:43] rye: bug #720707 [13:43] Launchpad bug 720707 in ubuntuone-storage-protocol "Bandwidth limit is not correctly enforced: Transmission delays are inserted between data chunk writes (of arbitrary sizes) (affects: 22) (dups: 5) (heat: 82)" [Medium,Confirmed] https://launchpad.net/bugs/720707 [13:43] nessita, awesome, thanks! [13:43] :-) [13:44] and I'm back! [13:45] alecu, nessita, mumble, right? [13:45] nessita, mandel, ralsina, Chipaca: mumble? [13:45] ralsina: you off the management call? [13:45] alecu: Chipaca and I are n another call [13:45] probably until about 11:30 ART [13:46] we always forget... [13:46] who wants to crash a conference call? [13:46] ralsina, so, in 45 more minutes. cool [13:46] rye, ah, I was searching for it [13:46] alecu: I answered your needs info in https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix-810053/+merge/67894 [13:47] ralsina, cool, will look at it. [13:47] facundobatista, my awesomebar has forgotten some of the bug reports so I am relying on search which does not include storage protocol :( [13:47] rye, I filtered #chicharra for ubuntuone-client, too, that's why I wasn't finding it :( [13:51] * nessita uses firefox search engine [13:56] /win 11 [13:56] damn irssi [13:56] fagan: I bet irssi thinks the same ;-) [13:56] nessita, ok, i know you know this one - the failure to create UDF in case other folder startswith the same characters. [13:57] rye: you should know that is fixed in trunk, but let me grab the bug for you [13:57] first i lost half of bugs during migration to ff4 then some more to ff6 [13:57] every time i upgrade firefox i become dumber [13:57] nessita: I love irssi but I do tend to make mistakes changing window sometimes which is very funny :) [13:57] bug #649945 [13:57] Launchpad bug 649945 in ubuntuone-client "Incorrect FolderCreateError (UDF's can not be nested) (affects: 4) (dups: 1) (heat: 23)" [High,Fix committed] https://launchpad.net/bugs/649945 [13:57] rye: ^ [13:58] nessita, thanks again! [13:59] :-) [14:01] alecu: when you have a moment, let me know about https://code.launchpad.net/~alecu/ubuntuone-client/unbreak-connect/+merge/68309 [14:03] nessita, sure. [14:11] alecu: if I wanted to see if an object is iterable we agreed that getattr(blah, '__iter__', None) is what we want to do, right? [14:11] mandel, ??? [14:11] mandel, did we agree on that? [14:12] alecu: well, in using getattr rather than hasattr, right? [14:12] I always get confused with this... [14:12] mandel, oh, yes. hasattr == caca [14:13] mandel, the bit about "iterable" is up to you. [14:13] alecu: ok, I though so.. [14:13] mandel, well, strings are iterable, but they don't have __iter__ [14:13] mandel, and unicode objects don't have __iter__ either. [14:14] mandel, on the other hand, lists, dicts, sets, tuples and custom made sequences do have __iter__ === zyga-food is now known as zyga [14:16] alecu: yeah. I was going after list, dicts, setc etc… and ignoring str and unicode because they do not have __iter__, is that hacky? [14:17] mandel, I don't understand what you'll be using this for, so I don't know if it's hacky :-) [14:17] mandel, if you want to consider str and unicode, you should do this: [14:18] try: iter(n); return true; except TypeError: return false [14:18] but use True and False :P [14:18] alecu: I'm writing a decorator that will make sure that all os_helper method that return paths or a list of paths do return them as str and wanted to use a single decorator... [14:19] you record I should use two diff decorators? [14:19] nessita: input welcome ^ [14:19] mandel, two, absolutely. [14:19] alecu: ok [14:19] mandel, I don't like apis in python that take a string or a list of strings as the same parameter. [14:19] mandel: I agree with 2, but for future reference: from collection import Iterable [14:20] isinstance(something, Iterable) [14:21] nessita: oh, nice [14:22] mandel, http://stackoverflow.com/questions/1952464/in-python-how-do-i-determine-if-a-variable-is-iterable [14:22] alecu: they copied me! :-P [14:25] mandel, I like the idea of doing this in a decorator, because (in a few months) we may short-circuit it when we need more speed, and only leave the check if DEBUG=True [14:25] mandel, but again, let's do that in a few months, after we have seen this all is working for some time. [14:25] :) [14:26] mandel: did you chat with alecu re: listdir? [14:26] hmm [14:27] yes, we are turning python into a static typed language. [14:28] alecu: so we have no option other than doing bytes -> unicode -> listdir -> bytes? [14:28] nessita: I did not... [14:29] mandel: you should! :-) alecu knows a lot, and I'm not 100% happy with the idea of building unicodes to listdirs [14:29] alecu: main problem is that you use listdir(byte) you loose some information that is needed to mangle the path so that illegla paths are synced to windows [14:29] alecu: let me get you the branch with the implementation [14:30] alecu: https://code.launchpad.net/~mandel/ubuntuone-client/fix-os-helper/+merge/68227 [14:31] mandel, looking [14:31] cheers! [14:34] alecu: the problem that we are working around is shown here: http://paste.ubuntu.com/647224/ [14:37] the evil detail is in the alst line [14:37] s/alst/last [14:41] mandel, I don't understand what's the evil in that [14:43] alecu: mumble? it does make more sense spoken than writen… nessita and I have already gone through that pain :) [14:43] mandel, ok [14:43] I'll join you [14:52] ralsina: you available for mumble now? [14:52] Chipaca: ^ [14:52] nessita: on the phone, will be in 5' [14:58] ralsina: you still in the management meeting? [14:58] nessita: no, phone with my mother [14:58] there. done :-) [14:59] Chipaca: you available to join us? [15:00] nessita: now i am [15:00] Chipaca: awesome! [15:00] nessita: yes, I am. [15:01] mandel: mumble ping? [15:01] nessita: so, my internet went nuts… did I miss something, is everyone in silience? [15:02] mandel1: we're waiting for you, can you hear uis? [15:02] no... [15:02] let me restart mumble [15:02] yeap [15:02] seems that its reconnect function is not great [15:03] ya oigo! [15:08] how to loose a team member in the first day => make him share a room with mandel [15:10] also mandel how to kill a team member drink with mandel [15:12] :P [15:15] nessita, alecu: pushing the branch [15:17] alecu, nessita: in spanish: ralsina no se quire quedar con el culo al aire… [15:17] :P [15:17] ralsina: I think is possible for tuesday :) [15:17] or earlier [15:30] nessita, alecu: I'm going to run the tests on linux, if lint and tests are ok, the branch is ready for review [15:40] bbiab, lunch [15:42] Need to head out a tad early ill make it up tomorrow morning I have to run to get something before 5 [15:46] alecu, nessita: please take a look at the smaller yet monster MP: https://code.launchpad.net/~mandel/ubuntuone-client/fix-os-helper/+merge/68227 [15:47] mandel: looking [15:48] mandel: this branch is bigger and bigger! [15:48] nessita: thx! [15:48] mandel: can you please split this in: [15:48] mandel, reluctantly looking! [15:48] nessita: yes… it started small but we kept improving it [15:48] * remove abspath only one branch [15:48] * add tests for access and can_write, another branch [15:49] * add decorators to os_helper, another branch [15:49] is that doable without too much trouble? [15:49] * mandel cries... [15:50] nessita: well, I can take a look at how to do it.. but is 17:50 here and I prefer not ro delay this anymore [15:51] mandel: ok, I can propose a couple of smaller branches based on yours [15:51] otherwise we can make mistakes reviewing since this is too big [15:52] nessita: how can I help? [15:52] mandel: keep working on your other branches :-), I may ping you for reviews [15:52] nessita: ok, will do that then [15:54] alecu: I'll split mandel's branches in smaller ones, so you may wanna pause the review [15:56] karni: http://askubuntu.com/q/53654/711 (btw) [15:56] * karni opens [15:57] Chipaca: ummm...... My post has been deleted. "We have this on our TODO list as it has also been requested via support email. Stay tuned!" - I suck at answering? [15:57] Maybe I should have started with "Hi! I'm Software Developer at Canonical" [15:58] karni: meh [15:58] karni: let me figure what's up with that [15:59] Chipaca: I can repy again, in a more formal way. [15:59] Chipaca: the author removed the answer [16:01] Chipaca: I improved my answer. [16:03] karni: looks like he deleted it again? [16:03] Chipaca: I edited the deleted answer and clickd "undelete" for the moderator to look at [16:04] Chipaca: I think getting this undeleted rather than answering again should be better? [16:04] * mandel is happy , the assertions in os_helper have started to be useful :) [16:05] Chipaca: Not to mention it's not even really a question heh ;) But that's just me picking on the guy. It should get undeleted soon. [16:06] mandel, alecu: trivialish -> https://code.launchpad.net/~nataliabidart/ubuntuone-client/no-custom-abspath/+merge/68418 [16:06] * mandel goes fast as lighting [16:07] ack [16:08] karni: check it out now :) [16:11] nessita: looks great, but can you remove comment in line 50 of the diff, I should have done it but I for got, the one that starts with '# TODO: because tritcask does not work well with \\?\ we are adding it [16:15] Chipaca: hahahah :) good [16:15] mandel: sure! [16:16] nessita, the branch looks great. I'm running the tests now. [16:18] mandel: comment removed (I removed a couple similar ones) and pushed to revno 1057 [16:18] nessita: ok, I've checked it on windows and you have a +1 from me [16:19] great [16:19] I'll be away 10 min, will be back though === mandel is now known as mandel_bbl [16:26] nessita: Diego's branch has a needsfixing from you: https://code.launchpad.net/~diego-sarmentero/ubuntuone-windows-installer/installer-ui/+merge/67788 [16:26] ralsina: hum, let me confirm [16:26] ralsina: not anymore! [16:28] nessita: have links to your branches handy? I can do some reviews now [16:28] ralsina: sure!@ [16:28] https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/build-on-linux [16:28] https://code.launchpad.net/~nataliabidart/ubuntuone-windows-installer/run-test-in-linux [16:28] ralsina: ^ [16:29] nessita: on it! [16:35] nessita: +1 on both, I took the liberty of setting them to approve [16:35] ralsina: awesome! [16:35] thisfred: did you see my note on your branch re: the merge failure? [16:36] nessita, approved the abspath branch. [16:36] alecu: thanks! [16:36] dobey: yup, fixed, just rerunning the tests [16:36] thisfred: ah ok, cool [16:37] dobey: and fixed [16:39] dobey: so could you remove your needs fixing? :) [16:40] alecu: sorry to bother you, but could you check https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix-810053/+merge/67894 [16:41] alecu: it's my "big" pending branch, I don't want that one to diverge [16:42] nessita: you promised me an approval for https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/show_spinner/+merge/68145 (the one with the puzzling click/clicked thing) [16:42] ralsina, looking [16:42] alecu: thanks! [16:43] ralsina: you're right [16:44] nessita: thanks! [16:47] thisfred: done [16:48] thx [16:49] ralsina, can I ask you to review this smallish branch? https://code.launchpad.net/~alecu/ubuntuone-control-panel/use-status-changed/+merge/68117 [16:49] alecu: of course! [16:50] alecu: I tested that on friday! MUst have forgotten to add the review :-( [16:54] ralsina, nessita: sorry to be bothersome, but I still don't understand why we need to have the ubuntuone-windows-installer running in the ubuntu-sso-client process: [16:54] https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix-810053/+merge/67894 [16:54] alecu: re-tested and +1 [16:54] alecu: because it's how SSO works. [16:54] alecu: is a mix about how sso is designed and how ralsina designed the wizard [16:54] alecu: the SSO UI always runs in the SSO process. [16:55] And since we need to have the SSO widgets inside the wizard... [16:55] alecu: basically, in the SSO API, we state [16:55] alecu: "give me your app name, your tc_url, etc, *and* your UI class and module and I take care of everything" [16:55] ugh! [16:55] alecu: the problem is that the UI module from the wizard is the wizard itself [16:56] nessita, ok, I understand that. [16:56] in gtk this is not so bad because you can embed widgets across apps [16:56] instead of having the auth process "separated" from the rest, which from my POV is what we should do [16:56] ralsina, right [16:57] The other way to do it would be to do most of ussoc's job in the installer process (which is how it was) but that was failing because it didn't do the ping. Rather than add the ping, we decided to go this way. [16:57] nessita, ralsina: anyway, I agree that the UI for this looks fine. I just wanted to understand why we are introducing this weird module dependency loopback [16:57] alecu: well, it's not really a loop [16:57] alecu: since sso nly depends on installer when used *by* installer [16:57] ralsina: you can't embed the qt widget in the qt app? :) [16:58] dobey: not accross process boudaries [16:58] mandel_bbl, alecu: another split from mandel's branch at https://code.launchpad.net/~nataliabidart/ubuntuone-client/add-tests-access-can-write/+merge/68428 [16:58] ralsina, hmm.... ok. [16:58] ralsina: surely qt has a plug/socket API [16:58] ralsina, approving. [16:58] dobey: nope. I mean, I *could* use ActiveX on windows to do this, but, really? [16:58] although i don't know what you mean by "boundaries" there [16:59] dobey: instantiating one widget on one process and its parent widget on another [16:59] ralsina: why would you use activex? is that what qt uses on windows to provide XEmbed functionality? [16:59] dobey: you understand xembed doesn't work on windows, right? ;-) [17:00] the way to provide "embedded apps" in windows is activex, yes. Or some variation of what used to be OLE [17:00] ralsina, approved. [17:01] alecu: cool, thanks [17:01] ralsina: xembed itself doesn't (well, unless you have X running also), but surely there is some API that provides the same functionality in qt, like there is in gtk+ [17:02] dobey: not that I know of [17:03] dobey: google tells me there is a reparent() call though [17:04] dobey: but only works inside a process :-( [17:05] oh, hmm [17:05] gtk+ only has plug/socket on X11 target it seems. guess it isn't ported to win32 then :( [17:07] dobey: that is really non-trivial to do on windows. SDL has it, though [17:07] ralsina: nevermind me then. but could stick the widget in a library and just talk to the backend over IPC, and embed the widget [17:07] ralsina: i wouldn't call it trivial to do on X11, either :) [17:07] ralsina: I'm testing the show_spinner IRL, can you please confirm when I should see the overlay 'loading'? [17:08] nessita: when you click on "login" for example, or in "set up account" [17:08] ralsina: when loading the TOS as well? [17:08] nessita: that is another branch [17:08] ah ok [17:09] ralsina: and when the overlay should go away? [17:09] because I submitted the form with empty fields, got the loading, got the errors, clicked on "ok" in the error warnings and the overlay will never go away [17:10] nessita: ohhhhhhhh ok, didn't try that! [17:10] nessita: nevermind then ;-) [17:12] ralsina: always try as many workflows as you can think of, remember 2 things: - users are "evil" - we don't have a separated QA process [17:12] ralsina: so we need to ensure the quality ourselves [17:12] nessita: yeah [17:12] nessita: I tried all the "positive" paths, missed the error ones [17:15] * mandel_bbl back [17:16] mandel: just a quick question: https://bugs.launchpad.net/ubuntuone-client/+bug/810521 [17:16] nessita: I'm back doing some work in the other branches, if you need a review et me know [17:16] Launchpad bug 810521 in ubuntuone-client "Typo in signal name (affects: 1) (heat: 343)" [High,Triaged] [17:16] mandel_bbl: I do, https://code.launchpad.net/~nataliabidart/ubuntuone-client/add-tests-access-can-write/+merge/68428 [17:16] mandel_bbl: if you tell me the names of those signals I will fix it :-) [17:17] mandel_bbl: there are some test failures in windows in listdir, the rest is green === mandel_bbl is now known as mandel [17:18] nessita: you mean in that branch? [17:18] ralsina: what's the problem [17:18] mandel: I need reviews for that branch, yes [17:18] nessita: and in that branch, there are errors in listdir, right? [17:19] mandel: one signal is written "on_equest" and I am guessing that's a missing r. Another one is called "request_queue_removed" which doesn't seem to follow the convention for the others ("on_request_whatever") [17:19] mandel: yeah, also in trunk [17:19] mandel: but all that is without your os_helper fixes [17:19] so I guess it makes sense? [17:19] nessita: sounds like it [17:21] ralsina: indeed, those look like typos [17:21] mandel: ok, I was not sure about the second one [17:22] teh signal is on_... [17:22] mandel: ok, will propose the trivial fix in a few minutes [17:23] ralsina: hurray! [17:23] mandel, alecu, nessita, DiegoSarmentero: we will have lots of syncdaemon help in the sprint from chicharra visitors [17:24] woohoo! [17:26] toma!!! [17:26] DiegoSarmentero: do not get scared by the spaniard while at the sprint, I'm like that always :P [17:26] nessita: running tests, if the pass will approve [17:27] mandel: well, they all will not pass [17:27] mandel: listdir test is failing here, remember [17:27] nessita: the ones you added do not pass? [17:27] they do! [17:27] ok, then if those pass :) === beuno is now known as beuno-lunch [17:32] Hello whitehats.. Hi guys.. does anyone here know if my synced files will get deleted from the ubuntu one cloud server upon subscription expiration? [17:33] mandel: trivial branch for review: https://code.launchpad.net/~ralsina/ubuntuone-client/fix-810521/+merge/68435 [17:33] wh1zz0: no they won't. You won't be able to upload anything until you are under the free 2GB though [17:33] wh1zz0, they won't be deleted right now, but this may change in the future. [17:34] Okie.. so if I have like 20gb and subsciption expires, I won't lose anything right, only thing is I might not be able to upload more, how about download existing files? [17:36] wh1zz0: that should still work [17:36] Thanks a bunch ralsina :) Ubuntu rocks! [17:41] mandel, jejeje sorry i was eating..... spookie! jeje [17:42] nessita: still waiting for the branc to finish… dammed 3g! [17:42] estupida isla sin internet! [17:43] DiegoSarmentero: :P [17:51] ralsina: did you run the test on windows? [17:51] nessita: approved, or so I believe... [17:51] 3g, not reliable... [17:51] mandel1: I did but I never know which ones *should* work today [17:52] ralsina: ok, I can give a code review.. but branching is taking aaaaaages :( [17:52] DiegoSarmentero: when you propose a branch, always remember to set the commit message field, or things never merge [17:53] ralsina, ok... sorry about that [17:53] DiegoSarmentero: we all forget every now and then :-) [17:53] that is what she said... [17:53] ralsina: I dare you to guess what I'm talking about ;) [17:54] mandel1: that's what you get for dating alzheimer-addled senior citizens [17:54] ralsina: hehe good one, but mine was waaaay dirtier :P [17:54] ralsina: and I don't like sushi when talking about this things.. [17:54] creo que me pase con esa... [17:55] * ralsina points in the general direction of the employee manual [17:55] I guess that is the EOD for me or I might be spelled due to the code of conduct... [17:55] ralsina: hahaha [17:56] a2! [18:02] ralsina, i saw that there are some conflicts trying to merge the branch, i'll review it as soon as i reach home [18:02] ralsina, is that ok? [18:06] rye: i have a user that is getting I/O error with tritcask metadata on an encrypted $HOME. Is the best workaround to delete the offending metadata and do a rescan? [18:08] DiegoSarmentero: sure [18:25] rye: found what i needed. [18:26] duanedesign, no no no, wait, is he on ecryptfs? [18:27] ralsina: can you review https://code.launchpad.net/~nataliabidart/ubuntuone-client/add-tests-access-can-write/+merge/68428 ? [18:28] nessita: can it wait 1 hour? I am about to have lunch [18:28] ralsina: sure [18:28] rye: yes, i believe so [18:28] gr [18:28] * rye is grrring [18:28] rye: i traced some of his error messages to bug #372014 [18:29] Launchpad bug 372014 in ecryptfs-utils (Ubuntu Karmic) (and 4 other projects) "errors in dmesg (dup-of: 509180)" [Low,Confirmed] https://launchpad.net/bugs/372014 [18:29] Launchpad bug 509180 in linux (Ubuntu Oneiric) (and 5 other projects) "ecryptfs sometimes seems to add trailing garbage to encrypted files (affects: 115) (dups: 9) (heat: 334)" [Medium,Fix released] https://launchpad.net/bugs/509180 === beuno-lunch is now known as beuno [18:39] duanedesign, awesome bug report, ecryptfs... why, oh why? [18:46] duanedesign, finished reading. splendid :-/ [18:47] verterok, in case tritcask file cannot be opened due to I/O Error, what happens? [18:48] rye: I think there is a bug about that /me search for it [18:48] rye: if I remeber correctly I think it dies if the broken file is the live data file [18:48] rye: yeah i just finished reading 509180 . Looks like a mess [18:49] verterok, okay, so it dies... so it does not write to that file? ecryptfs appends garbage to the file under some circumstances, which later on introduce IOErrors when reading. [18:49] rye: garbage is not the same as broken [18:49] beuno: I couldn't debug my issue, but I've found bug #813071 in the way of analysis the code [18:49] nessita: Bug 813071 on http://launchpad.net/bugs/813071 is private [18:49] duanedesign, have you understood how can this be recovered, is unmounting the private directory enough? [18:50] rye: tritcask should survive reading a file with garbage, it will not survive if it can't open the file [18:50] verterok, kernel sends IOError on open, nothing to read in this case [18:50] nessita, so not a total loss! I'll assign it to vds` [18:50] beuno: :-) [18:50] rye: then, the file is broken :) [18:51] verterok, ok, in the worst case, can we delete live tritcask metadata w/o some really awful consequences? [18:52] rye: it will cause the loss of that metadata [18:52] rye: image not being able to open a sqlite journal :) [18:52] all those changes are gone [18:53] verterok, well, yeah, what will happen next? :) [18:53] rye: one of the comments mentions "copying the lower file somewhere else allows me to recover the original file" [18:53] rye: need to manually recover? :) [18:54] rye: there is a bug about a specifc error where the live data file is broken (python can't open the file), which should be simple to fix [18:54] duanedesign, somewhere else?... Hard to imagine what "recover" means in this case, copy is just a read()... [18:54] rye: how to recover from that? the only way is to loss data or start over (remove everything and resync ) [18:54] rye: with data I mean "metadata" :) [18:55] verterok, in our case it can't open a file because it is overencrypted (i.e. encrypted and then a set of zero bytes appended) [18:55] rye: the kernel can't open the file? [18:56] verterok, the ecryptfs layer returns I/O error during open [18:56] rye: so, it's broken :) [18:57] as fopen fails :) [18:58] rye: remove the file, it's the only way to make it work [18:59] rye: the fix for the bug will be to discard the file, and create a new one [18:59] and log a big warning message [19:00] duanedesign, so yes, it looks like removal and creating new file is a workaround [19:01] beuno, you recently encountered a big crash of ecryptfs, has anything went missing or IOErroring ? [19:03] rye, not AFAICT [19:03] rye: thank you [19:06] bug #776386 [19:06] Launchpad bug 776386 in ubuntuone-client "File Sync error (IOError with tritcask metadata) (affects: 9) (dups: 2) (heat: 36)" [High,Incomplete] https://launchpad.net/bugs/776386 [19:08] is the U1 bug on this issue^ [19:13] alecu: can you do a review for me? https://code.launchpad.net/~nataliabidart/ubuntuone-client/add-tests-access-can-write/+merge/68428 actually is mandel's code [19:23] nessita, approved. [19:24] alecu: thanks! [19:24] and now I'll get me some food. [19:24] alecu: enjoy [19:41] nessita: I pushed show_spinner again with fixes for the error cases (all I could find, at least). https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/show_spinner/+merge/68145 [19:42] ralsina: with tests for those cases? :-) [19:42] I added a "critical" method that is used for popups [19:42] and that method is tested [19:43] I added a new controller for "forgottenpassword" that hides the overlay. That needs a test (sorry) [19:44] I noticed as I was writing that ;-) [19:44] ralsina: is ok, let me know when that's done [19:44] nessita: sure [19:54] nessita: pushed show_spinner with the missing test [19:55] ralsina: ack [19:57] nessita: need any reviews ? I have a slot now [19:57] ralsina: yes! [19:57] https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/letmeremove/+merge/68444 [19:57] nessita: on it [19:57] thanks [20:08] nessita: I get an insecurejelly error when running that branch IRL [20:08] http://pastebin.ubuntu.com/647596/ [20:09] is this the thing about clearing credentials failing that you mentioned in email? [20:09] ralsina: the traceback under 2011-07-19 17:06:49,773 - ubuntuone.controlpanel.backend - ERROR - process_unauthorized (clearing credentials): is this bug #813066 [20:09] nessita: Bug 813066 on http://launchpad.net/bugs/813066 is private [20:10] ralsina: what comes next, is clear_credentials failing, yes [20:10] nessita: if not, it looks much like it [20:10] alecu: have you gathered enough jellyfish-foo to confirm or deny that ^? [20:11] I do have tons and tons of repeated devices here, though, so may be unique to me [20:36] nessita, ralsina: that *should* be fixed when my latest branch lands [20:36] alecu: cool [20:36] nessita, ralsina: btw: do you have a windows 7 64bits and an xp? [20:36] I have a 7 64 bits and a xp [20:36] alecu: only 32bits win 7 [20:36] but my 7 has 32 bit python [20:37] ralsina, cool. Can you try this on both 764 and xp? http://pastebin.ubuntu.com/647568/ [20:37] alecu: sure [20:37] alecu: on 7, I get 1001 [20:37] ralsina, it should print the user id (for instance "1001" on my 732) [20:38] xp is going to take a few minutes [20:38] ralsina, I can wait :-) [20:39] alecu: I forgot , my xp VM is "clean" and has n python [20:39] no python [20:40] ralsina, ok, got it. Don't worry, I'll just assume this works on xp as well :-) [20:40] alecu: leave a message for fagan to try and tell you early tomorrow [20:40] he has a XP devenv setup [20:41] ralsina, cool [20:41] nessita: when you got 5 minutes I need to cry in your shoulder :-/ [20:41] ralsina: any other hint, so I can prepare psychologically? [20:43] nessita: have control panel trunk and lp:~ralsina/ubuntuone-windows-installer/fix_800376_2 handy [20:43] I *think* I am not doing anything wrong but I get a crash, without traceback, somewhere inside controlpanel when I use volumes_info [20:44] ralsina: why would you use volumes_info in the wizard? [20:44] :-) [20:44] nessita: it's called by validate_path_for_folder [20:44] then volumes_info calls sd_client.get_root but that fails [20:45] ralsina: do you have sd running? [20:45] nessita: I thught we had activation for that [20:45] but I have tried it with sd running [20:45] ralsina: not sure about the activation thingy [20:45] ralsina: ok, let me branch those [20:45] nessita: ok [20:46] ok, I need to be afk for 40 minutes or so. [20:46] will be back laters [20:46] cheers! [20:46] bye alecu! [20:48] ralsina: did you pushed to that branch? [20:48] I'm getting [20:48] ** Merging the branch [20:48] Nothing to do. [20:48] hmmm [20:48] ah wait [20:48] I have a typo [20:48] ralsina: false alarm [20:49] nessita: just in case, I did forget to push [20:49] lol [20:49] pushed now [20:50] ralsina: merging that branch in installer trunk generates 3 conflicts [20:50] Text conflict in ubuntuone_installer/gui/qt/gui.py [20:50] Text conflict in ubuntuone_installer/gui/qt/local_folders.py [20:50] Text conflict in ubuntuone_installer/gui/qt/main/windows.py [20:50] argh [20:50] then just branch the branch [20:51] ralsina: can you fix the conflicts? I think is better if I debug "the latest" code. If it's too much trouble, I ll branch the branch [20:51] nessita: it's ok, I'll fix them [20:57] ralsina: new review comments and questions added to show_spinner [20:57] nessita: ack === m_conley is now known as m_conley_away [21:04] nessita: pushed fix_800376_2 with the conflicts solved [21:05] ack, branching [21:06] alright all, i'm out. have a good evening :) [21:07] ralsina: ok, wanna point me to some piece of code? [21:08] nessita: to make it as easy as possible, comment out the call to _load_folders in line 93 [21:08] Then try it, and when you get to the local folders page, click on "add a folder". That will cause the crash [21:08] ralsina: parethesis: why you overwrote my changes to def default_folders(self)? :-) [21:08] ralsina: that was needed to be able to run the tests in linux [21:08] nessita: oops, sorry, I will revert that [21:09] since this ctypes.windll.shell32 is too window oriented :-) [21:09] ralsina: ok, I'll branch this in windows, I though no IRL was needed [21:09] nessita: well, unless you can figure out why it crashes by looking at it :-) [21:10] ralsina: well, the first thing I see is that you're not yielding on _load_folders [21:10] that's why I said, ignore it [21:10] I mean, comment it [21:11] ralsina: but if I comment that out, what crash will we have? commenting out that line is commenting out the crash, no? [21:11] ralsina: or I'm misunderstanding? [21:11] just using the button I take from controlpanel crashes it [21:11] ah, let's see [21:14] ralsina: I'm haivng delays since sso is now crashing [21:14] ok, if I let it crash there, I do get a traceback about not connecting to syncdaemon :-/ [21:16] so it seems the problem is activation is not working (http://pastebin.ubuntu.com/647641/) [21:19] naty, diego's branch has conflicts he was fixing them later at home [21:19] nessita: ^ [21:19] ralsina: ah, oh [21:19] ralsina: I saw 2 approves and I approved :-/ [21:19] yeah, I had tried that earlier :-) [21:19] ralsina: you marked as neesd fixing again? [21:20] there. I marked it [21:21] ralsina: this is almost impossible to test: I need to have credentials in order to have syncdaemon running, but I can't have credentials in order to the installer to show up :-/ [21:21] nessita: well, you should be able to start syncdaemon on the "successful login" page [21:21] ralsina: I think is time we fix this nightmare, and I think I have a quick-to-implement idea [21:22] ralsina: ok, idea -> later, I'll do that [21:22] nessita: that is, of course, even better ;-) [21:24] nessita: I am cnvinced with my latest trace that it's just not connecting to syncdaemon so I'd say it's not worth checking this [21:24] ok [21:24] is crashing for me too, with no trace at all [21:25] as long as it's not failing because of something I do wrong. Should I start syncdaemon manually before showing that page? [21:25] ralsina: not sure what you mean [21:25] nessita: u1cp manages to connect to sd. Obviously it's doing something installer is not doing [21:27] ralsina: not really, is just using SyncDaemonTool which, works... [21:27] ralsina: I'm not sure that the control panel activates SD though, I personally never tested it [21:27] nessita: ok [21:31] I know it does, because I can start control panel without anything else running, and it works [21:31] and syncdaemon starts running [21:31] but that's because I have .exes and the right keys in the registry [21:32] ok, I'll write the report and call it a day [21:36] There, reported, and EODing. I may put a couple hours late, depending on how sleepy I am (I slept like 4 hours last night) [21:37] Have a good evening and see you all tomorrow! [21:38] ralsina: you too [22:01] ok, I'm off crowd [22:01] see ya tomorrow! === zyga is now known as zyga-afk