[05:05] In a lot of my folders there is this header for Ubuntu One. Is there any way to remove it? I dont want to sync those folders [07:01] How do i get the nautilus integration to shut up? lol [07:01] i dont want to share these folders [07:07] hello slide [07:07] hello [07:09] slide: i think it is right-click in the folder and select Ubuntu One > Hide Ribbon [07:10] aaah yea thanks! [08:09] Good morning all! [08:09] hello JamesTait [10:22] * fagan break === teknico is now known as teknico_away [11:50] good morning people! [11:51] morning ralsina [11:52] So weird not having anyone online till lunchtime [11:52] now I know how mandel felt :) [11:55] fagan: It's because everybody's busy working :P [11:56] karni: and in argentina :) [11:56] * fagan was working too [11:56] fagan: you work with people from Argentina only? [11:56] karni: mandel is in spain [11:56] fagan: what do you mean 'in argentina', a sprint? [11:57] Mandel had some plane problems and will probably not be around today either [11:57] karni: mandel is the only other european in desktop+ [11:57] karni: well the entire team is in Argentina except me and mandel normaly [11:57] ralsina: hah! :) [11:58] but yeah there was a sprint too so was pretty lonely for me last week but I found stuff to do and learn [11:58] And he was in argentina until like 3AM your time today [11:59] ralsina: I didnt actually tell you what I did last week either did I, I was reading those books I got like the twisted one [11:59] ralsina: and writing up my report that I have to submit to my college [11:59] which is pretty huge [11:59] ralsina: I see [11:59] fagan: cool [12:00] ralsina: I was actually doing a bit of that earlier today too since I didnt have a specific thing to do [12:00] fagan: I guess interns can do that - I would feel guilty if I read (even Android related) books on company time ;) [12:00] karni: yeah I was kinda in need of learning [12:01] karni: and I didnt have any assignments so made sense :) [12:01] fagan: I'm in need of learning all the time. [12:01] karni: well, interning *is* an educational thing [12:01] ralsina: Indeed :) [12:02] ralsina: exactly ive learned stuff even when I wasnt expecting which is awesome [12:02] like mainly how companies work internally is great to learn alone [12:02] karni: and I am supposed to be training him and not doing much of it :-( [12:03] ralsina: Then he's excused for reading books! hahah ;D [12:03] ralsina: well it is incredibly busy because of the windows client [12:03] buenos días! [12:03] ok, gotta reboot back to windows, be back in a bit [12:03] I read a great book written by a guy who worked as a project manager at digital the company that flopped in the 80s [12:04] hi nessita , and bye! =D [12:04] oh crap, it was ralsina rebooting lol [12:04] It was all about software development practices that were used and how to budget, manage and all that [12:04] scratch that nessita , hello! :D [12:04] hey nessita [12:05] hi karni, fagan [12:05] fagan: srsly? that must have been boring. I'm reading up recently on TDD and testing practices, exciting!! [12:05] fagan: I mean, budget, manage.. bleh ;d [12:05] karni: well this was pretty interesting because it gave me insight into how people in upper management work to o [12:05] *too [12:06] like how they make the decision to go with a project [12:06] or not go with a project [12:06] and loads of other stuff [12:06] fagan: I had a separate class about organizations and management in college. Man that was boring. [12:06] karni: well this one had a load of stories and all that too [12:06] uhh.. peculiar bug me haz [12:06] so it was fairly interesting to read [12:06] fagan: cool! [12:07] and I could skip over the bits I did learn in college [12:07] like who needs to read about waterfall [12:07] :D [12:07] * fagan prays he never has to do waterfall [12:08] * fagan grabs some lunch before standup [12:13] hello #ubuntuone! [12:14] hello alecu ! [12:16] buenos ddos alecu [12:16] :D [12:17] fagan, "good distributed-denial-of-services"??? [12:17] alecu: yep :) [12:17] fagan, what does ddos mean to you? [12:18] alecu: well it ruined my favorite game but I was changing the spanish language for comic effect :D [12:19] oh, right. [12:19] buen dia alecu! [12:19] * alecu laughes [12:19] ralsina, buen día jefe! [12:19] how you doing today alecu? Better? [12:20] ralsina, yup. better. not muuuuuch, but better. [12:21] alecu: I missed the explanation because of IRC trouble [12:21] alecu: For Android 2.0+, this is what we'll be using in Ubuntu One Files as far as login/registration is concerned: lp:ubuntu-sso-java-client lp:ubuntu-sso-android-client (latter is an Android Library). They still need some love, but you can see how they're used in lp:ubuntuone-android-files - apart from regular Android library usage (pasting stuff to AndroidManifest etc), I did use a separate activity to tell the user (s)he needs to ... [12:21] ... add an account. Also, since this uses AccountManager, if there's U1Files or (soon) U1Music installed on the device already, there's even no need to use any of the two - a 3rd party app can simply request the token directly from AccountManager, and once the user grants access, it can use the OAuth token cached in AccountManager. Currently, although implemented, we've disabled registration in lp:ubuntu-sso-android-client, because it ... [12:21] ... requires copying the token from email, and pasting into the app, which is not user friendly. Instead, we pop the browser, until we have better SSO registration experience via the API. [12:21] ralsina, yup, I saw you disconnecting just as I said "hello!" [12:21] wow, that's a beefy message! [12:22] karni, nice!!!! [12:23] alecu: :) [12:23] karni, it makes me want to start up eclipse again and change droidcouch to use it. [12:24] * alecu has not started eclipse in at least 4 or 6 months. [12:24] alecu: I spent a little time making it also Ant friendly :) All the instructions are in HACKING.txt in each project [12:24] alecu: hah =) [12:25] ralsina, so, I've seen that mandel left a few branches up for review... [12:25] alecu: hi there! [12:25] alecu: yes, we're needing a review for https://code.edge.launchpad.net/~mandel/ubuntuone-client/fix-aq-tests-windows/+merge/69997 [12:25] alecu: yes [12:25] nessita, looking [12:25] and he won't be around today, so we should try to move them without him if needed [12:26] ack [12:27] ralsina, btw, regarding bug #803672 [12:27] Launchpad bug 803672 in ubuntuone-client (Ubuntu) (and 2 other projects) "Do not start two instances of this process (affects: 2) (heat: 22)" [Undecided,Invalid] https://launchpad.net/bugs/803672 [12:28] ralsina, did it happen when running sd using python, or using an .exe bundle? [12:28] * ralsina is all ears [12:28] alecu: python, but I have the exe in the place where the registry says [12:29] haven't tried it with just exe [12:29] ralsina, cool. I'll take a look after reviewing these branches. [12:29] ralsina, I have not tried with the .exe yet either, so I was curious, because using python I didn't got that error. [12:30] ralsina, but I'll try with trunk first, because some other thing might have changed. [12:30] ralsina, question: now that we are reimplementing setTitle and setSubTitle, should I remove the "title" and "subtitle parameters in the constructors and use just that function from outside the class, or leave both options, and use setTitle, etc inside the class too? [12:30] alecu: ok [12:32] DiegoSarmentero: we should avoid all code duplication, and so I think that we need to remove all the parameters from the classes otherwise we need to repeat the if to hide/show the labels in every class, right? [12:32] I say leave the parameters and then call setTitle / setSubtitle [12:33] ralsina, DiegoSarmentero: can we have those methods named with snake_case#? [12:33] ralsina, yes, that was what i was doing... so i leave the parameters and call that function instead the if [12:33] nessita, we could named with snake_case in the controller [12:34] not in SSOWizardPage [12:34] DiegoSarmentero: this is the sso-client branch, right? [12:34] because we are reimplementing Qt methods to hide the actual behaviour [12:34] nessita, sso and wizard [12:34] not yet ready [12:34] DiegoSarmentero: so, the setTitle is a method we define, right? [12:35] or is it inherited from Qt? [12:35] nessita, redefine [12:35] inherited [12:35] ah, ok [12:35] hey folks [12:35] am having problems with 'save to phone' [12:35] simply its not really working for me [12:36] if I just check one song, then hit 'save to phone' the downloading icon appears next to the track but keeps stopping and starting [12:36] and struggles to complete. [12:37] I find the only way i can get tracks onto my phone is to 'play' them, which caches them on the phone, (is that what the orange star means?) then, whilst cached, check and save to phone. [12:38] nessita: inherited methods, so camelCase [12:40] karni: ^ [12:40] alecu: do not bother doing the second review, I don t know why mandel proposed that since we agreed we'll skip file_shelf test suite [12:40] hmm [12:41] fagan: thank you [12:44] j0nr: You have filed a bug about this, right? [12:44] j0nr: Yes, orange star == song is cached [12:45] j0nr: I'll pass this on to Chad, who's been taking care of Ubuntu One Music. He might already know what's wrong. [12:45] j0nr: Thank you for reporting this. [12:47] nessita, by second review you mean this? https://code.edge.launchpad.net/~mandel/ubuntuone-client/fix-file-shelf-windows [12:47] alecu: yes [12:47] nessita, and I don't recall either: why skip it? [12:47] alecu: file_shelf is something used in old versions of u1client [12:48] alecu: and I'm not comfortable with the fix that mandel added using os.path.split [12:49] 10 minutes to standup [12:49] since is hiding the fact that fd.name returns different things in windows and in linux [12:52] nessita, also, that branch is missing the dependency on the previous one. [12:52] karni: ok :) [12:52] nessita, I can work on a branch to skip those tests after the fix that ralsina requested for activation. [12:52] j0nr: his nick is CardinalFang, he should be around soon. [12:53] alecu: thanks, but I'm adding that to my fix-tests branch. I have another task for you, if interested [12:54] nessita, shoot [12:54] alecu: either set_dir_readwrite and/or set_dir_readonly are not doing what they should. Calling readonly and then readwrite, do not restore the perms in a dir [12:54] alecu: I need to file a bug [12:55] nessita, great, point it to me, and I'll take a look. [12:56] alecu: bug #820350 [12:56] Launchpad bug 820350 in ubuntuone-client "WIndows: either set_dir_readwrite and/or set_dir_readonly are not doing what they should (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/820350 [13:00] me [13:00] me [13:00] me [13:00] dobey, nessita, alecu === m_conley_away is now known as m_conley [13:00] me [13:01] me [13:01] fagan, go [13:01] DONE [13:01] * more of my report for college [13:01] * Tried to figure out what was going on with 11.10 on my computers so I can file bugs (still haven't figured it out but going to fresh install and see if its still a problem) [13:01] TODO [13:01] * finish my report [13:01] * help with what ever any one wants [13:01] BLOCKED [13:01] * By bugs in 11.10 [13:01] ralsina: go [13:02] ralsina, ? [13:02] DiegoSarmentero: go! [13:02] DiegoSarmentero, go [13:02] might be having a connection issue again [13:02] DONE: [13:02] Modify SSO and Wizard to work with a reimplementation of "setTitle" and "setSubTitle". Test to validate this new implementation almost complete. [13:02] TODO: [13:02] Keep working on Installer with embed Control Panel branch. And UI bugs from Installer. [13:02] BLOCKED: [13:02] No. [13:02] DONE: bug #819917 bug #819878 bug #811246 bug #810053 TODO: bug #814686 and more [13:02] BLOCKED: no, NEXT: DiegoSarmentero [13:02] nessita, go [13:02] DONE: half day off, reviews, more fixing test_vm: found out that either set_dir_readonly and/or set_dir_readwrite are not doing what they should, so shares dir can not be removed [13:02] TODO: make more tests pass [13:02] BLOCKED: nopes [13:02] NEXT: alecu [13:02] DONE: took a sick day [13:02] TODO: reviews, work on bug #820350 [13:02] BLOCKED: not today [13:02] NEXT: dobey [13:02] Launchpad bug 819917 in ubuntu-sso-client "show_gui.py doesn't work anymore (affects: 1) (heat: 6)" [Undecided,Fix committed] https://launchpad.net/bugs/819917 [13:02] Launchpad bug 819878 in ubuntu-sso-client "On windows: "python setup.py install" fails (affects: 1) (heat: 6)" [High,Fix committed] https://launchpad.net/bugs/819878 [13:02] Launchpad bug 811246 in ubuntuone-windows-installer "The Control Panel is not opened in the last step (affects: 1) (heat: 10)" [High,Fix committed] https://launchpad.net/bugs/811246 [13:02] Launchpad bug 810053 in ubuntuone-windows-installer "Credentials obtained by the installer don't work. (affects: 1) (heat: 9)" [Critical,Fix committed] https://launchpad.net/bugs/810053 [13:02] Launchpad bug 814686 in ubuntu-sso-client "QT UI: 'Sign in' button can be clicked even if not values were entered, same happens with verification code (affects: 1) (heat: 13)" [Medium,In progress] https://launchpad.net/bugs/814686 [13:02] Launchpad bug 820350 in ubuntuone-client "WIndows: either set_dir_readwrite and/or set_dir_readonly are not doing what they should (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/820350 [13:03] LAGGGGGG [13:03] ralsina, stop leeching! [13:03] alecu: hehe === teknico_away is now known as teknico [13:04] I never asked how is the internet in AR? [13:04] any comments? [13:04] alecu, ralsina, Chipaca: are we having mumble today? [13:05] nessita, I'm -0 to that [13:05] Na I dont have comments [13:05] I'm ok with a mumble [13:05] Oh ralsina do you want to proof read my report when I have it done [13:06] fagan: usually, the same thing as everywhere else. In fact my IRC problems are because of a server in the US that is acting up [13:06] ralsina: ahhh ok [13:06] fagan: sure, send it over [13:06] ralsina: the report doesnt have anything internal on it its just a brief about the company [13:06] ralsina: cool [13:07] the only thing that would have been dodgy I asked about and im not going to write about it [13:12] me [13:12] λ DONE: yet more shim work [13:12] λ TODO: finish initial shim [13:12] λ BLCK: None. [13:28] brb [13:39] * nessita back [13:50] alecu, ralsina, Chipaca: shall we mumble? [13:56] grr, python-aptdaemon is being super annoying. [13:57] if it ain't one thing... [13:57] alecu, ralsina, Chipaca: hello? [13:58] nessita: in futures call [13:58] nessita: can it be in a while? I am having some local trouble and chipaca is in another call anyway [13:59] Chipaca, ralsina: would you please let us know when you're available? === m_conley is now known as m_conley_away [14:00] I will be in about 30 minutes [14:16] nessita, ralsina I need reviews for this branches: https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/titles-on-pages/+merge/70169 AND https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/titles-on-pages/+merge/70157 [14:17] DiegoSarmentero: on it! [14:17] ralsina, thanks [14:17] DiegoSarmentero: sure [14:18] nessita, great... now everything should be ok [14:19] ralsina, nessita mmmmm it seems that launchpad didn't rescan my branch yet [14:19] DiegoSarmentero: no problem, I'm branching locally [14:19] DiegoSarmentero: it takes a while sometimes [14:19] DiegoSarmentero: is mumble working for you? [14:20] nessita, no, about that...... is there a wiki or something where it explains how to configure... i don't know if you have to setup any special server or something... and i couldn't find it on the wiki [14:20] DiegoSarmentero: yes, let me give details in private [15:14] Chipaca, ralsina: hello bosses! any news re: mumble so we can schedule our lunches? [15:15] nessita: I am free now [15:16] lunch time here. bbiab :) [15:18] launchpad is not scanning branches. That means we can't merge stuff easily, right? [15:20] ralsina: not sure if LP is that related to tarmac [15:21] Anyway, if it ever comes back... https://code.launchpad.net/~ralsina/ubuntu-sso-client/validate-harder/+merge/70320 [15:21] ralsina: so, shall we schedule a time for mumble? [15:21] Chipaca: ^ [15:21] nessita: 39 minutes from now would wfm [15:23] Chipaca: sounds great. alecu, DiegoSarmentero, ralsina: meeting at 1pm ART? [15:23] nessita, ack [15:23] nessita, ack. Let's get some lunch then. [15:23] yeah! [15:23] As I said, I am free now [15:24] ralsina: would you be at 1pm? [15:24] ok, 1PM it is [15:45] nessita, ralsina, DiegoSarmentero: anybody has seen this? http://pastebin.ubuntu.com/658005/ [15:46] alecu: I did. You need to remove your xdg and install as the wiki says [15:46] ralsina, great, thanks. [15:50] alecu: ping [15:51] ralsina, great, thanks. [15:51] alecu: test_eq_inotify is failing for me in trunk, with a scary error regarding threads and a thread pool: http://pastebin.ubuntu.com/658012/ [15:51] alecu: any clues? [15:52] @ping [15:52] pong [15:53] nessita, there's nothing scary! [15:53] alecu: is scary to me [15:53] alecu: do you understand what's wrong? [15:54] nessita, the threadpool part of the stacktrace just says that it's failing in the bit that's being run on the watch thread [15:54] nessita, the other error I remember seeing it with mandel, when trying to open *something*. Not necesarily a file, but something. [15:56] alecu: so, shall the watch thread handle that? [15:56] nessita, so: it's trying to open the directory to set up the watch, and either the path is bad or the directory is no longer there. [15:56] nessita, it's the watch thread that's failing. [15:56] right [15:56] nessita, _watch is the function being run inside the thread. [15:56] right [15:56] nessita, and on the first line of that function it's failing [15:57] nessita, so, yes: we should somehow handle this error. [15:57] nessita, but right now it looks like an issue in the way the test was set up. [15:57] nessita, is that happenning on trunk to you? [15:57] alecu: yes [15:57] alecu: is it for you? [15:58] nessita, I'm running those tests now. [16:00] ouch, I forgot TRIAL_TEMP_DIR [16:01] nessita, all tests passed on trunk [16:01] alecu: among other things, I see calls to add_watch that are not being yield on [16:01] alecu: you up to date? seems like an ugly timing issue [16:01] mumble! [16:02] nessita, can you try setting TRIAL_TEMP_DIR to a clean folder? [16:02] alecu: yes, already did it, but I will do it again [16:02] nessita, what calls to add_watch are not being yielded on? [16:02] nessita, don't worry about that, I just tried with the same TRIAL_TEMP_DIR and it seems to be working. [16:03] alecu: confirmed is failing in a new temp dir [16:03] alecu: calls to add_watch inside test_eq_inotify [16:04] Chipaca: mumble? === m_conley_away is now known as m_conley === beuno is now known as beuno-lunch [16:30] ralsina: ping [16:30] hola mandel! [16:31] mandel, where are you now? [16:31] alecu: hola!!! [16:31] mandel: pong! [16:31] alecu: in madrids airport trying to get back home… I lost my connection flight and have to wait for 3 more hours [16:31] mandel, :-( [16:32] ralsina: hello, did you guys take look at my branches, I'm going to put some hours from the airport :P [16:32] mandel, we approved the first one, and rejected the second [16:32] alecu: yeah, is a pain, but I'm getting used to it, only bad things is that I smell a little hehehe [16:32] mandel, we are mumbling right now [16:32] mandel, hahahaha [16:32] mandel, yes, it sucks. [16:33] alecu: what was wrong with the second one? [16:33] * mandel goes to launchpad [16:34] mandel, nessita reminded us that the bits in the second branch are only used in old versions of syncdaemon. [16:34] alecu: funny thing, is my bday and I'm spending it at the airport, I feel like a real business man :P [16:34] mandel, oh!!!! right!!!! [16:35] happy birthday, Mr. Vice-President of Windows! [16:35] alecu: yes, indeed they are just used in old version, but there are tests everywhere that depend on those, and I think is better to have as many green ligths as possible :) [16:35] alecu: hehe, I'm vc of nothing ;) [16:35] mandel, she suggested skipping those tests on windows [16:36] alecu: that involves skipping tests for fsm too, which is a pain and is suppose to be multplatform [16:36] also, having deprecated code that does not work is dangerous in the case we ever want to use it, specially when it has already been fixed [16:36] mandel, and also she commented there's an issue with masking a filename that's absolute on windows, but relative on linux. [16:36] mandel, or something like that. [16:36] mandel: I don't like the adding to os.path.split [16:37] nessita: why? [16:37] mandel: if fd.name is returning different stuff in each platform, we need a more complex solution [16:37] nessita: is just in the test of a fake unpickle method, not in the actual code [16:37] not just masquerading this with os.path.split [16:37] mandel: then why fd.name returns different stuff in each platform? [16:38] (if it's a fake) [16:38] * DiegoSarmentero brb... lunch! [16:38] nessita: because open_file has the windows decorator which converts the path in an abspath with the \\?\ [16:39] mandel: and also, I don't understand how the test_fsm depends on stuff that are not being used un windows [16:39] nessita: yes, that is why I started fixing fileshelf, otherwise I would have not touched it [16:39] mandel: can we make test_fsm not depend on file_shelf for code that is currently being used [16:39] ? [16:39] nessita: we could, indeed [16:39] mandel: I think that is the solution [16:40] otherwise we're not testing what we're running [16:40] nessita: better than a solution that has already been implemented? [16:40] I think I am going to take a rather long lunch and put back the hours later, because I have some errands. I should be back in 2 or 3 hours :-( [16:40] nessita: I'd say we are testing all the API that sd provies, that it is not used I really don't care much (to a level ofcourse) [16:40] mandel: but, 2 things: [16:41] ralsina: before you go, swap day tom for me? [16:43] * fd.name should return the same in every platform. If that is not the case, we should fix why and not hide the issue using os.path.split [16:43] * you said that if we skip the file_shelf tests, we need to also skip test_fsm becasue test_fsm depends on that. So, I say, how come we're depending on running tests for test_fsm, which is a feature we do use in windows, on things that we're not running in windows (file_shelf)? [16:43] we want test_fsm to run all the suite using the features we will run in windows, no? [16:43] nessita, regarding the issue in the _watch function: can you print self._path before calling CreateFile? Also, print os.path.exists(self._path) and os.path.isdir( [16:43] self._path) [16:44] alecu: what is the problem with _watch? [16:44] mandel: sure, poor baby ;-) [16:44] alecu: in which file should I add the print for self._watch? [16:45] nessita, ubuntuone/platform/windows/filesystem_notifications.py [16:45] nessita: fd.name does return the same in every platform, the issue here is that the fd returns the path that was passed to open, because we use open_file, for reasons we both know, we call open with an abs path with \\?\, which makes fd to set the name to the abspath, we can change the code to call open with the abspath and then you will have the same tests, which is a compromise I'm willing to have [16:45] nessita, at the start of def _watch(), just before the call to CreateFile [16:45] nessita: since maybe using os.path.split was not the correct way to fix the tests [16:46] nessita, btw: CreateFile is in fact being used to "open a directory". talk about api naming... [16:46] mandel: that makes sense, but, I'm now worried that test_fsm will not work without file_shelf. We need test_fsm to work without it, since we will use it without or [16:46] alecu: on it [16:47] nessita: indeed, there ate tests in test_fsm that probable do not belong there, they seem to be testing metadata issues that should clearly be placed somewhere else, but we should ask chicharra about that [16:47] mandel: so we can skip those since are not fsm-related [16:48] nessita: yes, we skip the metadata ones in fsm and move forward, yet we fix the fileshelf tests no to use os.path.split (easy fix) and once those to are landed we talk with chicharra about why do we have this types of tests mixed [16:49] nessita: maybe moving them to test_metadata.py is a better approach rather than having them in tst_vm and test_fsm [16:49] I recall there were some metadata tests in test_vm too, but I might be wrong [16:50] mandel: vm metadata and fsm metadata are different metadata [16:50] *very* different [16:50] nessita: are they… joder!!!! care to explain me the diff? [16:51] mandel: let me confirm I'm not lying to you [16:51] facundobatista: ping [16:52] nessita: ok :) [16:52] mandel: it used to be for sure, but now with tritcask I'm in doubt [16:53] nessita: yeah, would be nice to have deprecation warnings etc.. so that we could know, could be an easy bug for chicharra [16:54] alecu: http://pastebin.ubuntu.com/658041/ [16:54] alecu: anyways, I see bad setupping in that test... [16:54] ok, lunchtime for me [16:56] alecu, nessita: that is easy to fix, call CreateFileW [16:57] alecu: the issue is that you are calling a unicode path with the CreateFile function instead of the unicode one from COM, which ends in W [16:57] alecu: mainly, puto COM, puto Windows, puto MSDN, yet easy to fix === jderose1 is now known as jderose [17:01] alecu: in case you need the docs: http://msdn.microsoft.com/en-us/library/aa363858(v=vs.85).aspx do a search for unicode (if the page is in english) [17:01] mandel, the very weird thing is that it works ok for me, and I'm also using unicode literal paths there. [17:02] alecu: really? dammed, well nevertheless we should be using CreateFileW [17:02] alecu: I though that nessita was using E: as here volume label, am I wrong? [17:05] nessita, let's do what mandel says: changing CreateFile to CreateFileW [17:10] @ping [17:10] pong [17:19] aquarius, heyo. I'm getting not infrequent 500 (and even occasionally 503) errors when uploading deja-dup backup files to U1. How do I go about helping to fix that? [17:19] mterry, I know vds has been working on a problem with upload failures -- vds? [17:19] mterry, what's the body of the error responses? === beuno-lunch is now known as beuno [17:20] aquarius, I haven't dug that out yet, would have to do some work to get it [17:20] mterry, a 503 is a temporary error on the server (that is: it might be due to overloaded database or something), and can be retried. [17:20] Which I'm happy to do, just haven't yet [17:20] aquarius, we do retry 5 times and take the last error [17:20] so I guess should just wait a bit [17:20] or back off longer [17:21] perhaps, yeah. We're working on some DB overload issues exactly now. Maybe exponential backoff or something rather than 5 in quick succession? [17:21] mterry, aquarius: I thin vds is off for 2day, maybe you can add somewhere a question for him to take a look… [17:21] a 500 -- it would be most useful to have the error body so we can see what the U1 server is complaining about [17:22] aquarius, I also note that last I checked the server returned generic 500 for "out of space". I think I filed a bug about that. Though that's unrelated to this issue [17:22] mterry, do you have an oops id? [17:22] vds: o/ hello [17:22] mandel, hello! :) [17:23] mterry, that means you're out of quota [17:23] vds, not on hand. I assume that would be in the body? [17:23] mterry, yes [17:23] vds, 500 always should mean out of quota? I seem to be getting it when not [17:23] vds, do we return oops ids in JSON responses? [17:23] mterry, having an oops id would be very useful [17:23]