[00:00] ralsina, ping..... just one question if you are still around [00:00] gatox: sure [00:00] ralsina, the script python bootstrap.py...... never ends?? just to know... because it stay at: "Generated script .... buildout" [00:01] gatox: it seems to go to background [00:01] gatox: after a bit, see if you have a bin\buildout.exe and that means it's finished :-) [00:01] ok thanks [00:31] gatox: ok, ese error es que para instalar twisted se necesita Visual Studio. Parece que voy a tener que hacer el egg primero [00:31] gatox: mañana la seguimos ;-) [00:32] ralsina, okas! no te jodi por aca porque crei que ya no estabas [00:33] no estoy ;-) [05:15] If I delete something locally, will Ubuntuone delete it too? [05:21] !delete [05:21] Factoid 'delete' not found [05:53] n2diy: yes [05:56] Ok, so it isn't a good idea to keep my backup box and this box synced to Ubuntuone, if I mess up this one, so goes the backup box!? [05:57] lifeless, thanks for being the only concious and alert life form on this channel! :) [06:01] yes, keeping in sync is u1's story :) [06:21] lifeless, roger that. I'm using it as an offsite backup, so... I need to understand how to play with it. === eu is now known as Guest108 [10:08] morning all! [11:01] good morning mandel! [11:01] ralsina, morning! [11:01] ralsina, I'm finishing your review right now, I have some small comments but most of it looks great :) [11:01] mandel: awesome [11:02] mandel: it's big enough that everything can' be right :-) [11:02] can't [11:02] ralsina, + I wrote some ot if hehehe [11:03] ugh, that is new, I swap letters between words.. [11:04] ralsina, done, comments written, let me know if all of them are clear and if you agree/disagree :) [11:06] mandel: the replace of @FOO@ is so it works like on linux [11:06] +1 on removing the glob import [11:06] or rather, let me work a bit on it and give a full answer [11:08] ralsina, sure, I also did this: https://code.launchpad.net/~ralsina/ubuntuone-control-panel/even_more_unique_than_the_other_snowflake/+merge/86912 [11:08] ralsina, that one is approve, yet you have to remove your name from the header since it was greed to do that in all files.. [11:08] although I miss seeing my name everywhere :( [11:08] mandel: yes, I am removing it, must have missed some [11:08] ralsina, just two of the, not much work :) [11:09] ralsina, and it is approve so you can approve the merge as soon as you removed those :) [11:09] mandel: cool :-) [11:10] ralsina, one question, do you know what is dirspec.basedir ? [11:10] mandel: dobey's relacement for a bit of xdg [11:10] mandel: s we can remove that crap from sso [11:11] ralsina, ah, superb, can it be replaced everywhere, I'm looking at https://code.launchpad.net/~diegosarmentero/ubuntuone-client/fix-links/+merge/86605 and Basedir is there [11:11] mandel: can I ask you a favour? Can you check if all the files in snowflake have the right EOLs? Something like find . -exec file {} \; | grep CRLF [11:12] mandel: we'll do it eventually, I suppose [11:12] ralsina, sure [11:13] ralsina, one question, do you know how to config squi3 not to use http1.1 ? [11:13] Is keeping the connections alive and I'm getting dirty reactor issues :( [11:13] mandel: want to not have keepalive? [11:13] mandel: I have a vague notion, give me 1' to check [11:13] ralsina, thx! [11:14] keepalive=1,1,1 (that should close any connection in about 2 seconds) [11:15] and tcpkeepalive=1,1,1 [11:15] ralsina, where do I put that? [11:15] in squid.conf [11:15] ralsina, anywhere? cool, let me try :) [11:15] no wait [11:15] good morning [11:15] at the end of the http_port [11:16] like this, mandel: http_port 3128 keepalive=1,1,1 [11:16] hi gatox! [11:16] ralsina, hi [11:17] gatox, morning! [11:17] gatox, feeling better? [11:17] gatox, I did some of your reviews, please check the branches [11:17] mandel, ok.... yep, much better [11:18] ralsina, I'm getting the following: [11:18] FATAL: Bungled squid.conf line 73: http_port 58557 keepalive=1,1,1 [11:19] mandel: well, it's in the docs: http://www.squid-cache.org/Doc/config/http_port/ [11:21] ralsina, he, stupid docs, in one part it says keepalive in other it says tcpkeepalive [11:21] ralsina, is the later.. [11:22] mandel: it happens [12:01] mandel, sorry..... didn't understand this comment: "One question, can the xdg from sso be replaced by dirspec.basedir?" [12:03] good morning everyone! [12:03] nessita, hi [12:03] gatox, so, apparently there is a pacakge implemented by dobey that has some oof the functions from xdg and is called dirspec.basedir so that we do not all depend on sso [12:04] gatox, I'm wondering if we can use that instead [12:04] nessita, morning! [12:04] hello mandel, gatox [12:04] mandel, any reference to that package to take a look if it can be useful in this case? [12:04] gatox, AFAIK is in the nightlies ppa already [12:05] mandel, i mean..... do you have a link or something to take look into the code or how to use it? [12:06] gatox, hm.. no idea of the name of the project, ralsina do you know the lp project for dirspec? [12:06] mandel: python-dirspec I think [12:10] gatox, ^ [12:10] mandel, yep..... i'm looking at the code [12:19] mandel, gatox: when you have a minute to try it, no priority, I have a new and improved buildout at http://u1.to/ralsina/O/build.zip [12:20] ralsina, ok.... i'll try it now...... now i can open 2 vms at the same time (gatox cry) [12:20] ralsina, I'm fighting with twisted at the moment (and winning) I'll be on that asap :) [12:20] gatox, but you only have one monitor! [12:20] buahahahaha [12:20] mandel, you always have to ruin the moment! [12:20] jejjee [12:20] gatox, that is what she said ;) [12:24] 'lo all [12:24] hello again! (/me had internet issues) [12:25] gatox: Virtualbox has an awesome mode called "stretch mode". You can have two large VM screens visible at the same time if you don't mind that they look like crap :-) [12:25] hello nessita, duanedesign! [12:25] ralsina, na, i prefer to use one desktop for each of them [12:26] hello ralsina [12:27] mandel, gatox: I just saw your comment in gatox's branch, re replacing xdg with dirspec [12:27] mandel, gatox: we're not doing that for now, we will migrate everything to use it in separated branches [12:28] mandel, gatox: specially because the package is not available in the ubuntu repos (yet) [12:28] nessita, ok...... not looking at dirspec anymore [12:28] nessita, ok [12:28] gatox: you can look, but please do not make the change for now [12:29] mandel: did the keepalive change i squid make a difference? [12:30] nessita, question..... is this really necessary?? "Can you please add tests to os_helper to reflect the changes you have made to path_exists, make_link and read_link?"...... [12:30] nessita it was TDD we already have the tests.... i change the code and keep working :P jeje [12:30] ralsina, no, but I found a way to fix the issue, I implemented the HttpChannel protocol in twisted so that it kept a reference to the instance currently used, during the cleanup I force the timeout of the connection directly from the protocol [12:31] mandel: ok [12:31] gatox: that's the problem, you changed the code without adding a test first that fails [12:31] ralsina, that cleans everything and lets the reactor clean [12:31] gatox: so, the first step would have been adding a test where you assert that path_exists(some_link_without_lnk) will give True [12:31] ralsina, the good/bad things is that the issue was just happening in P because we have squid3 :) [12:31] gatox: and that would fail, and just then you can modify the code [12:31] mandel: shouldn't those be cleared if you delete the instances that open them? as part of gc? [12:32] nessita, you mean "windows/os_helper", don't you? [12:32] ralsina, yes, but then we have an unclean reactor exception from trial [12:32] mandel: oh [12:32] gatox: no? (maybe I'm not understanding you?) [12:33] ralsina, we want to have no selectables of delay callbacks in the reactor since it make the next tests to work in an unclean env [12:33] mandel: makes sense [12:34] ralsina, do you remember the sso test that where failing? it is a very similar issues [12:34] nessita, if the test should be for test_os_helper for windows only, or the platform independent version...... because the changes were for windows only... if it is for windows only i can test better the changes i think [12:35] gatox: no, the test should be multiplatform, and in linux that test will pass, since os.path.exists will return True for links in linux [12:35] ok [12:53] mandel: fix_809873 ready for another pass, comments answered [12:53] ralsina, ok, give me some mins and I'll get to it :) [12:56] mandel: no rush :) [13:01] nessita, ralsina do I have to test that the proxy tests work for versions earlier than N? [13:02] mandel: they should, yes [13:02] ok [13:02] mandel: not M [13:02] mandel: but yes L [13:02] Although M should be no problem [13:02] mandel: if they don't work, we should have a clear understanding of why they don't, so we can fix after feature freeze [13:03] nessita, I'm making sure they work with squid2 and squid3, I'll run the tests on the previous releases to check what happens [13:03] mandel: awesome [13:13] Hi, a quick ubuntu one question, I'm running out of disk space on my PC, but have loads of room in U1, if I delete a synced folder from my PC, will it delete from my U1 account too? [13:16] rye, can you help with this ^ [13:18] NickL, if you first make it stop syncing from your desktop, and then delete it, it will become a cloud-only folder [13:18] that is, only if the folder isn't under ~/Ubuntu One/ [13:18] NickL, yes, unsubscribe the folder locally [13:18] awesome. Thanks beuno, rye. [13:18] NickL, using ubuntuone-control-panel-gtk [13:20] nessita, ralsina FYI the branch that was blocking alecu has passed all the tests in NMO and P, looking at L and will propose for re-review [13:20] * mandel is waiting for his L machine to update [13:20] mandel, nessita when you have a minute, please, re-review this: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/fix-links/+merge/86605 [13:21] sure [13:22] mandel: cool [13:22] ralsina, i'm getting this error now: http://paste.ubuntu.com/792688/ [13:25] gatox: more eggs that need VS to build. I can fix it! [13:35] alecu, ping? [13:36] hm.. I'm off to lunch [13:36] lets hope they see this: [13:37] alecu, dobey I updated https://code.launchpad.net/~mandel/ubuntuone-dev-tools/proxy-testcase/+merge/85019 to work on P with squid3, I have ran the tests from L to P and passes in all platform, may I have a re-review (including running the tests) [13:37] * mandel lucnk [13:37] * mandel lunch === mandel is now known as mandel_afk [13:41] hello all! [13:43] meh [13:43] hola alecu, dobey [13:43] kid's up, I gonna be a dad for half an hour [13:44] mandel_afk, great news; re-reviewing. [13:52] hola alecu! [14:12] nessita, ping [14:15] gatox: pong [14:16] nessita, question..... tihs issue is only for gtk: https://bugs.launchpad.net/ubuntuone-windows-installer/+bug/865176 ..... am i right? [14:16] Launchpad bug 865176 in ubuntuone-windows-installer "Handle errors more gracefully, show user friendly message and hide technical details (affects: 1) (heat: 2)" [Medium,In progress] [14:16] nessita, sorry [14:16] this one: https://bugs.launchpad.net/ubuntuone-control-panel/+bug/870270 [14:17] Launchpad bug 870270 in ubuntuone-control-panel "Windows tool tip doesn't display if "&" is in the folder name (affects: 1) (heat: 6)" [Medium,In progress] [14:17] gatox: yes [14:18] nessita, ok [14:18] just to be sure [14:18] gatox: but what is the question? [14:18] gatox: I said "yes" meaning "I see the bug report" [14:19] ahhhhh [14:19] nessita, if it's only valid for gtk [14:19] it seems [14:20] gatox: have you tried in the QT control panel? [14:22] nessita, i was looking at the description of the bug (and it's looks like controlpanel in linux).... i couldn't add a folder in windows.... i'm trying [14:24] gatox: you can try using the qt control panel on linux [14:31] @ping [14:31] pong [14:32] internet connection is acting up, apparently [14:32] ralsina: shall we skype? [14:33] nessita: it's a bit noisy here, but let's try [14:33] ralsina: you still in mdq? [14:34] nessita: yes [14:35] nessita: I see you as offline on skype [14:35] gatox: that bug is about the gtk control panel, yes [14:35] ralsina: calling you [14:35] nessita: ca you hear me? [14:35] ralsina: yes [14:35] ralsina: can you hear me? [14:35] then I can't hear you [14:36] ralsina: give one sec [14:36] ralsina: can you please call me? [14:36] sure [14:37] ralsina: you can hear me? [14:37] no [14:37] :-( [14:37] will you have mumble later? [14:37] ralsina: one sec, will install padevchooser [14:37] ok [14:37] ralsina: not in this machine :-( [14:37] dobey, thanks [14:38] ralsina, bug #911792 is for you! [14:38] Launchpad bug 911792 in ubuntuone-client "Messaging menu turns blue when I get a share, can't do anything about it (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/911792 [14:38] beuno: thanks! [14:39] (has been happening for a while, but thought it was time to file it to fix it for P!) [14:39] ralsina: I have no idea why audio input is not working, will try a little bit longer t fix [14:40] nessita, alecu, mandel, dobey, gatox: very very very trivial branch: https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix_911784/+merge/87483 [14:40] nessita: no rush [14:40] ralsina, on it [14:42] ralsina: restarting skype made it work [14:43] ralsina: calling you, if you're available [14:43] nessita: sure, go ahead [14:44] ralsina, +1 === mandel_afk is now known as mandel [14:45] ralsina, +1 [14:45] * mandel back [14:46] alecu, so, the way i fixed was to implement a version of the HTTPChannel and force a timeout of the connection, how do you see that? [14:46] alecu, does it sound hacky? [14:47] mandel, yup. [14:51] alecu, whenever you take a look let me know, but is the only way I found how to solve the issue in a reliable manner with squid3 [14:51] alecu, I'm moving to other proxy bugs, from this: https://bugs.launchpad.net/ubuntu-sso-client/+bugs?field.tag=u1-proxy which ones do you have? [14:54] mandel, sorry, where did you put the timeout? [14:55] mandel, I see that you've added a teardown_client_server, and that looks good. [14:57] mandel: hello [14:57] alecu, grep for self.site.protocol.protocolInstance.timeoutConnection() [14:57] ahhh, cool. Looking. [14:57] mandel: I have a user that gets this error https://pastebin.canonical.com/57717/ [14:57] alecu, I think that is ok, but I'm not 100% sure and you are the twisted guy :) [14:57] duanedesign, hello! long time no see [14:57] duanedesign, let me check [14:57] mandel: i saw a bug you worked on that appears to be this issue [14:58] duanedesign, which package is he using? [14:58] mandel, that's perfectly fine: you are forcing the mock webserver to timeout when closing. (I understand a different thing) [14:58] duanedesign, he is missing the pem file from ubuntuone storage protocol, which are in the storage protocol trunk [14:58] mandel, in fact, it's a lovely solution. [14:58] alecu, yes, that is what I meant to say :) [14:59] mandel: would this cause the sync to get to 'stuck'. And possibly produce the "DeadReferenceError 'Calling Stale Broker'" error? [14:59] mandel, I think you've said it right the first time, I understood cualquiera. [14:59] alecu, I'm glad to hear that, I though I was being stupid, which is common lately :P [15:00] duanedesign, hm.. it could possible mean that, is he getting the Stale Broker error from control panel? [15:00] duanedesign, that error usually means that the is a reference of the IPC in one of the UI applications, and sd gc that object [15:00] me [15:00] mandel, two small needsfixing, but the branch looks great! [15:00] me [15:00] https://code.launchpad.net/~mandel/ubuntuone-dev-tools/proxy-testcase/+merge/85019 [15:00] alecu, superb! [15:00] me [15:01] @ping [15:01] pong [15:01] me too [15:01] nessita, what's with all the @pinging? [15:02] ok! [15:02] mandel: Not sure what version they are running. Should updating to the latest version fix the missing pem file? [15:02] me [15:02] nessita, what's with all the @pinging? [15:02] ralsina, dobey standup? [15:02] me [15:02] nessita, I saw the first "me" before you disconnected. [15:02] alecu: my internet connection is not working right today, so I don't know when I'm off IRC (cuz pidgin shows me online the same) [15:02] duanedesign, should, also, we need to know which one he is using and find out why the pems are not there... [15:03] kk, thanks [15:03] meh [15:03] DONE: [15:03] Added some tests to my branches, investigating some issues. [15:03] TODO: [15:03] Bug #865176, Bug #870270, and several more. [15:03] BLOCKED: [15:03] No [15:03] Launchpad bug 865176 in ubuntuone-windows-installer "Handle errors more gracefully, show user friendly message and hide technical details (affects: 1) (heat: 2)" [Medium,In progress] https://launchpad.net/bugs/865176 [15:03] Launchpad bug 870270 in ubuntuone-control-panel "Windows tool tip doesn't display if "&" is in the folder name (affects: 1) (heat: 6)" [Medium,In progress] https://launchpad.net/bugs/870270 [15:03] mandel, go [15:03] duanedesign, let me know if you need a hand [15:03] DONE: Finally found a nice wayt to use localhost address to test the proxy testcases so that the reactor is cleaned if we use http 1.1 with squid3. A number of reviews for gatox and ralsina. [15:03] TODO: Fix the comments inthe proxy test case from alecu. Move to next proxy bug. [15:03] BLOCKED: no [15:03] nessita, go [15:03] DONE: lots of reviews, releases, catch up [15:03] TODO: keep fixing the tests for bug #834730 [15:03] BLOCKED: nopes [15:03] NEXT: ralsina [15:03] Launchpad bug 834730 in ubuntuone-client (and 1 other project) "Use SignalBroadcaster and RemoteMeta from ubuntu sso client (affects: 1) (heat: 8)" [High,In progress] https://launchpad.net/bugs/834730 [15:03] DONE: kept on pushing the windows packaging branch, first attempt at buildout-thingy, fixed minor string bug, reviews, 1-1 natalia TODO: fix more buildout, fix more bugs, some bug triaging sprint planning BLOCKED: no NEXT dobey [15:04] λ DONE: fix crashers in P, releases [15:04] λ TODO: uploads, fixes for gireactor from review, initial music store work [15:04] λ BLCK: No. [15:04] comments anyone? [15:04] oh [15:04] gatox: are you feeling better? [15:04] nessita skipped alecu [15:04] alecu: go :) [15:04] oh, sorry, I missed that me [15:04] nessita, yep...... [15:05] DONE: gave two hands debugging, worked on gsettings+qt, reviews [15:05] TODO: land proxy integration branch, a bit more on gsettings parser [15:05] BLOCKED: on mandel's branch that's coming real soon now. [15:05] EOD? [15:05] nessita, i'm trying to track some bugs..... it should be an easy fix..... but the tracking part is kind of confusing... [15:05] oops EOM? ;-) [15:05] gatox: what do you mean? [15:06] ralsina, I don't mind EOD :P [15:06] gatox: did you see my comment re: testing the qt control panel in linux? [15:06] nessita, yep..... but i'm with another one before that one [15:07] ok..... i need to go to the bank right now to hit some people..... i'll be back in a while [15:09] nessita: the bug that was about, was specifically for the gtk control panel though. [15:09] alecu, I just pushed the changes to the branch with the comments of the review [15:09] dobey: right, but before closing, I want to be sure it works right on linux as well [15:09] alecu, can you let me know which bugs from the proxy tag are free for me to tackle [15:10] nessita: also, it's filed about 2.0.0, so we might need to fix it in SRUs for older versions as well anyway [15:10] dobey: unless we backport the qt control panel :-D [15:10] nessita: though perhaps a lower priority. i don't think closing it as "well we have a qt panel now" is the answer [15:10] i don't even want to imagine thinking about backporting the qt panel yet :( [15:12] and i don't have to think about that sort of stuff until march anyway, i think === gatox is now known as gatox_away [15:16] alecu, are you working on this guy: bug 884975 [15:16] Launchpad bug 884975 in ubuntu-sso-client "replace urllib2.urlopen with common web-client (affects: 1) (heat: 9)" [Undecided,New] https://launchpad.net/bugs/884975 [15:16] oh, i guess i should have had "TODO: File MIR for dirspec" also [15:17] mandel, bug #884972 and bug #884975 [15:17] Launchpad bug 884972 in ubuntu-sso-client "replace urllib2.Request with common web-client (affects: 1) (heat: 9)" [Undecided,New] https://launchpad.net/bugs/884972 [15:17] Launchpad bug 884975 in ubuntu-sso-client "replace urllib2.urlopen with common web-client (affects: 1) (heat: 9)" [Undecided,New] https://launchpad.net/bugs/884975 [15:18] I wonder what would happen if I file a bug that in the title says bug #thesamebug [15:18] and then mention it here [15:19] mandel, I'll tackle bug #884973 next, since I already did some stuff to replace lazr. [15:19] Launchpad bug 884973 in ubuntu-sso-client "replace lazr.restfulclient with common web-client (affects: 1) (dups: 1) (heat: 12)" [Undecided,New] https://launchpad.net/bugs/884973 [15:20] ralsina, a black hole? [15:20] alecu, so I can move to bug #884972 and bug #884975, right? [15:20] Launchpad bug 884972 in ubuntu-sso-client "replace urllib2.Request with common web-client (affects: 1) (heat: 9)" [Undecided,New] https://launchpad.net/bugs/884972 [15:20] Launchpad bug 884975 in ubuntu-sso-client "replace urllib2.urlopen with common web-client (affects: 1) (heat: 9)" [Undecided,New] https://launchpad.net/bugs/884975 [15:21] mandel, yay! [15:21] hi [15:21] with ubuntu one can i have 5 people all accessing the same logon account [15:21] alecu, superb, I'll move to them now, let me know if you see any other issues with the test case after revno 80 [15:22] dobey, can you please review the proxy-testcase branch whenever you have the time :) [15:23] mandel: yes, and when the insanely cold weather stops affecting neuron transmission :) [15:23] hi [15:24] uh, bye then [15:26] dobey, ok :) [15:26] dobey, and thx [15:26] it is not supposed to be this cold here [15:29] mandel, I found this lint error: http://pastebin.ubuntu.com/792779/ [15:29] mandel, but it's also present on u1-devtools trunk, so it's not caused by your branch [15:29] dobey: any idea if the above lint error will cause any problem? [15:31] alecu: no. that is supposed to be there. it makes sure that u1lint doesn't exit non-zero for XXX/FIXME/TODO comments in code [15:31] dobey: cool, thanks. [15:31] alecu: which is why the comment is "test this doesn't fail" :) [15:33] dobey, I could have not explain it better :P [15:34] mandel, I've not run the tests on my P vm, because I still have to install more stuff, but I've run the tests on O, so I'm approving [15:36] alecu, superb, dobey you do have a P machine, right? [15:36] yes [15:36] dobey, would you run the tests in the P machine then, it works on in my P, but is better that in is ran in a diff env [15:37] when i get to the review :) [15:37] i have a bunch of ubuntu uploads to do first [15:39] dobey, yeah, no preasure, I just wanted you to know we needed a P === yofel_ is now known as yofel [15:41] alecu, the webclient_factory, does it take any *args or **kwargs? [15:51] brb [15:52] dobey, I keep forgetting that module that needed to be installed to run the sso tests so that you do not see the ui, do you mind refreshing my mind and tell me stupid at the same time? :D [15:53] xvfb [15:53] back === gatox_away is now known as gatox [15:54] dobey, thx [16:02] nessita, alecu why is the sso ping url unicode? [16:02] mandel: why or where? [16:04] alecu, please..... when you have a moment, please re-review this old branch: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/network-detect/+merge/77048 (only need your approval) [16:05] nessita, in ubuntu_sso/tests/tests_credentials.py, line 430 for example, the current webclient throws this: TypeError: url must be str, not unicode [16:07] mandel: what was your original question? why or where is the ping url code? [16:09] nessita: why is it unicode instead of str; not where is the code [16:11] nessita, why is the ping url unicode? [16:13] dobey, mandel: ah, misundertood the question [16:13] nessita, np :) [16:13] mandel: is unicode up to the point where we actually send it [16:14] mandel: so we can add to it any variable with any unicode value we need (since we append specific t to store in our servers) [16:16] nessita, hm.. have we ever tried that code with a funny char? [16:17] nessita, I might not understand the docs from here: http://docs.python.org/library/urllib2.html#urllib2.Request [16:17] mandel: if I recall correctly, we have tests for that [16:17] mandel: why? [16:17] mandel: why do you need to do, exactly? [16:17] nessita, I read that is must be a str, not unicode, and my brain is confused [16:18] mandel: yeap, that is correct. If you see the sso code, it will encode the url before using it [16:18] mandel: see utils.oauth_headers, it calls: [16:18] 99 url = url.encode('utf-8') [16:19] mandel: so, what do you need to do? I may help you [16:20] nessita, I was just wondering, I wanted to make sure that I indeed needed to do the .encode('utf-8') and that it was no error [16:20] alecu, ping [16:20] nessita, also, I think that if we do that, it should be in webclient, right? [16:20] mandel: yes [16:21] ok. need to get lunch. bbiab. [16:21] mandel: in the line of code closer to doing the actual web request [16:23] gatox, it seems that the last NeedsFixing in that branch was not fixed [16:23] alecu, no?? [16:23] let me check again [16:23] nessita, does this make sense then: #911844 [16:23] alecu, nessita bug 911844 [16:23] Launchpad bug 911844 in ubuntu-sso-client "Teh common webclient should be able to handle urls that are unicode and not str (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/911844 [16:24] mandel: besides the typo, I would say we should agree on an API where the webclient only accepts unicode urls [16:24] alecu: ^ [16:24] alecu, ahhhh sorry..... didn't see that one [16:24] nessita, fixed the typo hehe [16:24] nessita, mandel: a url is not unicode, its bytes [16:25] hmmm [16:25] alecu: is bytes before sending it thru the wire [16:25] * alecu is not sure of the thing he just said [16:25] * alecu googles. [16:25] alecu: I meant, is bytes right before sending it thru the wire, with some encoding. Before that, I would advice having it as unicode [16:29] nessita, alecu is a matter of agreeing, and how to make it so that it gets as bytes in the diff implementations, doing the decode in every diff implementation is a PITA [16:33]