[00:06] hello everyone, I am looking into this ubntuone , but I fail to see the point i guess [00:07] multiple computer access same files? [00:10] it's useful if don't have a computer wake up 7/24, it's an online (possibly public) private space [00:12] folks, my question stay open : is there anyway that u1 could be as accessible as, say, rsync.net ? [01:25] __lucio__, ping [04:23] hi === teknico changed the topic of #ubuntuone to: Have a question? Ask teknico | https://one.ubuntu.com | https://launchpad.net/ubuntuone | Please honk if you want a music store [12:41] hello every body [12:42] can some one tell me how to change my ubuntu one account's password ? === teknico is now known as teknico_away [13:03] how do i change my ubuntu one account's password ? [13:03] rodrigo_: hi. Regarding #DEBUG_MESSAGES, why do you want to enable debuging at runtine if at build time you have configured using --disable-debug-messages ? [13:04] adiroiban, for packages, so that we can get debug info from users installing packages [13:04] adiroiban, not sure if we want to build the package with --enable-debug [13:04] then just use ./configure [13:04] by default [13:04] debuggin is enabled [13:04] leohartx, hey; your Ubuntu One password is currently handled by launchpad.net. So, if you go to launchpad.net and log in, you'll be able to change your password there [13:05] so to remove #DEBUG_MESSAGES you will need to explicitly configure using --disable-debug [13:05] i'm giving it a try, thank aquarius [13:05] leohartx, click your name in the top-right corner after logging in, then "Change your password" on the right-hand-side. [13:06] adiroiban, yes, that's why I think we might just use the environment variable? [13:07] aquarius : i did it, thank you [13:07] what are the problems, drawbacks for the current implementation? [13:07] leohartx, no problem. [13:08] bye guys, gtg [13:08] you will just use ./configure [13:08] and debug messaes are enabled using the environmet variable [13:09] adiroiban, so, have --enable-debug in the package? [13:09] adiroiban, IU'm ok with that, not sure what the policy for that is for packages [13:09] rodrigo_: can you please rephrase „so, have --enable-debug in the package?” [13:10] adiroiban, I mean, with the current patch, we have to --enable-debug-messages in the package, and when we want to get debug info, tell the user to set the environment variable [13:10] if we build the package with --disable, no way to get debug messages at all, right? [13:11] yes [13:11] but by default ./configure [13:11] will enable debug messages [13:11] oh, right [13:11] so a package using CDBS will have them enabled [13:11] at compile time [13:12] so you want debuggin to be always enabled and instead of declare -x COUCDB_ENABLE_DEBUG to have declare -x COUCHDB_DISABLE_DEBUG? [13:13] I want an easy way to enable debugging, that's why IU ask you to add the environment variable [13:14] but if DEBUG_MESSAGES is enabled by default in configure, we're set :) [13:20] * rodrigo_ -> lunch [13:54] aquarius, would just calling that one method from a python interpreter work to pair with u1 ? [13:54] aquarius, also, maybe it might make sense to separate ui from functionality bits for pairing applet, so I could more easily provide an equivalent [13:54] homeasvs, I think so, but I'm not totally familiar with that side of the code [13:56] homeasvs, ah, it is separated, mostly. That function *should* be all you need to call; the call to that function is in ubuntuone-applet so that it automatically sets up a desktopcouch-to-Ubuntu-One pairing for you, but it doesn't have to be in the applet [13:56] it's just python [13:56] cardinalfang built most of the pairing stuff [13:56] aquarius, yeah, but the code to check if it's already paired is a little involved there [13:57] also, didn't see consumer_key (an arg to the function) actually being used [13:57] homeasvs, yes, yes it is. I agree with you :) [13:57] but I should look into it some more [13:57] no reply from the python-keyring guy yet, right ? [13:57] we should contact his gsoc mentor too :) [13:57] we should provide an is_this_replication_service_already_paired function, and we don't. [13:57] no reply to me, no. You haven't heard anything? [13:57] nope [13:58] also, dobey brought up an interesting point, which is: if we go python-keyring for desktopcouch, so it neatly stores into the most appropriate keyring, whatever that is...what do the C people do? === teknico_away is now known as teknico [14:21] homeasvs, I've just mailed the desktopcouch list about this problem. Your comments are eagerly sought :) [14:22] aquarius, go with the gnome one ? [14:22] not ideal on Windows :) [14:23] where do I subscribe to that list ? [14:23] http://groups.google.com/group/desktop-couchdb/ [14:23] aquarius, I'd just say 'What you C is what you get' [14:23] ie, 'should have used python' [14:23] thought you were already on the list! sorry :) [14:23] heh. "you gotta use python; suffer, loser" tends not to work that well as an argument on Android :) [14:26] sure, and if I had infinite power of influence I"d make you guys go write a Universal C-based Key Library [14:26] but I don't see that happening soon [14:45] you see the problem [14:53] adiroiban, ok, your 2 branches are now landed [14:53] adiroiban, please set the branch (not the merge proposal) status to 'Merged', so that they don't show up in the list of branches for couchdb-glib [14:55] hey homeasvs, congrats on the dc on n900! [14:55] homeasvs, will test it later [14:55] rodrigo_, ok, let me know, still waiting for someone to confirm that stuff works [14:56] ok [14:56] homeasvs, btw, did you have problems updating to the latest maemo5 update on the app manager? [14:56] homeasvs, it seems it forces you to reflash if you have installed packages by hand with apt-get [14:57] rodrigo_, zaheer told me to remove the qt packages, after that it worked without reflashing [14:57] rodrigo_: I think I need to delete the branch [14:57] homeasvs, I had installed gcc and other -dev packages by hand, and I had to reflash, did you? [14:57] homeasvs, ah [14:57] rodrigo_, I didn't install any dev packages [14:57] adiroiban, no, no need to delete, just set it to merged [14:57] rodrigo_, if I build stuff on the phone I do it in a chroot - I don't want to mess up my phone :) [14:57] homeasvs, ah, should have done that myself then :D [14:57] rodrigo_, well, now you can since you reinstalled it :) [14:58] yeah [14:58] rodrigo_: thanks :) [15:11] standup? [15:12] oh, you mean MEETING BEGINS ;-) [15:13] me [15:13] me [15:16] me [15:16] me [15:18] that's probably all of us today [15:19] rodrigo_: go [15:19] • DONE: Released libubuntuone 0.1.0. Couchdb-glib branch reviews and merging. Submitted fix for missing python-lxml dependency on the server. [15:19] • TODO: Conflict resolver tool in pair tool. Look at becoming a MOTU (https://wiki.ubuntu.com/UbuntuDevelopers). Make sandy's snowy test suite work with our server (http://git.gnome.org/cgit/snowy/tree/api/tests.py). Discuss with jdo and aquarius about oauth token per app, not per machine? Move JS code to its own file and add jslint tests to check: [15:19] • BLOCKED: no [15:19] next: teknico [15:19] DONE: set up a dev environ in a virtual machine, to work around bzr problems (#506974); more work implementing the mobile sync REST API for client app (#504689); some face duty [15:19] TODO: more face duty; land a branch to switch funambol_cared to a new server branch; more work implementing the mobile sync REST API for client app (#504689); fix a user's problems accessing the contacts web IU (#506584) [15:19] BLOCK: none [15:19] next: Chipaca [15:19] DONE: basked in the sun. TODO: read a couple of weeks worth of email, catch up with everything and everyone. BLOCKED: no way am I going to finish today [15:19] aquarius: [15:19] ⚀ DONE: redo music store URLs; adapt music store plugin prototype to use new URLs [15:19] ⚁ TODO: fix django-openid-auth so that I can actually use the DONE work above; have music library page send message to downloader daemon and row to database; make tomboy first-sync experience nicer [15:19] ⚂ BLOCKED: [15:21] MEETING ENDS, not with a bang [15:22] sorry, I could make up something cool to be blocked by other than my own stupidity if it makes the meeting end more impressively. [15:22] BLOCKED: laptop was eaten by a swarm of space wasps [15:23] :) [16:26] Hi. I'd like to develop an application against ubuntu one. How do I get oauth consumer key and secret for the app? === teknico changed the topic of #ubuntuone to: Have a question? Ask here | https://one.ubuntu.com | https://launchpad.net/ubuntuone | Please honk if you want a music store [16:33] tomm-i, hi [16:33] tomm-i, when you say "against Ubuntu One", do you mean to do with file synchronization, or to do with CouchDB and desktopcouch? [16:35] aquarius: to update notes. It would be another front end for it, somewhat similar to tomboy with some special requirements. [16:35] tomm-i, ah. then you don't need to get the existing key. Instead, you request a new one, authorising your application to edit the user's notes [16:36] it's standard 3-legged oauth [16:37] tomm-i, alternatively, you can use the existing oauthdesktop D-Bus service to get the existing key [16:38] tomm-i, make a d-bus call to com.ubuntuone.Authentication.login() and then wait for a NewCredentials signal [16:38] I need to write this up somewhere, don't I? [16:39] aquarius: yeah, you do :) And I need to learn more about oauth. Unfortunately I don't have dbus available. [16:41] tomm-i, ah, so your app isn't on a standard Ubuntu desktop? In that case, you need to do three-legged OAuth. The OAuth URLs used by Ubuntu One are in /etc/xdg/ubuntuone/oauth_urls on a standard Ubuntu desktop [16:41] tomm-i, http://paste.ubuntu.com/358573/ [16:42] * aquarius adds "how to get keys" to the plan for the Ubuntu One talk during Ubuntu Developer Week. Good reminder :) [16:42] aquarius: hmm, consumer secret is there? You said that I need to request a new one? From where? Oh... nice plan ;) [16:43] the consumer key and secret are always the same: ubuntuone/hammertime [16:43] kudos for someone for nice password [16:43] (in the original oauth design, consumer key and consumer secret were meant to be kept secret by the application that used them, so they'd be specific to your app. That doesn't work all that well when apps are open source :)) [16:43] but I think I got the hang of it, now thanks :) [16:44] tomm-i, I have to go now, but do please ping me (or others here) for helpif you get stuck! [16:44] Thanks! :) [17:58] hi everyone [18:00] Whenever I put a file into my ubuntu one folder it appends u1conflict to the filename (which i guess means there's a conflict). But I don't understand why there is a conflict, there's no files whatsoever in my ubuntu one on the web [18:15] ojii, hi. Does the filename contain colon ":" character? [18:16] rtgz: nope [18:16] ojii, is the file name secret or you can disclose it here? [18:16] erm i tried 'myfile.txt' (which is a text file containing 'hello world') and a picture called me_2.jpg [18:16] neither worked [18:27] rtgz: see http://files.ojii.ch/one.png [18:28] ojii, ok, then could you please pastebin the contents of ~/.cache/ubuntuone/log/syncdaemon.log ? [18:28] sure thing [18:30] there's no plaintext passwords or whatever in there I guess? [18:30] rtgz: http://paste.ubuntu.com/358621/ [18:30] no, only filenames are the most sensitive part [18:31] i don't see how filenames can be very sensitive... so no problem there [18:32] ojii, ok, looks a bit strange. Do you have some 15 minutes to debug this? [18:33] sure [18:33] ojii, if yes, then could you please shut down the syncdaemon (the daemon that is responsible for syncing the files) with u1sdtool -q [18:33] done [18:34] rtgz: done [18:34] ojii, then check that ps auxw | grep ubuntuone-syncdaemon does not return anything [18:34] nah still there, should i kill it? [18:34] if that returns some running process, (except grep) then kill it [18:34] hm, 2 syncdaemons?... [18:34] ops [18:34] was the grep [18:35] ojii, it is already dead so it is ok :) [18:35] always forget about that when i ps | grep something [18:35] ojii, yep, I should have said ps auxw | grep [u]buntuone-syncdaemon [18:36] o.O that's actually good to know [18:36] ojii, ok, now start the syncdaemon with debugging enabled: /usr/lib/ubuntuone-client/ubuntuone-syncdaemon --debug > ~/syncdaemon-debug.log [18:36] ojii, reconnect to the cloud via the applet and try putting some file to Ubuntu One directory [18:39] conflicted again, btw I get 'updating finished: ubuntu one finished updating 0 files' [18:39] and isn't the daemon supposed to go to background when i launch it from the shell? Or does it stay with debug on? [18:41] ojii, ok, now pastebin the contents of the generated syncdaemon-debug.log [18:41] ojii, 0 files.. yes, need to poke devs about it, It also says so when there are only files removed... [18:41] rtgz: http://paste.ubuntu.com/358628/ [18:42] rtgz: why does using the bracket on grep keep it from showing up in the results? [18:42] or rather, how [18:43] (not related at all with the issue, sorry!) [18:43] jamalta, "[u]" becomes a part of the command line, but grep will use "u" letter only, since [u] is simply u when regular expression is expanded [18:44] rtgz: hm.. interesting, i get the regular expression part.. not sure what it means for it to become part of the command line though [18:44] cool trick nonetheless [18:45] i agree :D [18:46] prevents newbies like me to think stuff is still running [18:46] jamalta, so grep sees "grep [u]buntuone" literally while it is searching for "ubuntuone" string [18:47] ojii, back to the log, reading it right now :) [18:47] rtgz: OH! RIGHT That makes sense [18:47] i feel like a fool for not seeing that at first :) [18:47] thanks [18:49] oh just btw, ubuntuone is written in python? (the log looks a lot like it uses python) [18:52] ojii: yeap [18:53] nice, this makes me like it again despite not working for me ;) [18:53] ojii: heh, well it's somewhat working [18:53] and if you're experiencing a bug, then it is likely you aren't the only one [18:53] so it's great that you're trying to discuss it here :) [18:53] as i said: not working *for me* [18:56] ojii: i understand, and i'm sure the devs will be glad to help you as soon as they can put some attention towards your issue === dpm is now known as dpm-afk [18:59] meh I'm sure rtgz will magically solve that problem [19:00] ojii, when did it start? [19:00] ojii, I mean when has it started malfunctioning? [19:00] what start? [19:00] ah [19:01] rtgz: well last time I checked (probably a month or so ago) it wouldn't even connect (crash on me trying to connect). Now I tried again yesterday and I get this problem [19:02] hi ojii, rtgz [19:03] hi verterok [19:03] ojii, hmm.. could you please pastebin the contents of /etc/xdg/ubuntuone/syncdaemon.conf [19:03] shot in the dark.. [19:04] rtgz: http://paste.ubuntu.com/358639/ [19:04] verterok, summary: ojii's syncdaemon loses the file and then renames it to conflict [19:04] rtgz: thanks :) [19:05] ojii: do you have multiple ubuntu one/launchpad accounts? [19:06] verterok: I don't think so.... [19:06] verterok, http://paste.ubuntu.com/358628 [19:06] but launchpad is confusing [19:06] verterok, syncdaemon w/ debug [19:07] ojii, rtgz: looks like the problem is that the MakeFile is failing, with a DOES_NOT_EXIST error, this means that the parent does not exist [19:07] in this case the parent is the root, so I'm quite sure it exists [19:07] verterok, you mean Ubuntu One ? [19:07] rtgz: yes [19:07] verterok, ubuntuone does exist, I am using it :) [19:08] my best guess is that the id of root stored in the client metadata isn't the root of the authenticated user [19:08] verterok, we had a very similar case to this after changing the user's auth stuff, right? [19:08] ojii: do you have any files in ~/Ubuntu One? [19:09] facundobatista: don't remember [19:09] verterok: yes, i put three files there for testing, i also have the Shared with me thing [19:09] verterok, so the file gets to the folder, then the parent is checked, the parent does not exist and everything breaks [19:09] rtgz: sort of :) [19:09] I'm pretty sure this is an early sign of 2012 :D [19:10] facundobatista: what auth stuff changed? [19:10] ojii, the sign of 2012 is that we will run out of IPv4 addresses. Syncdaemon is not related to apocalypse [19:11] verterok, so, if we remove metadata info, reinitialize syncdaemon then everything should be back to normal? [19:11] rtgz: good point, and I'm more scared of 2038 since I still have a server running on 32bit [19:11] verterok, user gets a root in the server, and after that something changes at user/auth/token level, and then it tries to put files on *other* roots [19:11] ojii: ok, as you don't have any files, let's do a backup of all syncdaemon metadata and files [19:11] yay [19:11] sounds like fun [19:11] facundobatista: somethin changes in the server side? [19:12] ojii: tar -cjf ubuntuone_metadata.tar.bz2 ~/.share/local/ubuntuone [19:12] oops [19:12] ojii: tar -cjf ubuntuone_metadata.tar.bz2 ~/.local/share/ubuntuone [19:12] lol [19:13] makes more sense [19:13] ojii: and: tar -cjf ubuntuone_cache.tar.bz2 ~/.cache/ubuntuone [19:13] done [19:14] now rm -Rf those I guess [19:14] ojii: and last: tar -cjf UbuntuOne.tar.bz2 ~/Ubuntu\ One [19:14] ojii: yes, rm -Rf ~/.local/share/ubuntuone ~/.cache/ubuntuone ~/Ubuntu\ One [19:14] done [19:14] ojii: quit syncdaemon first ;) [19:15] ojii: once you removed all that stuff, start syncdaemon [19:15] how do i do that best? [19:15] just ubuntuone-syncdaemon? [19:16] ojii: in order to get the debug logs: /usr/lib/ubuntuone-client/ubuntuone-syncdaemon --debug > ~/syncdaemon-debug.log [19:16] ah kay [19:16] can i force that to background? [19:16] I don't like to have open shells everywhere [19:16] ojii: once we check it's working, you can quit the applet and start it again,m and that will start syncdaemon [19:17] okay started daemon, reconnect and try to add a file? [19:18] ojii: yes, please [19:18] verterok: will you marry me? [19:18] hehehe [19:19] verterok: (= it works!) [19:19] but i guess you figured that out [19:19] ojii: cool! [19:20] indeed [19:20] thanks verterok and rtgz! [19:20] ojii: looks like syncdaemon was previously connected with a different launchpad/sso account, ATM syncdaemon doesn't support changin the LP/sso username [19:21] now i gotta find something useful to use this for :D [19:22] verterok, i changed LP account name several times, syncdaemon continued to work properly... [19:22] * rtgz YES! now all my text messages are fetched via bt/serial connection by smsd and dispatched to local sendmail... Niiice... [19:24] rtgz: ups, yes. you can change the account name, but you will break it if you remove the oauth tokens and create new ones with a different account ;) [19:24] can I just share a folder with 'everyone'? [19:31] ojii, no, you need to find someone to try sharing. Public access is not yet supported. But public urls for some files are planned, I guess [19:31] ah cool [19:32] wanted to use this with shutter so i don't have to upload the files all the time... [19:56] ojii: i'm waiting as well.. i love the ability to do this with Dropbox but I honestly would much rather do it with Ubuntu One [19:57] guys, just for the record. What's the official status of Tomboy notes web? [20:04] here, just for the record, what is the status of http://one.ubuntu.com/notes/ ? [20:05] * rtgz oO( "Ask here". Ok, asking "here" ) [20:41] hi rtgz: I'm not sure I understand your question about official status of http://one.ubuntu.com/notes/ [20:42] statik, short story: does not work [20:42] * statik logs in to the notes UI to check [20:42] statik, long story: it stopped working one day before christmas. New notes are created, but only one revision is saved. Subsequent edits get lost [20:43] oh man that is terrible [20:43] statik, no edits of the existing notes are possible [20:43] statik, Bug 501020 [20:43] cardinalfang has a branch open that tries to address a problem with how notes are versioned inside couchdb [20:43] Launchpad bug 501020 in ubuntuone-servers "Tomboy note sync and web-ui for notes are broken" [Undecided,Confirmed] https://launchpad.net/bugs/501020 [20:44] statik, yep, but this is like... "we support note sync!*" and at the same time "* Only one sync per note created on the web interface" :) [20:44] rtgz, this totally sucks. i see it now too "there was a problem with saving your note" [20:44] statik, I even awarded it a "christmas-bug" tag... [20:46] statik, http://twitter.com/UbuntuOne/status/6928891051 - This was the only notes-related info after that upgrade. Tomboy sync is broken, this is known, somehow, maybe, ... but the web UI unable to perform the edits might affect UbuntuOne credibility as this is one of the core routines. [20:47] rtgz: for sure. thanks for bringing it up. I wonder if it's related to this bug: https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/499595 [20:47] Error: This bug is private [20:47] oh, i need to make that public [20:49] https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/499595 [20:49] Ubuntu bug 499595 in ubuntuone-servers "Notes API causes ResourceConflict" [High,In progress] [20:49] statik, yup, looks like this is what happens behind the scenes [20:51] looks like cardinalfang has a branch to fix that, and it's been through code review but not yet landed [20:52] i need to check and see whether edge is still rolling out every 4 hours, might be able to test this very soon [20:53] statik, ok, thanks. 'cause it looks like nobody from the dev team have tested the /notes/ ui recently. [20:57] rtgz: this is a problem that we're trying to figure out a good strategy for. i'd like to have simple regression/acceptances tests running so it's immediately obvious if something stops working. simple is so easy to say and not so easy to accomplish though. We've got some EC2 machines that we could configure hudson on, and I've been talking to folks about setting up windmill - but also wondering whether we should be using something like cuc [20:57] umber or pyfit [20:57] so hudson would take care of running the regression/acceptance tests on edge every day [20:58] and we'd need to write a windmill script or something to login, edit notes, and confirm that things work ok [20:58] naturally there will always be some tests that need to be done manually - what i'd really like is a way to record and publish the results of any manual testing, so that we can see what hasn't been checked in a while, and make it easy to give people credit when they help with testing things [20:59] statik, or selenium until that's done? [20:59] oh yeah, selenium would be great too [21:00] statik, the problem is that there are bug reports about these issues, it's just not that clear how to "push" them so that they become more visible [21:04] rtgz: that is a good point, and something that i've noticed too. there are some important bugs that sometimes get overlooked for a while [21:06] rtgz: i don't have a complete solution yet, but I have reorganized the team a bit so that we have more people working on QA related activities, and I'm hiring a full time person to do bug triage and make sure that the most important bugs get prioritized or escalated correctly. What we had been doing before is having each developer take a turn one or two days a month doing bug triage. thats great for spreading knowledge but it hasn't been s [21:06] o efficient for staying on top of the bug triage inflow