[09:51] morning [10:30] still cant get a invite? [10:30] wtf i signed up day one [10:31] midkniht: many people did; we're releasing batches of new invites every day and monitoring system performance very carefully. [10:44] * midkniht nods [10:44] just thought about it again and decided to ask [10:44] ive just sat silently and waited as i will return to doing [10:49] I'm not an ubuntuone dev, but perhaps one is around and can check that you are in fact in the waitlist (if the web UI doesn't confirm that you aare for you) [11:06] never checked it since i signed up the first time [11:06] where is the page? [11:07] https://ubuntuone.com/account/ I suspect [11:07] ubuntuone.com [11:20] hell i cant login to launchpad at all now iguess [11:21] everything goes to this stupid openid login [11:21] not my real account [11:22] so i guess i do need someone to look it up [11:22] you can't login to lp? [11:25] its weird [11:25] i went directly to lp, i login its fine, then i go to ubuntuone and im a openiduser [11:26] oh wait [11:26] i found it [11:26] You are currently on the waiting list for Ubuntu One. As soon as we can, we'll be sending you an invitation code via email. [11:26] dunno why its showing openiduser at the top [14:07] midkniht: you logged in to ubuntuone via openid (it's the only method we support, and we currently only support openid from launchpad) [14:08] Hello Ubuntu One beta testers. I'm here to help anyone today that has questions or needs support with the service. === rmcbride changed the topic of #ubuntuone to: Today's Face of Ubuntuone: mattgriffin | File Sharing for Ubuntu | https://ubuntuone.com | https://launchpad.net/ubuntuone | Currently in limited beta, new invitations are sent regularly | Current Beta Client Revno is 41, Protocol Revno is 38 === rmcbride changed the topic of #ubuntuone to: Today's Face of Ubuntu One: mattgriffin | File Sharing for Ubuntu | https://ubuntuone.com | https://launchpad.net/ubuntuone | Currently in limited beta, new invitations are sent regularly | Current Beta Client Revno is 41, Protocol Revno is 38 [14:18] I have a question about the service: is there a limit on data transfers or it's on space only? [14:22] Brownout: there will be a limit on transfer. we haven't quite figured out what exactly that should be. [14:23] we don't want to prevent people from getting the most from the file sharing service... but on the other hand, as with any service, abuse could become a problem. [14:24] Brownout: is there any specific application of the service that a data transfer cap would worry you? [14:24] I was thinking mainly about upload. File serving abuse scenario didn't occur to me. [14:26] Brownout: today, file sharing would be difficult to abuse since we don't enable public sharing yet. it's something that we're looking into for the future though. [14:27] you could share access credentials, but I suppose it would be against TOS [14:27] Brownout: ah ... upload. i don't think you should be concerned. we will have a reasonable transfer cap. [14:33] midkniht: just read through the irc log. are you still experiencing openid issues with Launchpad? [14:33] dobey: an idea for naming your postinst restarter project, something based on the ultimate machine that claude shannon built: http://en.wikipedia.org/wiki/Claude_Shannon#Hobbies_and_inventions [14:34] mattgriffin: i have reinstalled my pc to karmic, and then installed the ubuntu one bits from the karmic repo. now when i start ubuntuone i get the icon, but clicking "connect" from the right click menu doesn't do anything [14:35] statik, dobey: Claude Shannon's "A Mathematical Theory of Communication" is a great read! http://en.wikipedia.org/wiki/A_Mathematical_Theory_of_Communication [14:39] popey: hello. thanks for being a beta tester. i haven't tested with karmic yet. i think this might have come up last week. i'll check Launchpad Answers and the IRC logs to see what I can find. [14:41] statik: heh. [14:41] mattgriffin: https://bugs.launchpad.net/ubuntuone-client/+bug/385237 looks like it, but it says that its fixed [14:41] Launchpad bug 385237 in ubuntuone-client "Possible issues with Karmic (dup-of: 383781)" [Undecided,New] [14:41] Launchpad bug 383781 in ubuntuone-client "[karmic] syncdaemon stopped working" [Critical,Fix committed] [14:44] in fact bug 385313 is also affecting me, but that also says its fixed [14:44] Launchpad bug 385313 in ubuntuone-client "[karmic] syncdaemon can't find python-configglue" [Critical,Fix released] https://launchpad.net/bugs/385313 [14:46] popey: you get that error when running ubuntuone-syncdaemon? [14:46] no not having problems i guess afterall mattgriffin [14:46] yes dobey [14:46] just looked weird to me [14:46] im just waitlisted [14:46] popey: what version of python-configglue doe you have installed? [14:46] 0.1-0~9.10-4+r9 [14:47] its in site-packages [14:47] Hmm. that should be the right version [14:47] huh [14:47] * rmcbride doublechecks [14:47] midkniht: sorry about that. we recently increased our invitation granting pace so it should arrive soon. [14:47] why are the configglue packages versioned oddly [14:48] dobey: that's the versioning that we have been using. I'm working on normalizing things with the REVU clean version, but we needed to get the PPA version fixed quickly [14:54] rmcbride: have you uploaded python-configglue to revu yet? [14:55] statik: I did a fixed upload this morning, it hasn't shown up in my summary yet [14:55] rmcbride: excellent! [15:00] rmcbride: it seems there was something wrong with your upload if it hasn't shown up yet [15:00] standup now, right? [15:00] rmcbride: someone on #ubuntu-motu will be able to fish it out of the rejects directory for you [15:00] mattgriffin, its cool thanks, keep up the good work [15:00] james_w: hmm. thanks [15:00] MEETING BEGINS [15:00] Hi, everyone who is here for the ubuntu one developer standup, please say 'me'. The format for this meeting is to paste 3 lines: TODO, DONE, and BLOCKED. Then say the person that should go next - we can try going in 'me' order. [15:00] me [15:00] me [15:00] me [15:00] me [15:00] me [15:01] vds, teknico, and aquarius are sprinting today i think [15:01] dobey should be here though [15:01] DONE: jingle/xmpp research, database deadlock stress, weekend security bug work. [15:01] TODO: Work on security bug with pfibiger. Work on OOPs with pfibiger. Chase lots of packaging work. [15:01] BLOCKED: None [15:01] next is jblount [15:01] DONE: Very little code, lots of email and IRC [15:01] TODO: Lots of code, very little email IRC (specifically I need to finish this file-usage-info branch and have a chat with urbanape about tackling the /files/ new ui) [15:01] BLOCKED: I'm having a hard time staying focused, so I may go darkish on IRC during the EST afternoon to try to focus a bit more, text or call my mobile if you need me after lunchish [15:01] * jblount passes to CardinalFang [15:01] me [15:01] done: subprocess leakage plugged. Graphite receives messages. [15:01] todo: replace mailer system: no MTA for devel, use transactional library. [15:01] blocking: none (except maybe some middlware for transactions?) [15:01] rodrigo_ -- tag! [15:02] done: more evo-couchdb/couchdb-glib work, some fixes in tomboy notes web interface [15:02] todo: integrate evo-couchdb in evo account setup, more tomboy notes work [15:02] blocked: none [15:02] urbanape: your turn :) [15:03] done: more writing on Firefox plugin. Want to get with aquarius a bit and get his thoughts [15:03] todo: work with jblount on /files/ ui [15:03] blocked: none [15:03] dobey: all yours [15:04] DONE: yet more nautilus extension in C (finishing today), micro-sprint with urbanape [15:04] TODO: Finish nautilus extension C port, Fix icontool package issues, package python-oauth, split ubuntuone-client packaging [15:04] BLCK: Need awesome name for dbus restarter magic. Awaiting approval on central-services. [15:04] nobody else? [15:05] i think that was everyone [15:05] thanks all! [15:05] MEETING ENDS [15:05] hey dobey. python-oauth is on my list as well, you need it doing urgently? [15:06] james_w: you were planning to package it? i was going to do it today [15:06] (assuming it's the same one) [15:06] james_w: the one from google [15:06] yeah [15:06] I need that for the new launchpadlib [15:07] want me to take a swing at it today? [15:07] james_w is a HERO [15:07] james_w: i just recently got clarification on the licensing for trunk, as there was some confusion about it, as there are several oauth impls hosted in the same svn on google code [15:07] ah, nice work [15:07] james_w: but yeah, if you're going to do it, that would be great :) [15:08] I suspect the actual packaging will be very quick for me, but if there are license issues then it could be a while afore we can get it in to Ubuntu [15:09] dobey: feel free to ping me later to test the C nautilus plugin [15:09] jcastro: cool [15:10] james_w: actually, i already started packaging it a while ago, but ran into the license clarity issue, and only recently did Leah respond to my inquiry about it :) [15:10] well if you've got something perhaps we should start from that [15:10] james_w: the license issue is resolved. trunk should have the appropriate (C)/license header at the top of oauth.py now, with the license as MIT [15:10] urbanape: can you make (if you haven't already) a lightweight spec with some basic timelines for that firefox/couchdb plugin ? [15:11] will do [15:11] it's a trivial package, since it's python with distutils [15:11] I'll keep adding to the page. I know it needs to get finalized this week to get into Karmic, yeah? [15:11] urbanape: are you planning on using my couchdb-glib lib? [15:11] though I would like to test my changes to dh_make to make it python aware :-) [15:12] james_w: well go for it. the packaging is super trivial anyway [15:12] rodrigo_: I figured it would all be done over HTTP, but I'd be happy to consider alternate approaches. [15:12] it's one py file installed with stuptools/distutils :) [15:12] (via javascript) [15:13] urbanape: well, I'm adding code to couchdb-glib to not only deal with the HTTP stuff, but with the internal format we use [15:13] dobey: did you ask about the license of the examples/ directory? [15:13] urbanape: for contacts it's quite complicated, so we don't want people to be dealing with the format in N apps [15:13] urbanape: I guess for bookmarks we should do the same [15:14] i probably wouldn't use couchdb-glib for firefox bookmarks... and i don't think we can use it from a firefox extension directly anyway [15:14] sicne we're considering a schema for bookmarks, yeah, we might want to use that. Can I pick your brain in a bit? [15:14] urbanape: yes, of course, the code is in gnome's git if you want to have a look [15:14] mostly, I just figured we'd be talking to localhost:xxxx and dumping/loading resources. [15:14] but perhaps writing a js lib that people can use with the appropriate bookmarks schema could be good [15:15] and perhaps we'd need js for thunderbird too, unless we patch the code [15:15] 8GB of RAM makes for a very happy lappy [15:16] * jblount promises not to buy any laptops for at least 6 months [15:16] james_w: i didn't. i presume they are the same though, and there is a LICENSE.txt now too [15:16] jblount: me, too. Done. [15:17] dobey: I assume that too, but I wonder if the lack of explicit declaration in the files will bite us [15:17] jblount: what time would you like to get together and talk /files/? [15:17] dobey: why you can't use couchdb-glib from an extension? [15:18] urbanape: It might be a good idea to convince pfibiger to do a skype with us, maybe we can shoot for post lunch, I've still get 1 1/2 branches to sort and I think he and statik are hacking on some security bug. [15:19] james_w: i just sent another mail to ask about them. [15:19] jblount, speaking of hardware, I CC'd you on a pal's question about Thinkpad hardware. [15:19] rodrigo_: because extensions are written in javascript and xml, not C? :) [15:19] dobey: thanks [15:19] dobey: hmm, really? [15:19] I thought you could write in C++ also [15:20] rodrigo_: yes. if they were C/C++, they couldn't be cross-platform. [15:20] CardinalFang: I saw that come in, I'll respond in a bit. [15:20] dobey: are we sticking with pyinotify in the client or will we be switching to gio? I'm reviewing the MIRs right now and there is one for pyinotify [15:20] rodrigo_: plug-ins are usually in C++, and you can interact with them via javascript, but that's a mess we would rather avoid i think [15:20] No hurry. /me curses spammers and email lossiness. [15:21] statik: i think we should switch to gio. i don't know how soon that is doable, and what all the details are though. [15:21] jblount: sounds like a good idea. [15:23] dobey: yeah, i'm not sure on the timelines either [15:24] dobey: remind me how to run ubuntuone-client from a branch, with a dev version of ubuntuone-storage-protocol too - do I just change the protocol symlink to point to the branch that has the code I want? [15:25] statik: you can ./setup.py build --protocol-source=/path/to/protocol/trunk [15:25] ah, thanks [15:25] statik: and then PYTHONPATH=. ./bin/ubuntuone-syncdaemon [15:25] (and likewise for applet) [15:26] running nautilus extension from a branch is a bit more tricky [15:26] cool. i'm just after u1sync and then the syncdaemon, so this should be fine. thanks! [15:28] Ugh, this Unicodification of Python is a headache. I'm not sure what is already encoded and what needs it. === pfibiger` is now known as pfibiger [15:51] hrmm, i think i might need to get a smaller bluetooth mouse for my laptop [16:00] Hmm. How do I get a percentage using two floats? [16:01] jblount: p = x/y [16:02] jblount: but it depends on what normal values for those floats are [16:04] dobey: I've got used bytes / storage bytes, it returns 0.005711137782782316207885742188 but I need it to be .5% or similar [16:04] I guess I just need to multiply it by 100 and truncate? [16:05] jblount: well multiply by 100 will give you something printable as [0-100]% [16:05] jblount: and you probably only want to show 1 or 2 decimal places after that [16:08] dobey: http://stackoverflow.com/questions/783897/truncating-floats-in-python/783900#783900 mentions around "round" function, seems perfect. [16:08] That way it will round up to be more accurate. [16:09] jblount: i don't know. in C you can just say "%0.2f" as the format string, but i don't think python works that way :) [16:09] dobey: :) [16:12] dobey: python-oauth is in the NEW queue [16:13] james_w: awesome! [16:17] jblount: take anything you see on stackoverflow with a very large grain of salt [16:17] jamesh: Acknowledged :) [16:18] jblount, do you want a float or a string as result? [16:19] CardinalFang: String is fine [16:20] Then, yes dobey has it best. "Blah blah %.1f%% blah" % (f*100,) [16:22] CardinalFang: Sure, but this give me a very long number. I just want xx.x% [16:23] oh, python is similar i guess [16:23] >>> "%1.2f%%" % (1.128348129349125,) [16:23] '1.13%' [16:23] * jblount rewrites [16:26] CardinalFang: Ah, works fine now. I must have done it badly / differently the first time. [16:26] jblount, mind the ".". It matters. [16:26] has anyone noticed if they computer boots slower with ubuntu one? [16:27] jdobrien: Haven't noticed, but I don't reboot _much_. [16:27] * CardinalFang eyes his uptime. [16:29] jdobrien: I think you recently mentioned a middleware for some kind of transaction. Whatchewtalkin'bout? [16:30] > [16:30] ? [16:30] CardinalFang: if I mentioned it here, it would have been accidental === dpm_ is now known as dpm [16:30] jdobrien: In a bug, iirc. [16:31] CardinalFang: oh. we're having some database transaction isolation issues. [16:32] CardinalFang: we have middleware we're using on our website to handle transactions for web requests. [16:33] CardinalFang: and it appears (although we're not sure) this is causing some problems for the file upload server [16:41] jdobrien: roger. [17:19] have some spare time? take a look at the wishlist and tell us what's missing from the Ubuntu One FS beta. http://tinyurl.com/mwkx8w [17:22] bbiab, lunch [17:26] rmcbride: did those ppa builds ever go through? [17:27] jcastro: yes, they went up on Friday [17:27] jcastro: the fix for python-configglue went up as well, but I'm getting reports of things still not being right on at least one testers' machine [17:27] me either, which PPA should I be using? [17:28] jcastro: either one. I posted them to Beta and to Nightly [17:28] ok [17:29] jcastro: I'm working on getting a karmic machine running, but I think the optical drive on my other test box died. I'm working on getting that running so that I can test stuff here with that issue [17:29] burned 3 CDs before I decided that the problem was not the CDs :/ [17:29] * rmcbride wishes sony hadn't crippled the Virtualization extensions in VAIO bios [17:30] rmcbride: 0.1-0~9.10-4+r9 ? [17:32] jcastro: Yep [17:32] k, didn't work for me. [17:32] still can't find the configglue [17:33] jcastro: yea I'm going to have to get this test box going. It should work, as I fixed debian/control as james_w indicated. [17:33] jcastro: but clearly something is still wrong [17:34] jcastro: want to pastebin the output of "dpkg -L python-configglue"? [17:34] swapping the optical drive out while I eat [17:35] http://paste2.org/p/267367 [17:36] rmcbride: no worries, just holler at me if you want to test. [17:36] ok [17:37] there's something wacky still going on with that package [17:38] james_w: any input gladly accepted on that [17:39] I guess it's good that the upload to REVU didn't take [17:39] heh [17:39] I'm just getting someone to look at that for you now [17:40] Cool I was going to go into "pest" mode after lunch :) [17:48] Hmm. My typing class would have been a lot more useful if they taught 10-key stuff using a laptop keyboard and Fn [18:05] james_w: just seeing your REVU comments on the package now and fixing those issues. [18:06] james_w: thanks very much for the detailed feedback and instructions. [18:06] np [18:19] rmcbride: so this should fix some Ubuntu One issues with Karmic? [18:20] james_w: python-oauth is going to main, right? [18:21] dobey: it don't quite work like that [18:21] mattgriffin: The current REVU activity is needed to get the package into Karmic. It may well fix the issues in teh PPA version of the package as well. I'm not completly clear on what is broken for Karmic and not for Jaunty in that regard, but I believe that following the directions I've been provided on REVU will help [18:21] james_w: well universe, and then main? [18:21] james_w: i mean, the goal is that it will be in main? [18:21] dobey: if it needs to be in main then we can make that happen, but you don't choose at this point [18:21] mattgriffin: I'll need to do a new PPA version of the package that pulls the appropriate changes in (currently the package workflow differs for the PPA) [18:21] yeah [18:22] it will need an MIR at some point [18:22] james_w: yeah, i understand the process. i just want to know where it's planned to end up :) [18:22] james_w: because we'll need it in main for ubuntuone as well, though i presumed it would end up there if launchpadlib is going to use it [18:25] yeah [18:26] james_w: great, thanks! :) [18:30] Sorry for the interruption, just one quick question: I have a folder that I am unable to delete: It says it was: "Modified Sun. June 0 2009, 21:1" [18:31] and according to the log, it stucks at the part when it starts working on the metadata [18:32] so it looks like because of the corrupted metadata [18:32] any ideas how to solve this? [18:40] toros: hrmm. it's a national holiday in .ar, so the guys who would be best to answer that, aren't really around today. but if you could file a bug, i'm sure they would love to help diagnose the problem further when they return :) [18:41] dobey: thanks [18:42] luckily it's not my own account, just someone asked me to help :) [18:44] heh [18:49] rmcbride: is your new configglue upload ready for re-review? [18:49] james_w: nearly. I have one issue left to fix (the directory structure comment) [18:56] james_w: statik is going to do a new upstream release with LICENSE included. I've uploaded the current version of my package if for no other reason than to get the automated checker to tell me if I've forgotten anything [18:56] :-) [18:57] you know about lintian? [18:57] james_w: debuild runs it on the package, yes? [18:57] yeah [18:57] you can run it on the binary packages as well though [18:58] Yea I got a warning on the standards version you wanted, and also it says that my debian/watch file lacks a version. I'm looking into that now [18:58] well, if you build binaries with debuild then it will do it of course [18:58] ah, that's just a format marker for the debian/watch file itself [18:58] "man uscan" [18:58] version=3 at the top IIRC [18:59] Ah. fixing now [19:00] hooray for 300dpi screens [19:00] surprisingly, i can actually read most of the text in this terminal [19:02] dobey: I think you may be capable of reading negative point-sized fonts. [19:02] heh [19:03] i read fonts in infrared [19:08] james_w: is it ok for MANIFEST.in to be in the orig tarball? It surprised me that setuptools included it [19:08] hey statik [19:08] what does that file do? [19:09] I don't think I've seen it before [19:09] james_w: its the file that tells setuptools or distutils what extra files to include [19:09] like LICENSE [19:09] ah [19:09] sounds like it should be then [19:10] MANIFEST.in is appropriate if we have on [19:10] one [19:11] dobey: thanks! i noticed that the tarball for ubuntuone-storage-protocol didn't included MANIFEST.in (at least when I just built it locally) [19:12] statik: oh? hrmm [19:12] boas tardes [19:12] dobey: statik: [[]] [19:13] statik: indeed it doesn't. did you list MANIFEST.in inside MANIFEST.in? :) [19:13] dobey: nope. it looks like setuptools picks up any files that are under version control [19:14] ugh [19:14] i hope not [19:14] well, that doesn't quite match. but i mean it looks like MANIFEST.in is getting grabbed because it's versioned and I didn't exclude it [19:15] thank you guys, I reported the bug [19:15] happy hacking :) [19:15] bye [19:15] statik: it could be that setuptools includes it, but distutils doesn't i guess? [19:16] dobey: yeah, thats what I'm thinking. do you have a preference whether I included it or filter it out? [19:17] do you use setuptools_bzr? [19:17] statik: i don't think it matters. having it there is better i think, because it means you can re-build the tarball from within the tarball [19:17] ok [19:17] does one of you have it installed and the other not? [19:18] james_w: we're not using setuptools_bzr, no. i'd like to find out more about that sometime though [19:18] james_w: i'm certainly not using it, but i haven't messed with configglue, and storage-protocol uses distutils [19:18] ah [19:19] hi BUGabundo btw [19:19] eheh hey back dobey [19:19] hi james_w [19:19] hi BUGabundo [19:21] man this screen is freakin' nice. it's too bad the video is poulsbo instead of sane intel gma === thekorn_ is now known as thekorn [19:27] BUGabundo: http://wayofthemonkey.com/pics/newtoys.jpg :) [19:27] * BUGabundo checks [19:27] rmcbride: you've got a version 0.2 upstream release of python-configglue that includes LICENSE and MANIFEST.in. happy packaging :) http://pypi.python.org/pypi/configglue/0.2dev [19:28] statik: awesome, thanks! [19:28] dobey: is that a pad or a dockstation ? [19:28] that's his other laptop [19:28] BUGabundo: it's my old laptop (the 10" fujitsu) [19:31] cmf [19:31] you really like them small [19:31] rmcbride: if the release is 0.2dev then please version the package as 0.2dev-0ubuntu1, and adjust the watch file to include the "dev" in the group. [19:32] james_w: already done, unless I munged the format [19:32] cool [19:32] BUGabundo: eh, 7" would probably be optimal, but there aren't any 300dpi 7" models. the keyboard is a bit off on the u820 because it's so small. another inch of space,and it could be perfect though [19:34] 9" is great for me! smaller and I can't type /me big finger [19:35] heh. i have a foldable bluetooth keyboard, and a flexiable/foldable usb keyboard on the way (so i can type fast, and the usb for on planes) [19:36] james_w: new package has been uploaded to revu [19:36] eheh [19:51] james_w: btw. does specifying a new Standards-Version (3.8.1) break building on older (hardy) distros? === BUGabundo1 is now known as BUGabundo [20:03] statik: ping [20:03] hi dobey [20:04] statik: hey. so i just realized that i totally forgot i need to do some oauth stuff in the nautilus extension, to create shares, as it interacts with the web directly, rather than going through syncdaemon and the protocol [20:05] statik: this means that we'd need the liboauth C library, which is also not yet packaged [20:05] dobey: i have an idea [20:05] statik: unless we add some dbus API to get an oauth signature [20:06] statik: i prefer the latter of those two, but it's going to take a little more time :) [20:06] dobey: what if we move the share offer stuff behind the dbus API completely? then it could stay in python, and easily move into the protocol itself (where it really should be) without needing the nautilus stuff to change [20:06] statik: so how do you feal about having the extension in C, but with the regression of not being able to create shares, for a couple of days? [20:07] so we can get the extension/packaging done, and then add the share creation back really quickly after that [20:07] dobey: not being able to create shares from the desktop for a couple of days is fine. [20:08] coolio [20:08] i'll disable that bit for the moment, and get on with packaging bits :) [20:30] statik: should we call the magic dbus restarting thing "majikthise"? :P [20:31] heh [20:32] dobey: no, it doesn't [20:34] james_w: great, thanks [20:34] Hello there! Anyone have an idea of how long it takes to receive aproval (or not?) for an ubuntuone account? [20:35] Savago: 0 < time < ∞ :) [20:35] dobey, heheheh... right. :-P [20:37] Savago: we're monitoring server performance, and fulfilling inivitation requests at a pace we think make sense based on load [20:37] Savago: we should hopefully have all the pending invites fulfilled soon though [20:37] Savago: hi Savago. we just went through a process to improve our infrastructure and have recently started granting requests again. i think we're up to about 500 per day so should be soon. [20:37] Savago: sorry for the delay [20:38] Cool, thank you guys for your reply. :-) [20:42] Is there any place where the client protocol of ubuntuone is described? [20:42] https://launchpad.net/ubuntuone-storage-protocol [20:42] I was expecting an article describing it, but I failed to find the link... [20:44] Savago: see if this has what you you're looking for: https://launchpad.net/ubuntuone-client [20:47] the client doesn't have it [20:47] Savago: what sort of description are you looking for? [20:48] dobey, something like this would help: http://code.google.com/apis/calendar/docs/2.0/developers_guide_protocol.html [20:49] but for ubuntuone, of course.. [20:50] oh, like, documentation :) [20:50] we don't have any real documentation yet, but we will be working on some soon i think [20:50] Savago: it's on the todo list. [20:50] yes... I heard that in some point you got to write the protocol spec... and so on... :-D [20:51] And then any interested party can write an implementation by simply reading the spec and writing the code... [20:51] james_w: just pushed a new icontool to review incorporating your changes. i'll file a bug requesting man pages though, but i don't think it's important at the moment :) [20:52] nope [20:52] Savago: the protocol spec is defined using protocol buffer, so getting a C/Java/etc implementation is justa matter of compiling ine ;) [20:52] someone else would have said it if I didn't [20:52] s/ine//s [20:52] james_w: indeed [20:52] mattgriffin, dobey: ATM, the alternative would be to read the current client library code, right? :-) [20:54] verterok, sorry about asking... but what is 'protocol buffer'? Is this the same as this: http://code.google.com/p/protobuf/ [20:54] ? [20:54] 'Google's data interchange format' [20:54] Savago: yes, sorry. google protocol buffers [20:55] Savago: if you want to use it, probably. but the .proto file has the description of the protocol that protobuf-compiler turns into python code that we wrap, and call [20:55] Savago: http://code.google.com/apis/protocolbuffers/ [20:55] Interesting. [20:55] hence the dep on python-protobuf, and build-dep on protobuf-compiler :) [20:56] Savago: as dobey pointed out, you could compile a Java/C++ (a more extensive list: http://code.google.com/p/protobuf/wiki/OtherLanguages) [20:57] Savago: http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntuone-storage-protocol/trunk/annotate/head%3A/canonical/ubuntuone/storage/protocol/dircontent.proto [20:57] and http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntuone-storage-protocol/trunk/annotate/head%3A/canonical/ubuntuone/storage/protocol/protocol.proto [20:57] At this moment, the official ubuntuone client implementation uses python? [20:57] james_w: also pushed up fixes to lp:~dobey/ubuntu/karmic/icontool/karmic which should be "correct" now :) [20:58] Savago: yes [20:58] And how is the integration with some C apps (i.e. evolution) done? [20:59] we don't currently integrate with evolution [20:59] the e-d-s backend for contacts will talk to couchdb though, which will use couchdb's replication protocol to replicate to the server or other systems of the user's [21:00] Savago: but if you want to talk with the syncdaemon, it provides a DBus API [21:00] Ah... I see. [21:01] Another question: why to create a new protocol format? There are some many (syncml, google data, etc). [21:01] * Savago admits that is a somewhat naive question... [21:02] none of those have anything to do with sharing data really [21:03] err, sharing files even [21:03] and aren't necessarily developed to work well with off-line mode, horrible/slow connections, etc... [21:03] Savago: the format of the protocol is the same as any other protocol that uses "Google protocol buffers" :) [21:05] rmcbride, dobey: both packages reviewed and almost advocated [21:05] Another question: is there any roadmap (technical side) about ubuntuone? [21:05] dobey: apologies for not catching that issue first time around [21:06] james_w: looking at review and acting on any items. THanks! [21:06]