=== afk is now known as mthaddon === afk is now known as mthaddon [01:45] * BUGabundo $ sudo shutdown -t now ; echo 1 lamb, 2 lamb, 3 lamb 4rrrr lammbbbb..... === jdo_ is now known as jdobrien === mthaddon is now known as afk [07:48] What is the best way to reset Ubuntu One? I've deleted everything on one computer, but it's still on the Web frontend and the other computer. And on the web frontend, the "delete" button has no effect. [07:49] hm, probably the simplest way to do this is to use u1sync [07:49] create a directory somewhere -- not under "Ubuntu One" [07:49] (I want to start with a pritine setup to analyse the bugs.) [07:49] s/pritine/pristine/ [07:49] run `u1sync --init` on that directory [07:50] then run `u1sync --action=clobber-server` inside that directory [07:50] that should delete everything from the server [07:51] (u1sync is sort of a "side band" way of dealing with your storage) [07:51] Okay, I'll try that, thanks! [07:51] (the first command sets up that directory to be a mirror of your storage, and the second does a sync which forces the server to match what it has locally [i.e. nothing]) [07:52] after you've done that you can delete the directory you used with u1sync [07:53] Is ~/Ubuntu\ One/My\ Files/ already initialised this way? [07:53] no [07:53] this is a separate system [07:53] Okay, this explains it. :-) [07:54] mainly used for testing and server stuff [07:54] ~/Ubuntu One/My Files is managed by a daemon (ubuntuone-syncdaemon) [07:54] it's better not to mix the two on the same directory(s) [07:55] s/separate system/separate client/ [07:55] Mmm... since most of my problems will probably have to do with the daemon, it's better to get the My Files reset. [07:57] probably the best way to do that is to kill the daemon [07:57] and remove ~/.cache/ubuntuone/syncdaemon [07:57] then restart the client applet (which I believe should restart the daemon) === tcole1 is now known as tcole [08:21] Does anybody else have trouble with uploading files through the web interface, too? I get proxy errors always. [08:21] https://bugs.launchpad.net/ubuntuone-client/+bug/388827 [08:21] Launchpad bug 388827 in ubuntuone-client "Proxy error with upload" [Undecided,New] [08:22] (The reported package is certainly wrong, but I can't find a better one.) [10:34] Now I have the situation that on both client machines, My Files is empty, the daemon was restarted, the Web interface is still full of the files, but the delete button has no effect. Great. :-) Apparently I srewed up the thing totally. What's the remedy? [12:18] Well it appears that in order to a large download...lots of files. You need to disconnect/connect the client over and over [12:19] hmm: 1010.5 MB of 10.0 GB Used (9.9%) i would have thought the django fileformatter would have made that a partial GB [12:22] question, i think this is a decent feature request, but wanted to talk about it. [12:22] i have 3-4 jaunty worksations that particiapte in ubuntuone sharing with my fileset [12:23] when it is scanning and syncing, it would be nice to have some sort of mouseover progress indicator [12:23] maybe even some way to give particular files priority [12:23] thoughts? [12:28] slestak: that's a good suggestion...it's also in our to-do list :) [12:29] slestak: theres bug #330769 [12:29] Launchpad bug 330769 in ubuntuone-client "Upload/Download Status Information" [High,In progress] https://launchpad.net/bugs/330769 [12:30] doh [12:30] slestak: and bug #377960 [12:30] Launchpad bug 377960 in ubuntuone-client "ubuntuone-client enhancement" [Wishlist,Triaged] https://launchpad.net/bugs/377960 [12:30] oh cool...one of them is in progress...sweet [12:31] tell me if I'm being silly here. I have personal work and work work. I have two sets of hg repos. I am having a hard time not thorowing my repos into ubuntuone [12:32] i wouldn't recommend doing that [12:32] i know I can clone, but i start so many small, one time use things, i may just like to say here is all my stuff, and I will clone my work stuff, but I want all my personal src to be synced [12:32] at any given moment, I can be on any one of 4 workstations [12:32] slestak: switch to bzr, use launchpad ;) [12:32] hehe [12:33] i have aix in the mix, and hg installs easily in win32, linux, and aix [12:33] also bzr has a way to announce a branch on a local lan so one computer can host it..like launchpad [12:33] hmm, that looks compelling [12:33] does hg have something like bzr so you can setup a repo accessible over a lan [12:34] yeah, it has hgweb, but i was using ssh for updating, and just using the web portion for non tech users ability to browse repo and research [12:35] jdobrien: the "announce branch" part sounds different. is that sth that bzr has exclusively, or is that just serving upa repo by http? [12:36] the problem with using u1 for would be that files are in a different state of upload as it's designed to shove files over the internet at a slower speed than possible over a local lan [12:36] with a dvcs one prob i have is rememebring what code is on what workstation in what level of completion [12:37] thats what i hoped u1 could help with [12:37] let me look up the plug in [12:37] if i knew a branch was not completely synced, i could find sth welse to do while waiting, but the script i started the night before in a coffee shop, would be automatically brought to my work pc [12:38] two things i found: https://launchpad.net/bzr-dbus [12:38] thats seems to just be a commit notification method. not liek a zeroconf announcer [12:39] if i connect to the lan and had some repos available, would be cool if sth would say, ive got some python over here, heres some here too. here is some unibasic [12:39] and http://doc.bazaar-vcs.org/bzr.0.18/server.htm [12:40] yeah, hg allows clone and update via ssh and http, im just still learning them [12:40] ahh [12:41] by the time i know i need some code off one of my ubuntu workstaitons, it may be powered off at home [12:41] but for u1, I think you would run into problems given it's current design [12:41] looks liek for my workflow i need access to an always available server to push to [12:42] have you tried a shared folder? [12:42] well, ive taken you OT. ty for pointing out the relevant bug reports [12:42] well, the shared folder only works if the hosting machine is powered on [12:43] hehe...indeed [12:43] thats what i was thinking u1 could provide, an always on shared folder [12:43] stupid computers [12:44] slestak: it would work as long as you didn't expect instant synchronization [12:44] maybe a feature for u1 could be to assign priorites to folders in metadata so we could say, please do this first, always [12:44] slestak: no waste of time at all...its interesting to know what users want to use a tool for...helps us determine how to prioritize features too [12:45] slestak: thats an interesting concept [12:47] my stuff is definitely low risk, just learning python, so I may try it for a while [12:47] i'll report back if it all breaks [12:49] i had a strange u1 prob this morning. using nautilus, i used drag-n-drop to copy komodo-edit tarball out of u1. i accidentally moved it instead of copy, so i immediately copied it back to u1. well, now I appear to have two copies, one with a .conflict suffix [12:50] before I copied komodo.tgz back into u1, it was definitely not present, but it probably was just marked as deleted instead of physically removed. There was only like 10 seconds between removing it and adding it back [12:51] should it have detected that the same file was just readded and it already had it available? could save syncing time to compare md5 or something if it thinks it has it [12:53] hmm, not that i look at it, i have zero byte .conflict files for lots of files, but the komodo.tgz has a 37.6MB .conflict file [12:54] How much can a company firewall affect U1? [12:55] bronger: if it blocks out going connections it could [12:56] I don't think so, but apart from ssh, all incoming connection are blocked. [12:56] bronger: there are no incoming connections [12:56] bronger: are you seeing a specific problem [12:57] Yes: If I say "touch foobar" in the My Files directory, the file is created and uploaded to U1. And in *this* moment, the daemon starts downloading all the file which are only visible through the web interface. So I wondered whether notifications from outside were blocked. [12:59] bronger: what do you mean by "the daemon starts downloading all the file which are only visible through the web" where did these files come from? [13:00] I had uploaded a *lot* of files (my mail dir actually) to U1. Then, I deleted them on the client. However, they were not deleted in the web interface. [13:01] By the way, U1 is not good at many small files. ;-) [13:02] Hi all [13:02] bronger: yeah i have found i need to stop/start the client to get it to work [13:03] bronger: I am currenltly uploading 2.6G of files [13:03] hi facundobatista [13:06] mattgriffin: Morning! [13:06] jblount: good morning :) [13:31] jdobrien: i take it your not on the 2g beat plan :) [13:32] s/beat/beta [13:33] slestak: :) [13:33] slestak: I didn't know using up my 10GB storage would be so much work ;) [13:34] 1.2 GB of 10.0 GB Used (11.9%) [13:34] i could use 10g, i want to move my ~/Music folder [13:34] but i should just setup ampache. would be more bandwidth friendly [13:34] are there going to be more choices than 2g and 10g? [13:35] slestak: I think we're planning on doing something a little more a la carte (get 10 extra GB for x dollars a month) [13:36] slestak: For this first release having _any_ sort of paid plan was important because we wanted to make sure all the moving pieces to make that happy were in place :) [13:38] can we discuss my .conflict issue from about 1 hr ago? [13:39] slestak: Sure, do you have updates? (it'll take me 1 second to read the traceback) [13:41] slestak: sorry..i had to look at another issue [13:41] facundobatista: yo! What should users do with .conflict files? Is there somem way to resolve them? [13:42] jblount, they should use them to fix the original file... (in the best case they just move it into the normal file, but there was a bug in the move operation in that case... don't know if it's already solved... __lucio__?) [13:43] <__lucio__> facundobatista: not yet, in progress [13:44] __lucio__, if the user deletes the original file and renames the foo.conflict to foo, it works? [13:44] <__lucio__> facundobatista: should work. [13:45] slestak: ^^ (emphasis on _should_) [13:46] i have a _lot_ of zero byte .conflict files, and the one .conflict for komodo that is == to the original file [13:46] it was a quick copy out, put back into u1 [13:47] slestak: Are they all named the same thing? (the zero byte .conflict files) [13:47] named after the original files, I can supply a screenshot or pastebin it you want to see [13:47] slestak: That'd be great, thanks! [13:48] i just got a work email i need to handle, i may be ~1 hours [13:48] slestak: No worries, we'll be here ;) [13:55] * jblount dances around to Alkaline Trio [13:55] jblount: http://pastebin.com/f44fa82bc [13:59] slestak: Nice, thanks. It seems the zero byte conflict files are "server knows about the file, but didn't actually get it" and the fullsize komodo .conflict is "this is the version the server has, it's different than your local machine and we weren't sure which was right" [14:00] slestak: Why don't you try to delete the zero byte conflict files and see what the client does on your machine? [14:04] hello world [14:04] statik: hiya! [14:07] I am getting a constant spinning logo [14:07] hola [14:07] and every two minutes in the log I get this [14:07] buenos morningos [14:07] 2009-06-18 14:06:10,863 - ubuntuone.SyncDaemon.Main - INFO - ---- MARK (state: READING_WITH_NETWORK_WITH_BOTHQ; queues: metadata: 26; content: 19; hash: 0) ---- [14:07] alanbell: what does the tooltip on the spinning icon say? [14:07] jdobrien: please stop breaking the syncdaemon :-P [14:08] alanbell: READING means it's not yet connected, just doing hashes of your local stuff [14:08] dobey: Ubuntu One: Scanning [14:08] alanbell: has it been in that state for long? [14:08] Chipaca: several days [14:08] also, why is it READING with a hash queue of 0? [14:08] there are some files on the web it hasn't pulled down [14:08] Chipaca: no! [14:09] hmm [14:09] alanbell: could you submit a bug, attach the logs, and then restart? [14:09] dobey: Would you mind pointing your massive brain at this bug when you get a few minuts? #378707 [14:10] Chipaca: when our near release time, I was thinking of promoting the "I can break U1 Challenge" [14:10] alanbell: the logs I'm interested are the ones in ~/.cache/ubuntuone/log/syncdaemon.log* [14:10] alanbell: as it's been "several days", the one submitted by apport probably won't be relevant [14:11] alanbell: if you can attach the one where it lands in that state, that would be great [14:11] ubottu: bug #378707 [14:11] Launchpad bug 378707 in ubuntuone-client "UbuntuOne client applet silently fails to connect" [Undecided,Triaged] https://launchpad.net/bugs/378707 [14:11] Chipaca: OK, will look back through it [14:11] alanbell: you can grep for '-> READING_WITH_NETWORK_WITH_BOTHQ' [14:12] jblount: hrmm [14:13] dobey: I got soemone poking me about it on twitter, just wondered what it seemed like to you. [14:14] jblount: my .conflict files are ll older than the original files, see Centos-5.zip, freelink...zip, etc [14:14] Chipaca: seems to have gone to that state at 2009-06-13 09:58:10,863 [14:15] jblount: well the problem is that the syncdaemon is crashing. the applet doesn't know /why/ it happens, it just gets the dbus error, which is pretty much entirely useless, and logs it. :-/ [14:15] hi all, just got my invitation to ubuntu one [14:15] dobey: grumble. [14:15] javi: Neat! [14:15] does symlinks work? [14:15] dobey: jblount: the syncdaemon is crashing? [14:15] javi: no [14:15] jblount: i can probably write some funky "crash dialog" hack, but it almost certainly won't work in all cases [14:16] any plan on supporting it? [14:16] Chipaca: http://launchpadlibrarian.net/28002187/.home.michal..cache.ubuntuone.log.syncdaemon.log.txt [14:16] javi: yes. And there's a bug you can +1 to prod us to get it done faster :) [14:16] javi: Not yet, maybe someday. Some people are ln -s ing into their Ubuntu One folder though. [14:16] symlinks are very very difficult to get right [14:16] excellent, thank you [14:17] dobey: +1 (that symlinks are difficult) [14:17] hmm, symlinks dont work? i wasnt aware of that. not using them, but i guess i should take that off the table [14:17] because the correct behavior is not something we can guarantee to be able to do on every system where ubuntu one is running [14:17] Chipaca: http://pastebin.com/f3dfb74f2 [14:17] dobey: last time I saw that, the user had inadvertently previously run the syncdaemon as root [14:18] alanbell: nope, before that [14:19] Chipaca: that log isn't the original reporter's same problem, but the resulting experience is the same [14:19] dobey: right [14:19] Chipaca: perhaps for that case though, we should handle the OSError, check the permissions with stat(), and propagate a useful error message to the client via dbus [14:20] dobey: didn't we have a bug about having a crash handler let dbus know? [14:20] slestak: I see that they are older, does that change how you feel about deleting the zero byte .conflict files ( context: http://pastebin.com/f44fa82bc ) [14:20] Chipaca: what am I looking for? the full log file is 2.4Mb and contains lots of file names I don't want to attach to a bug [14:21] jblount: i have no issue with deleting them. i just dont understand the product yet, and didnt want to screw it up w/o knowing [14:21] Chipaca: i don't know. currently we don't have any crash handler sending a message over dbus. syncdaemon exits, dbus reports the failed to start error as a totally useless error message [14:21] alanbell: hmm... something logged at ERROR, probably [14:21] and the client applet just fails to connect [14:21] dobey: that can be construed as suboptimal [14:21] 2009-06-13 09:57:28,120 - ubuntuone.SyncDaemon.local_rescan - ERROR - in the scan: ('/home/alan/Ubuntu One/My Files/BellLord/The Open Learning Centre/Roundlens/Roundlens_project brief_v0.1Draft.pdf') [14:22] grr [14:22] yes, yes it is [14:22] facundobatista: ping [14:23] Chipaca, pong [14:23] facundobatista: when you have a minute, could you give alanbell a hand at reporting a bug wrt local rescan, without him uploading the full logs (privacy concerns)? [14:23] slestak: Like facundobatista said, I would suggest getting rid of the zero byte .cofnlict files, copying your komodo file to a *safe place* and renaming the komodo .conflict to Komodo-Edit-5.1.3-3592-linux-libcpp6-x86.tar.gz [14:24] alanbell, do you have a traceback after that ERROR line? [14:24] Chipaca, ok [14:24] slestak: But do be aware, we are truly in beta. I wouldn't want you to lose time / work because of a bug :) [14:24] alanbell, in the log, I mean [14:24] facundobatista: sure, will pastebin . . . [14:25] can I safely chmod u+w Ubuntu One folder? [14:26] ooh, a python traceback yes [14:26] Chipaca: ^^ (can javi chmod u+w the Ubuntu One folder?) [14:26] alanbell, exactly [14:27] javi: you can, but I think syncdaemon 'll fight you and change it back [14:27] http://pastebin.com/f10e0125d [14:27] verterok: Fight to the death? (welcome back from vacation!) [14:27] boas tardes [14:27] for reference that was "grep -n -A 30 -B 10 ERROR syncdaemon.log.2009-06-14_09-55-13 |pastebinit" [14:27] jblount: hi! thanks! [14:28] BUGabundo: j0! [14:28] jblount: don't know if to death, but it 'll try ;) [14:28] hey jblount [14:28] verterok: ok, thank you [14:28] statik: https://edge.launchpad.net/changeup [14:29] dobey: Nice! [14:29] jblount: what verterok said [14:30] javi: or, to put it another way, you can, but you might break things down the road [14:30] jblount: Pantomime Horses fighting to the death? [14:31] dobey: heh [14:31] dobey, great! [14:31] dobey, Chipaca: Mind if I copy / paste a bit of the conversation you had about that bug into a commment on it? I want the people waiting on it to know you are thinking about it actively. [14:31] javi: also: exactly why do you want to do that? :) [14:31] Chipaca: if that happends, do i report it as a bug or should i as i'm doing something not supposed do be done? [14:32] Chipaca: it's just to change the folder icon :-D [14:32] javi: oooh, an actual, real, sensible use case! [14:33] Chipaca: i don't plan to make any other changes to the Ubuntu One folder :D [14:33] javi: things breaking, you mean? When/if that happens, if we don't warn you such that you know not to, it's a bug and you should report it [14:33] jblount: no, I don't mind. [14:34] javi: to set a custom icon in nautilus? [14:35] dobey: it's under kde actually [14:35] ah [14:35] i don't now how konqueror or dolphin works [14:36] but nautilus stores the metadata for setting a custom icon in ~/.config or somewhere like that i think, so you don't need to have write permissions to set a custom folder icon [14:36] dobey: basically you need to add a .directory file to the folder you have customized, so you need write permisions [14:36] because .directory files are deprecated... :) [14:36] hrmm [14:36] oh well :) [14:38] javi: I'd love to have that use case documented if you had a few minutes to file a bug: http://bugs.launchpad.net/ubuntuone-client [14:38] jblount: of course, no problem [14:39] javi: Thanks! [14:39] anyway, what's the point in making it read only? the daemon looks only for this folder name? [14:40] dobey: in which network are you trying to reach segphault? [14:40] irs.arstecnica.com ? [14:40] javi: It's because the folders users should be interacting with are ~/Ubuntu One/My Files and ../Shared With Me respectively [14:40] lololol [14:41] s/irs/irc/ duh [14:41] javi: ~/Ubuntu One/ is just kind of a place holder, although I *belive* we'll be changing that at some point to work in a bit more sensible way. [14:41] BUGabundo: well i don't see him on freenode or gimpnet [14:41] he is not! [14:41] dobey: only on "his" own irc network [14:41] yes, /whois told me that much [14:42] dobey: The #gwibber chan is on arstechnica's thing. [14:42] s/thing/irc network [14:42] he was on here durring UDS :) [14:42] jblount: I said as much [14:42] dobey: didn't saw him [14:42] guess it was a fluke ? [14:42] he was remotely participating in the uds session about gwibber [14:43] BUGabundo: He was on freenode because we were chatting with him about how best to integrate gwibber with Karmic [14:43] yeah I guessed as much [14:43] dobey: his IM is ryankpaul@gmail.com [14:43] whoops, that should have been PMed, spam time! [14:43] jcastro: LOL [14:44] ok, but shouldn't it be a user choice? I mean, I understand that for the new user the easyest is to identify Ubuntu One folder with it's Ubuntu One service, and My Files and Shared with Me with those folders, but for a user that knows what he is doing, why can't he/she make a "Sync" folder and use it, or selecting other path instead of ~? [14:44] jblount: ahg i didn't read you before sending [14:45] jblount: that's great [14:45] javi: there are bugs about this already, as well :) [14:45] alanbell, the lines 9813 to 9816 in the pastebin, before the error, are actually together in the real log? Nothing in the middle? [14:46] javi: Ease of use is more important to us _right now_ than making it work for more advanced users, but that's totally a valid point. [14:47] jblount: i understand it completely, user behaviour can be frustrating, like someone wanting to change the folder icon... lol [14:47] javi: :D, but I agree with you, if you want to change the icon you should be able to. It's just a matter of what to focus on first. [14:47] facundobatista: err, yes. [14:48] unless grep is lying to me [14:48] javi: I'm having a hard time finding the bugs about "sync folders, it shouldn't matter where they are", but they do exist (like dobey said) [14:50] why thank you pidgin, because your error message is so bluntly obvious what the problem is [14:50] Wow, that "iwatch" package is useful. [14:50] "Could not send message (Code 503)" [14:51] dobey: Switch to empathy, same great taste, less filling :) [14:52] uhm, no. [14:53] at least, not on jaunty [14:54] well I can confirm that after exiting the client and starting it again, the permissions are only-read again as you said [14:59] CardinalFang, using iwatch to see if your zcml is being read? [15:00] statik: Something similar. [15:00] meeting time! [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 [15:01] i think vds, teknico, and aquarius are still sprinting this week. urbanape, dobey, jblount, rodrigo_? [15:01] me [15:01] me [15:01] me [15:01] me [15:01] DONE: lots of phone calls and discussion for Karmic Feature Definition Freeze today. Screensharing design discussions. Pushed a patch upstream for license headers in the spawning wsgi server, fixed the make -j 4 bug in ubunet, saw our first contributor license agreement signed! [15:01] TODO: More Karmic planning, work on OOPs with pfibiger, look at glib-couchdb with rodrigo. [15:01] BLOCKED: None [15:01] next is CardinalFang [15:01] DONE: Got tests running better; some things dislike karmic still. Bypassed ZCML and interface for zope.sendmail in favor of our own config. [15:01] TODO: Verify that everything is working. [15:01] BLOCKED: Nyet! [15:01] jblount: tag! [15:01] DONE: paired up on the /files/new UI with urbanape, figured out the fix for our squid caching stuff [15:02] TODO: FACE duty! [15:02] * jblount tags rodrigo_ [15:02] BLOCKED: I think I'll need to hand off the squid fix to someone else, but I'm not sure who (and I'd like to have it done today before we do a rollout) #388232 [15:02] done: more evo-couchdb work and debugging of CORBA issues [15:02] todo: fix CORBA issues with my backend, and once that works, add the few evo backend methods missing [15:02] blocked: corba :) [15:03] alanbell, it's *very* strange.. it's kike if you had the metadata for that file twice... it's very strange... [15:03] urbanape: your turn [15:03] DONE: Worked with jblount a bit, but we spent a lot of time reading YUI docs together. [15:03] TODO: Continue with /files/new UI and FF extension [15:03] BLOCKED: None, other than familiarizing myself with YUI more [15:03] dobey, you're up [15:03] DONE: Client build system hacks, tarballs released, protocol and client upload to REVU, Created ChangeUp project on LP for app restarting [15:03] TODO: move share creation via web api code to syncdaemon, split ubuntuone-client packaging, backport python-oauth, get some code in ChangeUp and Central Services [15:03] BLCK: None. [15:04] alanbell, if you quit the client (close it completely), and restart, does it happen again? [15:04] urbanape, can you work on the squid caching thing with jblount? i'm available for a pre-implementation call to discuss the solution if you want [15:04] sure thing [15:04] hi teknico, are you joining the standup meeting today or is your client just connecting automatically? [15:04] dobey, congratulations on the uploads! [15:04] i guess thats everyone, MEETING ENDS [15:04] facundobatista: will give it a go, are we interupting the meeting here? [15:04] statik: not automatically, I just realized that I didn't show up at all these days :-) [15:05] oh, guess not :-) [15:05] facundobatista: is closing the client from the toolbar sufficient? [15:05] teknico, no worries, i don't expect people to attend standup meetings while they are at a sprint, but if they do it's a delightful bonus [15:05] statik: thinks [15:05] err [15:05] thanks [15:06] facundobatista: well restarting the client by closing it in the toolbar and relaunching it did nothing. Still spinning and reporting the same thing in the log [15:06] alanbell, try "ps -eaf | grep syncdaemon" to be sure it's not in memory [15:07] facundobatista: it is down [15:08] oh, now it has stopped spinning [15:08] odd [15:08] alanbell, our client fixes itself... "import skynet" magic [15:08] facundobatista: +1 :D [15:09] ah, it started spinning again [15:09] statik: Yay, contributors. [15:10] but it appears to be doing constructive work now [15:10] alanbell, always check logs [15:10] doing a tail -f of the log [15:12] __lucio__: yo! When you get a moment, can you point your brain at this bug? I think marked as "later" is wrong, but wanted to ask: https://bugs.edge.launchpad.net/ubuntuone-client/+bug/357572 [15:12] Launchpad bug 357572 in ubuntuone-client "Top level hierarchy seems to be ignored" [High,Triaged] [15:13] Is there any details on the couchDB side of things? I know Daimien Katz back from the Lotus days and I was writing a python GUI client for couchDB a while back. [15:15] alanbell, lots of details! we're working on making couchdb available on the desktop by default in karmic [15:16] statik: that would be fantastic. [15:16] and are working on documenting some good patterns for apps to integrate with it instead of sqlite or flat files [15:16] alanbell: https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-karmic-integrating-with-ubuntu-one <-- a blue print of stuff that we're planning on [15:16] there has been no discussion of ubuntuone on the couchdb mailing lists up to now [15:16] and it isn't listed here http://wiki.apache.org/couchdb/CouchDB_in_the_wild (for the next 5 minutes or so at least . . .) [15:18] alanbell: We're still not _quite_ there yet, but I think you'll start to hear noise on it pretty quickly on the couch mailing list and similar places [15:18] alanbell: wow, we aren't very good at tooting our own horn ;) we've got some trivial patches into couchdb already, it was discussed a bunch at UDS and various meetups [15:19] alanbell, let me find the public blueprints for you [15:19] https://wiki.ubuntu.com/DesktopTeam/Specs/Karmic/IntegratingWithUbuntuOne [15:20] alanbell, how are you using couchdb these days? i'm always interested to hear about what people are doing with it [15:21] statik: well I am not really using it much. I wrote some of the early documentation and a LotusScript library for it. [15:22] back in the pre-JSON XML days [15:22] I then did a client called Valance which I never really finished [15:23] https://code.launchpad.net/~alan-dominux/valance/trunk [15:24] http://www.dominux.co.uk/index.php/2008/04/08/and-another-workspace/ [15:25] facundobatista: my client has now gone to a disconnected state [15:25] alanbell, ¿? reason? (you have the reason in the log) [15:26] facundobatista: nothing obvious. no ERROR [15:26] alanbell: look for a SYS_NET_DISCONNECT o similar [15:27] verterok: nope, haven't got one of them. [15:27] alanbell, what are the few last lines of the log? [15:28] http://pastebin.com/f7eed341c [15:28] statik: it's hard to keep up with "the web" :) [15:30] alanbell: it was connected and then disconnected or you just restarted the client...you need to manually connect it [15:30] verterok: well it just spent 10 minutes or so doing stuff [15:31] alanbell, facundobatista: I think that was just the local rescan [15:31] verterok: ah, I see. [15:32] alanbell, oh, yes... the local rescan is done without connecting, and the client does not connect automatically, so... [15:32] alanbell, grep for ERROR in the new log [15:32] but it did have SYS_NET_CONNECTED when it first restarted [15:32] no ERROR [15:33] alanbell: yes, that's NetworkManager talling the syncdaemon that there is network available [15:33] alanbell, don't know if SYS_NET_CONNECTED is that it's connected to the network, or that it has a connected network that it could ues [15:33] *use [15:34] it's can use [15:34] alanbell: to check if it connected or if a connection was requested, look for a SYS_CONNECT [15:34] if it's connected to the network and doing nothing, the state is IDLE [15:35] it probably went READY -> LOCAL_RESCAN -> READY [15:35] verterok: no SYS_CONNECT until just now when I told it to connect. [15:35] alanbell: ok, so that was the reason :) [15:36] it is now working [15:36] doing push events [15:39] what does fsm stand for (apart from Flying Spaghetti Monster) [15:39] alanbell: finite state machine, or file system manager [15:39] alanbell: depends :) === yofel_ is now known as yofel === Chipaca_ is now known as Chipaca [15:40] <__lucio__> jblount: back. So, we are not paying attention to folders in Ubuntu One that were not created by the system. If someone wants to create a folder, he ahs to do it in My Files. or inside a share he has rw access. This has proven to create confusion, so we are planning on changing it. To what is still not clear. So thats why its marked as later. [15:41] __lucio__: Cool, I just wanted to poke you about it since I saw that joshuahoover triaged it. I'll poke you about it again in a few weeks when you've worked off more of your invite queue :) [15:41] <__lucio__> :) [15:44] __lucio__, jblount: also the user must explicitly change the Ubuntu One folder permissions to create this folder :) [15:44] jblount: triaged it, or moved it from beta2 to later? [15:45] jblount: everything that had the beta2 milestone was moved to 'later' [15:47] dobey: Right, but previously on the same bug statik mentioned wanting to do this soon. Just didn't want to lose track of that particular bugs importance. [15:47] dobey: The client re-write you just landed is in C yeah? or is that C++ ? [15:51]