[00:29] <daveisadork> is there any way to prevent ubuntu one from replicating a specific database?
[00:37] <daveisadork> i'm writing an app and was hoping to use desktopcouch for the database... didn't realize ubuntu one would try to sync it
[00:37] <daveisadork> and when i delete the local database, ubuntu one re-syncs it (and all 20k documents) from the cloud
[12:18] <petur> Hello
[12:18] <Guest33632> If a file on my harddrive gets corrupt, (it's md5sum will be diffrent) because of bad sectors on my drive
[12:18] <Guest33632> And that file is in a folder which is to be synced with ubuntuone
[12:19] <Guest33632> erhm.. i guess what i'm trying to ask it
[12:19] <rtgz> Guest33632, the file will be synced to Ubuntu One if its checksum differs
[12:19] <peturdk> Alright
[12:20] <peturdk> What if the readonly flag is set?
[12:20] <rtgz> peturdk, if the file is readable then it will be synced. There was one such case discussed here with power failure and zeroed file
[12:20] <peturdk> Is there any way i can tell the file "STAY"
[12:21] <peturdk> I see
[12:21] <rtgz> peturdk, I think that this is not possible at the moment.
[12:21] <peturdk> And if i chmod it to 000 after it's been put into ubuntuone
[12:21] <peturdk> will it still sync on changes?
[12:21] <rtgz> Though it needs better investigation. The bad part for this is that it is hard to emulate that in virtual machine...
[12:21] <rtgz> peturdk, if ubuntuone can't read the file to be synced... hmmm...
[12:21] <rtgz> let me check
[12:22] <peturdk> oki
[12:23] <rtgz> peturdk, erm.. it will not sync the file, right
[12:23] <rtgz> but syncdaemon becomes crazy
[12:23] <peturdk> Oh okay i see
[12:23] <peturdk> And a final question, would 'rm -r /home/petur' delete all files from ubuntuone as well?
[12:24] <rtgz> constantly running hasher routine to get the hash, it fails and it retries, causing pretty high CPU load
[12:24] <peturdk> This is bad for me, because if any one of my workstations is to behave strangely, it would loose all my data
[12:25] <peturdk> You have been very helpful, thank you
[12:25] <rtgz> peturdk, if ubuntuone-syncdaemon is running and it is fast enough to propagate the changes to the server then yes. Chances are that it will face unexpected condition when cache will be removed and then it might try to create conflicts
[12:25] <rtgz> so... It is better to shut syncdaemon down before rm -rf your home
[12:26] <peturdk> Yes you are correct
[12:26] <rtgz> peturdk, and thanks for the hint for unreadable ( chmod 000) file - it looks that I will need to poke devs regarding CPU usage
[12:27] <peturdk> Always trying to think outside of the box :)
[12:27] <rtgz> peturdk, the idea to get multiple versions of the file has been raised here, but I don't think it is considered at the moment.
[12:29] <peturdk> Store, sync and share. Backup your life with ubuntu one.
[12:29] <peturdk> Offsite backups, as the page ads, should be safe from local disasters no matter how complicated
[12:30] <peturdk> I would love to see it improved :)
[12:30] <rtgz> Desktop+ guys, hasher becomes crazy when it is unable to hash the file, it tries to hash the file with no hold-off period. This leads to enormous CPU usage. Observed on 1.1.1+r321-0ubuntu1~ppa1~karmic
[12:31] <rtgz> peturdk, yes, there are a lot of rough corners at the moment, but the team works to get it all sorted, so that it would be really possible to back up someone's life
[12:31] <peturdk> :)
[12:37] <rtgz> peturdk, ok, the CPU problem is now confirmed: bug #491950
[12:48] <peturdk> rtgz: Great
[12:49] <rtgz> peturdk, now we need to get version support in.
[12:49] <peturdk> What is version support?
[12:49] <rtgz> frankly speaking, the versions might occupy the same 2Gb/50Gb space, i.e. it would be possible to "clean" older versions.
[12:50] <peturdk> Ahh
[12:50] <rtgz> peturdk, I mean when new version file gets uploaded to the server, the old one does not get lost, it is still there.
[12:51] <rtgz> okay, it is still there when file is removed, since the actual removal happens not in the realtime, but the upload might create new block, put the "latest revision" there, mark old one as obsoleted and show some hint for file version
[12:51] <rtgz> this can then be embedded in nautilus as some file propery tab and the web interface should give this info as well
[12:52] <rtgz> pretty handy for files that get updated often.
[12:52] <rtgz> If the space consumed by the obsolete files is counted towards user quota then it should be possible to "clean up" older file versions.
[12:53] <rtgz> but this is an informal description; need to get it written into the wishlist; But in your case that would help a lot.
[12:53] <rtgz> peturdk, since in case a corrupted version gets uploaded, it would be possible to recover the "known good version"
[12:54] <rtgz> okay, me needs to become away
[12:54] <peturdk> :D
[12:54] <peturdk> Good idea
[13:04] <diverse_izzue> the web frontend has some usability issues that seem like they would be very easy to fix. 1) file names are truncated even if there is more than enough space to show them entirely 2) instead of 2days, 1 month ago, say "1 month and 2 days ago"
[14:04] <rtg|away|z> diverse_izzue, 1) Bug #451997
[14:04] <rtg|away|z> diverse_izzue, the second issue does not seem to have any associated bug
[14:05] <diverse_izzue> rtgz, thanks
[14:06] <jblount> rtg|away|z, diverse_izzue: https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/457298
[14:07] <rtgz> don't know.. ubuntuone-servers might need a web-ui projecct or something; keeping HTML/JS related bugs as private simply makes more duplicates
[14:07]  * dobey likes duplicates
[14:08] <dobey> "me too" is totally useless
[14:09] <rtgz> dobey, need to tell users to use standard "mee-too" button
[14:09] <dobey> no
[14:09] <rtgz> jblount, nope, the bug is too private for now, can't see it :)
[14:09] <jblount> Sorry, should be public now: https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/457298
[14:10] <dobey> whether they say "me too" in a comment or click the "me too" button is useless information
[14:10] <jblount> rtgz: Just wait until Monday :)
[14:10] <rtgz> o_O
[14:10] <jblount> dobey: Isn't konwing how many people are effected by something useful?
[14:11] <rtgz> dobey, but it is much better than... bug #492645
[14:11] <dobey> jblount: only if they are actually affected by it. 99% of the "me too"s we get aren't actually the bug
[14:11] <dobey> rtgz: filing a bug in the wrong place is one thing.
[14:11] <rtgz> which is simply, erm.. a default bug text :)
[14:12] <jblount> dobey: Yeah, I guess that's true.
[14:12] <dobey> rtgz: yes, well, lp ui isn't really translated very well i guess, and not everyone understands the meaning
[14:14] <rtgz> dobey, translations... yes. Translated UI is one part. Translating content is what makes it that complicated.
[14:14] <dobey> rtgz: translating the content isn't the problem here.
[14:15] <rtgz> but that is beyond the scope of the original discussion, really; ubuntuone-servers bug reports are private by default and there is no "tick" when ticket is criated to make it public. So it goes like post, see that it was committed to lp and then run and hide, w/o marking the bug as public.
[14:16] <rtgz> Though I have no info on the number of ubuntuone-server bugs;
[14:16]  * rtgz needs to get rid of semicolon
[14:16] <dobey> rtgz: and bugs filed with apport that attach logs are marked private by default as well
[14:16] <dobey> rtgz: for Ubuntu
[14:16] <rtgz> true
[14:17] <dobey> and ideally, the web should never be involved in filing a bug
[14:21] <jblount> Is there a master bug for the notes web ui being broken ?
[14:21] <rtgz> jblount, there is a joint one for webui and tomboy sync, since at that time it all was failing in a similar way
[14:22] <rtgz> jblount, bug #501020
[14:22] <rtgz> will not do joined bug reports for such things anymore, I promise
[14:23] <jblount> rtgz: :)
[14:23]  * rtgz will do such joined reports for other things only
[14:28] <rtgz> jblount, bug #499595 is related to this, as statik said earlier
[14:30] <jblount> rtgz: Nice, thanks!
[14:31] <rtgz> jblount, those are tied together by christmas-bug tag
[14:31] <verterok> dobey: I marked #492645 dupe of #368626
[14:32] <jblount> Heh, all we got for Christmas was more bug reports!
[14:33] <dobey> verterok: ok
[14:59] <Lili^> hello
[15:00] <jblount> Lili^: Hi!
[15:03] <rtgz> ok, time to clean-up my mess with metadata-loading. Rebooting to reproduce the issue with stock karmic sources
[15:03] <aquarius> CardinalFang, the watch-the-changes-feed function that got built at UDS Dallas -- is that in released desktopcouch?
[15:05] <CardinalFang> aquarius, It's "released", but that version isn't packaged in Ubuntu.
[15:05] <aquarius> CardinalFang, cheers
[15:06] <CardinalFang> So, tarballs only at present.
[15:07] <dobey> ahem
[15:07] <urbanape> itzat time
[15:07] <urbanape> Desktop+ folks: MEETING BEGINS and/or MEETING STARTS - say 'me' to stake your claim
[15:08] <urbanape> me
[15:08] <jblount> me
[15:08] <dobey> meh
[15:08] <aquarius> me
[15:09] <CardinalFang> me
[15:09] <vds> me
[15:09] <rodrigo_> me
[15:09] <teknico> me
[15:09] <urbanape> DONE: Finished my subsequent client first time sync branch and went to test it - D'OH! Bindwood no worky with Firefox 3.6! Started digging into it.
[15:09] <urbanape> TODO: Lordy, fix Bindwood to work with 3.6 - current Bindwood has maxVersion of 3.5.*, so will likely need an emergency fix released soon, or we need to let people know not to use 3.6 with karmic.
[15:09] <urbanape> BLOCK: None, but frustrated a bit.
[15:09] <urbanape> jblount: HAMMERTIME
[15:10] <jblount> DONE: Got (nearly) back in the swing of things from being sick, Bug day
[15:10] <jblount> TODO: FACE DUTY, call Ubuntu Start page next steps
[15:10] <jblount> BOCKED: NOPE
[15:10] <jblount> dobey: GOGOGO
[15:10] <dobey> ☺ DONE: Debugged branch landing problems, Got branches landed, Filed #512552, Handed off tarmac stuff to ops+, Worked on #510351 (enable bipedal oauth in server)
[15:11] <dobey> ☹ TODO: Fix some bugs I filed (#510351, #510353, #510348, #510355)
[15:11] <dobey> ☹ BLCK: None.
[15:11] <dobey> aquarius: romane ite domum
[15:11] <aquarius> that would be "romani" :)
[15:11] <aquarius> ⚀ DONE: discuss music downloader design; have music library page send row to database; write and deliver Ubuntu Developer Day talk
[15:11] <aquarius> ⚁ TODO: split library page into three, write tests, and commit; test suite for music downloader design; have music library page send message to downloader daemon; make tomboy first-sync experience nicer
[15:11] <aquarius> ⚂ BLOCKED:
[15:11] <aquarius> CardinalFang, speak your wisdom
[15:11] <dobey> romani eunt domus
[15:12] <CardinalFang> DONE: added ~7 tests for server notes code that had no test coverage at all.  Found a few bugs.
[15:12] <CardinalFang> TODO: finish with tests and fix notes bug#511341
[15:12] <CardinalFang> BLOCKED: None
[15:12] <CardinalFang> vds, por favor
[15:12] <vds> DONE: landed branch to fix sms configuration in trunk, helped beuno in configuring working copy for mobile sync, triaged bugs for bug day, attended two Ubuntu Developer Week lectures, started new branch to implement the business logic to configure a mobile handset.
[15:12] <vds> TODO: ask a question to funambol support
[15:12] <vds> BLOCKED: nope
[15:12] <vds> rodrigo_: go go go!
[15:12] <rodrigo_> • DONE: Contacts picker work based on design team's designs. Looked at python-lxml version update for hardy. Moved JS code to its own file in libu1. Music store discussions with Stuart, Lucio and Alejandro. Looked at oauthdesktop for music store. Added categories field to contacts for evolution-couchdb unstable
[15:12] <rodrigo_> • 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? Add jslint tests to check. Remove autosave code in notes web editor. U1 client must intercept notregistered URL and use d-bus to call oauthdesktop.
[15:12] <rodrigo_>  U1 client adds back button to pages. U1 client interrogates library page to update download progress. Test suite for download daemon. geoip detection on server to forward to appropriate store
[15:12] <rodrigo_> • BLOCKED: no
[15:12] <rodrigo_> teknico, go!
[15:12] <teknico> DONE: bug triage; more debugging an error at first login (#513280); some testing funambol cared deployment (#511316)
[15:12] <teknico> TODO: more testing funambol cared and funambol exchange deployment (#511316); fix a user's problems accessing the contacts web IU (#506584)
[15:12] <teknico> BLOCK: still cannot access the web interface due to an error at first login (#513280)
[15:12] <teknico> next: MoonRiver
[15:13] <aquarius> wider than a mile?
[15:13] <teknico> aquarius, exactly :-)
[15:14] <jblount> teknico: My small brain can't solve that problem you mention, but you can subscribe at /plans/ if thtat helps while running locally :)
[15:15] <teknico> jblount, interesting, I tried that now, and whoever is logged in is getting somewhere :-)
[15:16] <jblount> teknico: :)
[15:17] <jblount> teknico: Probably the user you are logged in as is one of the ranodom users that you get. Sure smells like a open ID problem.
[15:19] <teknico> jblount, yes, same thing as jdo said
[15:22]  * jblount must be learning something if he said something jdo said :)
[16:26] <rtgz> my cleanup brings up the following questions: Bug 492610 - why would something crash if it receives dbus timeout; I could not reproduce that
[16:31] <CardinalFang> rtgz, Synchronous dbus functions raise exceptions in Python, on timeout.  Callers should catch them, or (Better Yet!) just use asynchronous calls.
[16:34] <rtgz> CardinalFang, yep, but I could not reproduce the crash itself; applet was sitting there, doing nothing, while timeout occured. In my case it remained doing nothing.
[16:35] <rtgz> ok, I will need to become away again for next 2 hours; In case there is something I broke with the duplicates, please let me know
[16:35] <CardinalFang> rtgz, no idea.  Maybe a different timeout preempted the first in your case.
[16:37] <CardinalFang> Damn.  Desktop is frozen on big computer.
[16:37] <CardinalFang> I blame jblount.
[16:38] <CardinalFang> Something in USB use causes it.  Second time it's happened this year.
[16:39] <CardinalFang> Wow, it took 3 minutes before caps+scroll LEDs started blinking.
[16:44] <CardinalFang> ah, appears to be nvidia driver.  Fun.
[17:32] <jblount> CardinalFang: Apologies, it was probably my fault. I break everything.
[17:33] <fagan> aquarius: I installed the music store plugin and I got "FIXME failed to correctly auth to the music store"
[17:33] <aquarius> fagan, that's because it's not ready yet. :)
[17:33] <fagan> Ah
[17:33] <fagan> I wanted to have a look :)
[17:33] <aquarius> fagan, it's targeted for alpha 3.
[17:34] <fagan> aquarius: ill be waiting :)
[17:34]  * fagan is excited since the UDS meeting about this 
[17:34] <aquarius> I'll make a big deal when it's ready
[17:36] <fagan> aquarius: are we allowed to know who the music provider is yet?
[17:36] <aquarius> mattgriffin, ^
[17:36]  * fagan has a bet that its amazon 
[17:37] <aquarius> we're in the process of putting together a FAQ about various questions, one of which is that one, so keep your eyes out for that
[17:37] <fagan> I will thanks aquarius
[17:40] <CardinalFang> fagan, I don't think it's secret, but we want to get the best deal we can out of them, so we want to make them worry a bit.  So we don't talk about that part yet.
[17:41] <fagan> the phrase "we got them right where we want them comes to mind"
[17:41] <fagan> Whoops quotation error there
[17:42] <fagan> "we got them right where we want them" comes to mind
[17:42] <CardinalFang> I love your optimism.
[17:43] <fagan> CardinalFang: thats my biggest quality
[17:43] <fagan> I dont know enough not to be anything but
[17:43] <fagan> Its hard not to be interested and optimistic if you know nothing about it
[18:02] <mattgriffin> fagan: very soon ;)
[18:02] <fagan> mattgriffin: Ah I can wait
[18:03] <mattgriffin> fagan: thanks. we think everyone will be very pleased
[18:04] <fagan> mattgriffin: well I was pleased it was even happening and at the UDS I was very happy how far the plans were
[18:04] <mattgriffin> :)
[18:04] <fagan> I was person talking way too much :)
[18:04] <mattgriffin> hehe
[18:28] <statik> wow, vish is tearing through the bug list, way to go
[18:31] <dobey> is that what you named your script?
[18:32] <statik> no, my script is named blammo
[18:37] <beuno> statik, so when are we printing Ubunto One t-shirts to give away to these productive community people?  :)
[18:38]  * beuno secretly wants an ubuntu one tshirt
[18:38] <dobey> you mean the community people can't be productive and print the shirts for us?!
[18:38] <dobey> :)
[18:38] <beuno> sure, and then we'll triage bugs for them!  ;)
[18:39] <dobey> ok
[18:39]  * dobey writes a 10 line python script to triage the bugs
[18:39] <CardinalFang> Triage *correctly*.
[18:40] <dobey> oh sorry
[18:40]  * dobey writes a 20 line python script to triage the bugs
[18:40]  * CardinalFang thinks it should be a 300 line C program, anyway.
[18:40] <beuno> if only...
[18:40] <dobey> beuno: what do you mean?
[18:41] <beuno> ...that was possible
[18:41] <beuno> :)
[18:41] <dobey> the python script?
[18:41] <dobey> it is
[18:42] <dobey> anything is possible in software
[18:42] <beuno> I'm tempted to ask you to prove it
[18:42] <beuno> :)
[18:43] <dobey> dude, we already *have* a python script running on one of the servers that triages some bugs for us. it's called apport-retrace
[18:43]  * CardinalFang casts Summon Gödel spell.
[18:43] <dobey> heh
[18:46] <statik> beuno, shirts would be cool
[18:48] <dobey> beuno: and if i didn't have so much work to do on this client app thing, i'd probably have written 3 such scripts already
[18:48] <dobey> verterok: ping
[18:49] <verterok> dobey: pong
[18:50] <dobey> verterok: hey. it's not quite clear to me. i want to add a dbus call to syncdaemon to get the free space info, since there isn't one already. but i'm not sure how to tell the action queue to do a specific command
[18:51] <dobey> verterok: do I need to add some code to volume_manager.py or something to make the action queue calls?
[18:52] <verterok> dobey: hmm...let me take a look
[18:53] <verterok> dobey: the call is: action_q.inquire_free_space(volume_id)
[18:53] <verterok> dobey: in which object you want to expose this method?
[18:53] <verterok> dobey: in which DBus object
[18:54] <dobey> oh ok, and it's self.dbus_iface.action_queue
[18:54] <verterok> dobey: right
[18:54] <dobey> verterok: well the signal for FreeSpace is on Shares, so I presume it should go there as well
[18:54] <dobey> verterok: though maybe both of them belong somewhere else
[18:55] <verterok> dobey: we can change that, as we now have UDFs
[18:55] <verterok> dobey: right
[18:55] <verterok> dobey: what do you think about Status?
[18:56] <dobey> it should probably be next to the account info bit i guess. but i'm not sure why it's currently under Shares
[18:57] <verterok> dobey: no idea why it's there
[18:57] <dobey> hmm, and why aren't the free space, and total space, in the same message in the protocol
[18:58] <verterok> dobey: the free space signal are actually free space for a specific share
[18:58] <verterok> dobey: in the case of UDFs it's the same value for UDFs and for the root, but each share has it's own free space value
[18:59] <dobey> well, shares to you show the free space of the owner's account i would guess
[18:59] <dobey> but why doesn't it show used space also
[19:00] <dobey> the storage protocol method of getting this info is currently pretty expensive :-/
[19:02] <dobey> and i don't know if we should change the protocol right now
[19:03] <dobey> blah
[19:17]  * rtgz is no longer away
[19:18] <rtgz> is there any bug report regarding ubuntuone sharing clashing with samba sharing; One can share folders via ubuntuone but "Share" tab on file properties will show that sharing is not enabled
[19:19] <joshuahoover> rtgz: i don't ever recall seeing a bug along those lines
[19:19] <rtgz> joshuahoover, he he... freshbug
[19:26]  * dobey thinks WONTFIX
[19:26] <dobey> at least, not for lucid
[19:27] <dobey> though i'd be happy to see nautilus-share not be installed by default
[19:27] <rtgz> dobey, I would be happy to see nautilus-share NOT using samba, or being pluggable
[19:28]  * rtgz thinks that Nautilus Plugin that uses other plugins is not a good idea
[19:28] <dobey> rtgz: i already started working on something like that years ago
[19:29] <dobey> rtgz: extensible and simple, but alas
[20:06] <dsager> hello everyone, a question regarding the notes stored in the U1 cloud: are the notes somehow encrypted? i'd like to sync personal notes (like PW lists, bank accounts) with U1 but I'm not sure if the data is save...
[20:07] <statik> hi dsager: they are not encrypted. the notes are transmitted over SSL so not likely to be sniffed, but they are not encrypted.
[20:08] <dsager> ok, so anyone with access to the U1 filesystem could read them, right?
[20:08] <statik> dsager: the sysadmin who has access to backup the couchdb could read them if he wanted to get fired and go to prison. nobody else could read them.
[20:09] <mandel> statik, love that answer ;)
[20:09] <dsager> hehe, thanks for the info!
[20:09] <statik> dsager, we've been getting feature requests to allow people to share notes, so in the future you might be able to grant access to other people. but for now it's only you that can access them.
[20:10] <dsager> ok, that sounds good! thanks again!
[20:44] <CardinalFang> rodrigo__, you around?
[22:10] <jugger90> hi
[22:11] <jblount> jugger90: hi!
[22:12] <jugger90> I have problem on note saving from web ui yestaday
[22:12] <jugger90> i can save a note and del a note but can't edit
[22:13] <jugger90> i can creat
[22:13] <jblount> jugger90: Let me test that for my account and see if I can recreate.
[22:13] <jblount> jugger90: Do you get any error msg?
[22:13] <jugger90> no i don't get any msg
[22:14] <jugger90> just say There was a problem with saving your note!
[22:14] <jblount> jugger90: Do you see, "There was a problem with saving your note!" at the bottom?
[22:14] <jugger90> i can see that
[22:16] <jblount> Oh, ok. This is probably related to some known brokeness, let me find you the bug.
[22:16] <jugger90> thank you
[22:16] <CardinalFang> jblount, I have this bug.
[22:16] <CardinalFang> jugger90, thank you.  We know about it and will have a fix very soon.
[22:17] <jugger90> i see
[22:17] <jblount> jugger90: https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/501020 seems to be it, I think if you follow that bug you'll know when it is resolved.
[22:17]  * jblount high-fives CardinalFang 
[22:17] <CardinalFang> jugger90, there were two bugs.  We fixed the one already, but discovered there was another waiting behind it.
[22:18] <jugger90> oh...
[22:19] <CardinalFang> Bug #501020 isn't your problem.  Yours is a private bug, unfortunately.
[22:19] <jugger90> I will mainly use web ui from windows so,this bug make note useless