[00:29] <phxheat1> I would like to expand my Recent Documents from 10 to about 20 showing in the list. Places>Recent Documents. A link or a path would be fine, if you can even change it
[05:53] <foxxtrot> Hmm, this is strange, the last desktopcouch-replication-log update on my 9.04 box is from the end of November.
[06:21] <foxxtrot> Wait, not 9.04, 9.10.
[06:33] <foxxtrot> And nevermind, just added a record to one of my replicated databases and it synced.
[08:30] <rtgz> honk-keepalive
[14:21] <jblount> urbanape: I need your help when you get available on taking the public urls rest api and connecting it to the interface with js
[14:57] <urbanape> jblount: k
[15:00] <jblount> Desktop+ MEETING BEGINS
[15:00] <jblount> Hi everyone, you know the drill:
[15:00] <jblount> me
[15:00] <aquarius> me
[15:00] <CardinalFang> me
[15:01] <urbanape> me
[15:01] <dobey> me
[15:03] <jblount> DONE: Got most of the public urls web ui working, got frustrated at js
[15:03] <jblount> TODO: Get urbanape to help me connecting the dots between the rest API and the js, work on ui bugs in this list: https://bugs.edge.launchpad.net/ubuntuone/+bugs?field.searchtext=&orderby=-importance&assignee_option=any&field.assignee=&field.bug_reporter=&field.bug_supervisor=&field.bug_commenter=&field.subscriber=&field.omit_dupes.used=&field.omit_dupes=on&field.has_patch.used=&field.has_cve.used=&field.tag=gd&field.tags_combinator=ANY&search=Search
[15:03] <jblount> aquarius: YTMND
[15:03] <teknico> me
[15:03] <jblount> BLOCKED: Nope
[15:03] <aquarius> ⚀ DONE: review day; add music store downloads table to database
[15:03] <aquarius> ⚁ TODO: continue to spec file delivery downloader daemon; have music library page send message to downloader daemon and row to database; make workitems of outstanding todo items; make tomboy first-sync experience nicer
[15:03] <aquarius> ⚂ BLOCKED:
[15:03] <aquarius> CardinalFang, you're up!
[15:04] <CardinalFang> DONE: prpopsed fix for #499595.  tried to get attention on karmic desktopcouch package.
[15:04] <CardinalFang> TODO: get karmic desktopcouch updated today.  make bugs and blueprints.
[15:04] <CardinalFang> BLOCKED: None
[15:04] <CardinalFang> urbanape, thaw your fingers.
[15:04] <urbanape> DONE: Cleared up problem with server-side filtering by doing filtering client-side. Ran into other unexpected events happening.
[15:04] <urbanape> TODO: Track them down, clean them up. Bindwood most likely won't make it in as a package today.
[15:04] <urbanape> BLOCK: None.
[15:04] <urbanape> dobey: yo
[15:04] <dobey> ☺ DONE: Releases for Lucid A2
[15:04] <dobey> ☹ TODO: More new UI code and tests
[15:04] <dobey> ☹ BLCK: None.
[15:06] <jblount> teknico: rocknroll :)
[15:06] <teknico> DONE: planned contacts work with statik, booked flight for the February phone sync meeting in London, more setting up a development environment in a virtual machine, more fighting for net access
[15:06] <teknico> TODO: propose and land two branches for configuring funambol for sending sms messages (#418048), start implementing the mobile sync REST API for client app (#504689)
[15:06] <teknico> BLOCK: none
[15:06] <teknico> next: ?
[15:07] <jblount> I think EOM for now, no one else has jumped in (yet).
[15:08]  * rtgz is setting up ipv6 tunnel to attack ubuntuone software from unexpected positions
[15:08] <statik> CardinalFang, I saw that two debian developers are working on packaging desktopcouch (they posted to the erlang packaging list).
[15:09] <CardinalFang> rawk!
[15:09] <CardinalFang> Though, I'm worried if it takes two people.
[15:09] <statik> i think it's a co-maintaining packages kind of thing
[15:09] <CardinalFang> I'm worried if it takes more than a tents of one person.
[15:09] <CardinalFang> "tenth"
[15:10] <jblount> CardinalFang: A tense person? :)
[15:10] <dobey> shouldn't they just backport the ubuntu packages? :)
[15:10] <CardinalFang> !!
[15:10] <jblount> urbanape: When you get a chance to look at this, let me know. It's half done, but I think looking at the diff you should see where I'm at (and probably where I've gone wrong) lp:~jblount/ubuntuone-servers/public-files-webui
[15:11] <statik> dobey: i think they are starting from the ubuntu packages, yes.
[15:12] <urbanape> jblount: grabbing it now
[15:14] <dobey> i guess there might be differences with our erlang/couchdb packages, that aren't backported to debian, which would be harder
[15:21] <aquarius> statik, yep: ITP at http://lists.debian.org/debian-devel/2010/01/msg00183.html
[15:22] <statik> dobey: i've tried to port the changes in couchdb package to debian, and gotten rejected for no good reason
[15:24] <jblount> CardinalFang: That fruit you brought with you has me on a tangerine kick. Thanks (I think).
[15:26] <dobey> statik: maybe because you weren't a DD? :)
[15:27] <statik> dobey: I joined the erlang packaging team in response to a call for volunteers to maintain couchdb in debian, ported a bunch of other changes which were accepted, and this other dude who is also not a DD objects to the split between couchdb and couchdb-bin, for no good reason. It's depressing really :(
[15:28] <dobey> statik: oh
[15:28] <dobey> statik: well, isn't it debian policy that separate binaries generally get their own package? :)
[15:30] <CardinalFang> jblount, welcome.  I have a few more to share tomorrow.  They're losing moisture, though.
[15:30] <statik> i dunno - there are plenty of other packages that work this way, and the person who recommended splitting the package that way was slangasek, the ubuntu RM who is also a longtime DD. I argued as best I could, but eventually dropped it because I don't want to start a fight over there.
[15:31] <statik> __lucio__, want to join #pycon? I want to ask about how to get our ubuntuone/desktopcouch sprint on the official list so that people who are registering for pycon now can sign up for it
[15:40] <urbanape> anybody having trouble getting their local dev env to work properly?
[15:40] <urbanape> I'm getting can't connect to host after make start
[15:41] <urbanape> freshly updated sourcedeps, to.
[15:41] <urbanape> too
[15:43] <jblount> urbanape: Is apt happy? I had a problem with RabbitMQ that statik fixed and it started with cranky Rabbit and ended with me not having u-s-d installed.
[15:43] <urbanape> hm.
[15:43] <urbanape> u-s-d?
[15:44] <jblount> ubuntuone-sourcedeps? I forget the package name.
[15:44] <jblount> Grr, not soucedeps. The one meta package to rule them all.
[15:45] <urbanape> I think I'm missing something, thne.
[15:46] <urbanape> then. I don't see anything meta looking when I apt-cahce search ubuntuone
[15:47] <statik> jblount: ubuntuone-developer-dependencies ?
[15:48] <jblount> statik: You win!
[15:48] <jblount> urbanape: ^^?
[15:48] <urbanape> yeah, don't see that at all.
[15:48] <urbanape> is that in another PPA?
[15:48] <statik> it's in the hackers ppa
[15:49] <statik> it's just the metapackage that drags in postgres and rabbit and all that junk
[15:50] <urbanape> launchpad says that archive has been disabled.
[15:51] <urbanape> https://edge.launchpad.net/~ubuntuone-hackers/+archive/ppa ?
[15:51] <dobey> urbanape: ~ubuntuone/+archive/hackers
[15:53] <urbanape> normal? W: Failed to fetch http://ppa.launchpad.net/ubuntuone/hackers/ubuntu/dists/karmic/main/binary-i386/Packages.gz  404  Not Found
[15:53] <dobey> urbanape: you need your key/token
[15:54] <dobey> although i don't know why it would give a 404
[15:54] <adiroiban> rodrigo_: can you hint me to branch used for maintainig the couchdb-glib package? I would like to have a package with the latest code in a PPA
[15:54] <rodrigo_> adiroiban, lp:couchdb-glib
[15:54] <rodrigo_> ah, the package, sorry
[15:54] <adiroiban> hm... I remember you just removed
[15:54] <adiroiban> ah. ok
[15:55] <rodrigo_> http://bazaar.launchpad.net/~ubuntu-desktop/couchdb-glib/ubuntu/
[15:56] <rodrigo_> I submitted on friday a branch to package the very latest stable one
[15:56] <urbanape> dobey: did that, still get a 404
[15:56] <rodrigo_> bzr+ssh://bazaar.launchpad.net/~rodrigo-moya/couchdb-glib/0-5-3-release/
[15:56] <rodrigo_> adiroiban, ^^
[15:56] <rodrigo_> adiroiban, the unstable stuff is not yet packaged, so great you're doing it
[15:57] <adiroiban> thanks. can you take a look at bug 505623 ?
[15:58] <adiroiban> for me, while using vala, is a blocker
[15:58] <rodrigo_> adiroiban, yes, renaming them to ...Session
[15:58] <rodrigo_> I'll do it as soon as I finish one thing I'm doing, so you should have it this evening as latest
[15:58] <rodrigo_> adiroiban, check the git master if I'm not around
[15:58] <adiroiban> rodrigo_: no hurry.
[15:58] <adiroiban> i just that before having the packing in PPA
[15:59] <adiroiban> I would like to see that problem fixed
[15:59] <rodrigo_> well, I don't want to block you more
[15:59] <adiroiban> I fixed the problem on my local system
[15:59] <rodrigo_> adiroiban, did you see my comment on your debug messages branch?
[15:59] <adiroiban> and since I'm using bleeding edge code
[15:59] <dobey> urbanape: oh, yes. it's a private ppa
[15:59] <adiroiban> I don't expect to have a general public release soon
[16:00] <adiroiban> rodrigo_: yep. Just submited a fix
[16:00] <rodrigo_> cool
[16:00] <dobey> urbanape: http://launchpad.net/people/+me/+archivesubscriptions
[16:00] <adiroiban> you should receive it via email
[16:00] <dobey> urbanape: the normal lp ppa page looks like it gives the wrong info for private PPAs
[16:01] <rodrigo_> adiroiban, yeah, got it
[16:02] <adiroiban> rodrigo_: are we going to see desktopcouch-glib in Lucid?
[16:08] <urbanape> dobey: gotcha. okay, seems to be cleared up now.
[16:08] <dobey> urbanape: bug #404279 is what you hit :)
[16:08] <urbanape> statik: it's still called ubunet-developer-dependencies, btw
[16:09] <statik> ah right
[16:13] <rodrigo_> adiroiban, yes, I hope so
[16:14] <rodrigo_> adiroiban, I just need to make sure evo-couchdb works as in karmic with it
[16:14] <rodrigo_> adiroiban, so yeah, 90% sure
[16:14] <adiroiban> and will be build from the same source pacakge, or do you plan to split it?
[16:14] <adiroiban> I can help with testing
[16:16] <urbanape> jblount: k, I'm up and running
[16:22] <rodrigo_> adiroiban, from the same source package
[16:22] <adiroiban> rodrigo_: ok. am trying to add the required changes for packaging
[16:27] <rodrigo_> adiroiban, ok
[16:57] <rodrigo_> adiroiban, renaming to ..Session done
[16:59] <adiroiban> rodrigo_: thanks. I will build the packages using a PPA and test evolution. this should keep me busy for awhile
[16:59] <rodrigo_> ok, cool
[16:59] <rodrigo_> let me know of any problems you find
[17:22] <adiroiban> rodrigo_: sure. any idea why couchdb-glib package is at compat 5 ?
[17:22] <rodrigo_> adiroiban, no
[17:24] <adiroiban> rodrigo_: I will set it to 7 to avoid passing --sourcedir to dh_install
[17:24] <rodrigo_> ok
[17:34] <rodrigo_> bbl
[18:35] <mandel> aquarius ping
[18:35] <aquarius> mandel, pong
[18:38] <mandel> aquariues, quick question about XDG, if an app wants to create views, should it just create the appropriate json docs there, right? Does the app have to take care of the different version of the doc? I assume that the app is the only one allowed to modify those design docs right?
[18:39] <mandel> aquarius, ups typo, that was meant for you
[18:39] <aquarius> you shouldn't need different versions of the doc
[18:40] <aquarius> you just overwrite your existing map.js with a new one, if you want
[18:40] <mandel> aquarius, what if the we are talking about a plugin app in which plugins add new views?
[18:41] <aquarius> no reason why the plugins shouldn't write views into the same folder
[18:41] <aquarius> if you're worried about plugin views and your app views having the same names, require plugin views to be in a design doc named after the plugin
[18:42] <mandel> so, if I have to views I do plugin1.js and plugin2.js ? (I have some problem getting my head around this)
[18:42] <aquarius> no. you have plugin1/map.js and plugin2/map.js
[18:43] <aquarius> or, plugin1/viewname/map.js if you prefer each plugin to have its own design doc (which I'd recommend)
[18:43] <aquarius> er, got the path wrong :)
[18:43] <aquarius> your app should create $XDG_DATA_DIR/desktop-couch/databases/YOUR_DB_NAME/_design/DESIGN_DOC_NAME/views/VIEW_NAME/map.js
[18:44] <aquarius> I'd then suggest that you tell plugins to create $XDG_DATA_DIR/desktop-couch/databases/YOUR_DB_NAME/_design/PLUGIN_NAME/views/VIEW_NAME/map.js
[18:44] <aquarius> so that each plugin gets its own design doc, and can create as many views as it wants in that design doc
[18:44] <mandel> ok, thanks for the help I get it now... I understood it completely wrong
[18:44] <aquarius> see /usr/share/doc/python-desktopcouch/README on Ubuntu
[18:45] <aquarius> or http://www.freedesktop.org/wiki/Specifications/desktopcouch/Documentation/DesignDocsFilesystem
[18:46] <mandel> I was reading the freedesktop one but I was getting a bit confused. I got everything now :D
[18:47] <mandel> I i'd like a feature method to be added (bulk put) should I commit a patch or just post a question?
[18:50] <aquarius> submit a patch
[18:50] <aquarius> unless you're not sure how to design it at all, in which case we should get into a discussion about it :)
[18:50] <mandel> k
[18:52] <mandel> aquarius, I've got an idea, the only things is that pythion-couchdb uses a single request but does not consider a bulk update a single transaction. I'm not sure that is what we/I want
[18:52] <aquarius> yeah.
[18:53] <aquarius> urbanape ran into this with bindwood, too
[18:53] <aquarius> might wanna ping him about it?
[18:54] <mandel> sure
[18:54] <mandel> ping urbanape
[19:00] <urbanape> yeah, ended up doing individual requests.
[19:00] <urbanape> the bulkSave didn't do what I wanted.
[19:03] <mandel> urbanape, what where you looking for?
[19:04] <urbanape> I was hoping that the documents saved in a bulkSave would show up in a single sequence in the _changes feed.
[19:05] <urbanape> also, Lex is home for the afternoon, so I'll be popping in and out doing wrangling
[19:07] <mandel> urbanape, I was looking for something similar plus the idea of single transactions.. well I take a look of how to get both, cheers
[19:30] <urbanape> mandel: it turns out my model allowed me to deal with disparate transactions just fine.
[19:37] <mandel> urbanape, I just asked in the couchdb mailing list about transactions  list and the current implementation is not there. I guess that it can be fake if needed, I'll just send a patch that uses the bulk update with the CouchDataBase class
[19:57] <dobey> verterok: ping
[20:01] <verterok> dobey: pong
[20:01] <dobey> verterok: do you see anything odd in http://pastebin.ubuntu.com/355144/ ?
[20:02] <verterok> dobey: looking
[20:02] <dobey> verterok: i keep getting AlreadyCalledError and timeouts trying to run those tests :/
[20:03] <verterok> all the tests?
[20:03] <verterok> ups, it's one :)
[20:04] <dobey> verterok: that one test, yes :)
[20:07] <verterok> dobey: there is and weird error at the top: exceptions.AssertionError: reactor already installed
[20:08] <verterok> dobey: looks like it's executing: bin/ubuntuone-login
[20:08] <dobey> verterok: oh, right. i tweaked ubuntuone-login to do that in __main__ only, instead of where it is currently
[20:08] <dobey> maybe i should just commit/push
[20:09] <dobey> but i hate committing half-assed branches that don't work :)
[20:09] <verterok> dobey: nm, why are you doing execfile of ubuntuone-login?
[20:10] <dobey> verterok: trying to load it as code so we can run tests on the bits in it
[20:10] <dobey> verterok: it just loads it into a namespace, as if it were imported
[20:11] <dobey> but we can't import because of the - in filename
[20:11] <PerJ_> Matt: How do I make Canonical give me back my 50GB account - it vanished the 23. of december
[20:11] <verterok> dobey: just move it into a module, and make ubuntuone-login, just a caller of a function ;)
[20:11] <verterok> webm0nk3y: ^
[20:11] <dobey> no
[20:12] <dobey> having a script that just does import foo.main, main.main() is lame :)
[20:12] <webm0nk3y> PerJ_, let me check what may have happened to your account
[20:12] <verterok> dobey: it's just a script
[20:12] <dobey> script logic belongs in scripts, not modules
[20:12] <PerJ_> web: you can find me as PerJensen
[20:12] <verterok> dobey: put the script logic in ubuntuone-login, and the module logic in a module
[20:13] <PerJ_> a withdrawal on the 19. was not executed
[20:13] <dobey> verterok: that's how it is now
[20:13] <verterok> dobey: and why you need to import it?
[20:13] <verterok> :()
[20:13] <verterok> :)
[20:13] <PerJ_> web: so the payment for january has not been made
[20:13] <dobey> verterok: to run tests on the code in the script
[20:13] <verterok> dobey: ok, if I remove that, I can run the test?
[20:14] <mattgriffin> PerJ_: just got the email from you... sorry. a bit behind on responding to customer messages.
[20:14] <PerJ_> web: no emails from Canonical as to why the withrawal from my creditcard was not performed
[20:14] <dobey> verterok: let me see what all i changed. i think i had to move a couple more calls around too
[20:14] <verterok> dobey: also, constants in a script? ;)
[20:14] <PerJ_> matt: yes you are *S*, the one from 23. has not been answered yet
[20:14] <issyl0> Hi there.
[20:14]  * issyl0 is wondering how Ubuntu One works.  Is it like an online place to store your files that can obviously be accessed from anywhere by logging in with your lp id ?
[20:15] <issyl0> mattgriffin ?  :)
[20:15] <dobey> verterok: yeah, i had to change some more so we can use FakeLogin() in the test
[20:15] <mattgriffin> issyl0: hi. that's part of it. the key think is synchronization.
[20:16] <mattgriffin> issyl0: the desktop software syncs your files and data with your personal cloud storage and with your other computers setup with your account.
[20:16] <issyl0> mattgriffin: oh now that is brilliant.  :D
[20:16] <verterok> dobey: ok. I don't see anything "wrong" in the tests, but would like to execute it. could you pastebin the full traceback?
[20:17] <dobey> sure
[20:17]  * issyl0 has a laptop too and it's certainly better than carrying around usb sticks/emailing files to yourself
[20:17] <dobey> grr, firefox didn't restart when i clicked the restart button
[20:17]  * issyl0 will try that tonight with some of her homework :)
[20:17] <webm0nk3y> PerJ_, I am talking to you in a private channel btw
[20:21] <dobey> verterok: http://pastebin.ubuntu.com/355158/
[20:22] <verterok> dobey: found it!
[20:23] <verterok> dobey: gimme 1' to double check it
[20:26] <dobey> verterok: yay!
[20:27] <dobey> verterok: i pushed to lp:~dobey/ubuntuone-client/test-login btw
[20:27] <verterok> dobey: looks like tearDown it's being called twice, not actually tearDown, but cleanup_signal_receivers.
[20:27] <verterok> dobey: the callback are beign called more than once p/signal_receiver
[20:27] <verterok> *callbacks
[20:28]  * verterok pulls
[20:29] <dobey> right, i am not sure why
[20:30] <dobey> i think maybe we need to move the syncdaemon\-specific bits of DBusTwistedTestCase to another test case, and have DBusTwistedTestcase just do the twisted/dbus bits that are necessary there
[20:32] <verterok> dobey: let me try somethign before doing that :)
[20:35] <dobey> sure
[20:40] <verterok> dobey: is this using any gtk stuff? it's quite weird :(
[20:42] <dobey> verterok: the script uses gtk+ to create an error dialog for the errors
[20:43] <dobey> verterok: i was also having some weirdness with overriding the dbus signal handler methods in the test... it's weird
[20:44] <verterok> dobey: I think the u1login signal receivers might be the cause of this stuff going crazy
[20:46] <verterok> dobey: If I remove all the signal cleaup in the tearDown of your test it works
[20:47] <dobey> hmm
[20:49] <urbanape> aquarius: you about?
[20:49] <urbanape> or thisfred or CardinalFang
[20:50] <thisfred> I am
[20:50] <urbanape> can I configure my desktopcouch to exempt additional databases from being replicated?
[20:50] <verterok> dobey: I "think" I found the problem :)
[20:50] <urbanape> I'm testing Bindwood, and I'd like to use a local db that will not not not get replicated to u1.
[20:51] <thisfred> urbanape: you can, manually
[20:51] <CardinalFang> yep.
[20:51] <thisfred> there's a database for the pairings, which also allows you to specifically exclude dbs
[20:51] <CardinalFang> It's a feature that isn't well tested and not exposed anywhere.
[20:51] <thisfred> CardinalFang might know exactly how, or I could dig in the code
[20:52] <verterok> dobey: LoginMain.__init__ register the signal receivers/handlers and then you override those methods, and register new handlers
[20:52] <dobey> verterok: yes, because of the weirdness i mentioned earlier
[20:53] <dobey> verterok: it seems as though the overridden methods don't get called, unless i register additional signal handlers :(
[20:53] <verterok> dobey: so, the first handlers are still registered, as the method is still there (the reference to reference)
[20:53] <dobey> verterok: it runs the original method code
[20:53] <verterok> dobey: right, because it get registered at __init__ time
[20:53] <verterok> dobey: python methods are objects ;)
[20:53] <dobey> right
[20:54] <verterok> dobey: doing self.foo.a_method = my_method just updates self.foo.__dict__
[20:54] <verterok> dobey: but all references to the old method are still alive
[20:54] <dobey> hmm
[20:55] <verterok> dobey: yes, and dbus keeps a reference to the old one :/
[20:55] <CardinalFang> urbanape, thisfred, in the pairing record(s), it's something like "excluded_names".  I think it should be a list of db names.
[20:56] <dobey> hrmm
[20:56] <thisfred> either that, or sacrifice a goat
[20:57] <dobey> am still getting errors after moving that code
[20:57] <dobey> AlreadyCalledError even
[20:58] <dobey> :(
[20:59] <urbanape> CardinalFang: are the pairing records in management?
[21:02] <CardinalFang> urbanape, Yes.  Management DB, with record_type  http://www.freedesktop.org/wiki/Specifications/desktopcouch/paired_server  .
[21:05] <urbanape> okay, so just create a new field called 'excluded_names', make it a list of db names?
[21:07] <jamalta> hi there, i have a question.. mostly asking for opinions
[21:07] <verterok> dobey: moving what code to where? :)
[21:07] <mattgriffin> jamalta: how can we help you?
[21:07] <urbanape> cool. So, I think the weirdness I was seeing was the test db from U1 being replicated back.
[21:07] <dobey> verterok: moved the signal connections out of __init__
[21:07] <urbanape> Hey, wow. My code works.
[21:07] <jamalta> i've been using keepass for password management for a while now.. but i'm wondering if it would be a good idea for this to be based on desktopcouch
[21:08] <jamalta> obviously, each password is encrypted
[21:08] <urbanape> Bindwood 1.0 is gonna be pretty cool, I think.
[21:08] <jamalta> but it would be awesome to be able to integrate that into the application instead of having a third-party application
[21:08] <urbanape> k, I have to go for a bit. jblount: I should have something for you to look at tomorrow.
[21:08] <jamalta> so firefox could have a plugin that reads from desktopcouch for it's passwords.. and any other application that needs passwords could do that as well
[21:08] <jamalta> i just wanted to see what some smarter people thought about that :)
[21:09] <mattgriffin> jamalta: this is one area that was discussed at UDSLucid
[21:09] <jamalta> mattgriffin: oh really? i must have completely missed that
[21:09] <mattgriffin> jamalta: the ubuntu one team is trying to do something about it - https://blueprints.launchpad.net/ubuntu/+spec/lucid-ubuntu-one-keyring-sync
[21:10] <mattgriffin> jamalta: is that what you mean?
[21:10] <dobey> verterok: i think i need to split up the TestCase code to not do syncdaemon stuff there, but in a separate SDDBusTwistedTestCase or something
[21:10] <jamalta> mattgriffin: oh sexy ;) thanks! i guess i just have to sit tight and hang on
[21:10] <jamalta> mattgriffin: yeah, that's even better than what i was thinking
[21:10] <mattgriffin> jamalta: :)
[21:10] <verterok> dobey: you can try a quit hack, by stripping down DBusTwistedTestCase, but I'm not sure that's the problem
[21:11] <jamalta> it makes more sense to do it through the keyring...
[21:11] <jamalta> mattgriffin: thanks!
[21:11] <mattgriffin> jamalta: are there any other apps that you use that could benefit from this?
[21:12] <jamalta> mattgriffin: well, somewhat.. it would be good if there was a way to figure out the data source (not sure if gnome keyring is already able to do this)
[21:13] <jamalta> for example, firefox should know that i want my identi.ca password when i'm at identi.ca
[21:13] <jamalta> and then, gwibber, should also be able to pull my identi.ca password from the keyring..
[21:13] <mattgriffin> +1
[21:13] <jamalta> or if i open evolution, and create a gmail account.. it should know that i have a google password in the keyring and try to pull that
[21:14] <verterok> dobey: ok, I think I found it (the third is..)
[21:14] <verterok> :)
[21:15] <verterok> dobey: looks like it's a bug in testcase.py itself :)
[21:15] <jamalta> mattgriffin: well thanks for pointing me to that.. i'm glad i came and asked
[21:15] <verterok> dobey: looks at the *** lines in: http://pastebin.ubuntu.com/355181/
[21:16] <verterok> dobey: seems like a weird scope issue
[21:16] <mattgriffin> jamalta: cool. come back if you think of more ideas or good applications for Ubuntu One and the underlying technologies. we're always eager to chat.
[21:16] <jamalta> mattgriffin: thanks, will do :)
[21:20] <verterok> dobey: or a bug in dbus-python itself...who knows! :)
[21:23] <verterok> dobey: fix http://pastebin.ubuntu.com/355184/
[21:27] <verterok> dobey: so isn't a scoping issue, it's some dbus weirdness
[21:58] <dobey> verterok: ah, workarounds eh :)
[21:59] <verterok> dobey: :)
[21:59] <dobey> verterok: i noticed in the test case itself, in tearDown, it was doing d.addBoth(self.tearDown)
[21:59] <dobey> verterok: which probably isn't very helpful either :)
[22:00] <verterok> dobey: that's ok, tearDown should always finish
[22:00] <dobey> oh
[22:00] <dobey> no, it's calling _tearDown
[22:00] <verterok> dobey: if not, it might break the next test
[22:00] <dobey> i missed the _
[22:00] <verterok> yes :)
[22:01] <verterok> dobey: I think the deferred it's called several time because there are multiple matches of the signal receiver "name", I'll digg in dbus-python as soon I get some spare time to check if there is a bug regarding this
[22:02] <verterok> dobey: I know the testcase it's using private dbus-python stuff, but's the only way I found to actually make it work :(
[22:02] <dobey> well, now i'm getting unclean reactor, and it's mentioning syncdaemon.event_queue.MyReader
[22:03] <verterok> hmm, let me check your tearDown again
[22:03] <dobey> verterok: i copied the code from test_dbus.py
[22:04] <verterok> dobey: ok, but something is hanging around that makes tearDown not finish OK
[22:04] <dobey> verterok: maybe because i'm not actually doing anything with SD, but the testcase somehow expects me to?
[22:05] <verterok> dobey: no
[22:05] <dobey> (since the testcase was written to test the SD dbus bits)
[22:05] <dobey> hmm
[22:06] <verterok> dobey: the test is failing with a TimeoutError
[22:09] <dobey> fun, looks like LP is not wanting me to push my changes now
[22:10] <verterok> dobey: change the  addCallback calls in the tearDown to addBoth
[22:11] <dobey> verterok: i did
[22:11] <dobey> but LP is having a fit right now :(
[22:12] <verterok> uh :(
[22:12] <dobey> doh
[22:13] <dobey> yay, it works now
[22:13] <dobey> but i think doing all the SD stuff is still a bit heavy for this :)
[22:14] <dobey> verterok: pushed my branch with working test
[22:14] <dobey> verterok: if you could look over it real quick again, and let me know if i did anything dumb/crazy/etc... that'd be awesome
[22:14] <verterok> dobey: sure, gimme 1'
[22:14] <dobey> verterok: sure, no problem
[22:17] <verterok> dobey: looks ok, but I have a recommendation for the signal stuff and avoid replication ;)
[22:18] <dobey> verterok: i am open to suggestions :)
[22:18] <verterok> dobey: instead of moving the add_signal_receiver call to main, move them to a method, eg: _register_dbus_signal_handlers or whatever
[22:19] <verterok> dobey: and in the test create the LoginMain instance, override the methods and call _register_dbus_signal_handlers
[22:19] <dobey> ok
[22:19] <dobey> yeah, that makes sense
[22:19] <dobey> thanks :)
[22:19] <verterok> np
[22:19] <verterok> :)
[22:35] <dobey> later
[22:35]  * CardinalFang Zzzzz.
[22:40] <statik> urbanape, can your new-folder-inline-edit branch be landed?
[22:40] <statik> CardinalFang, sweet dreams. I sent one of your old branches to pqm, and your notes revision branch is sitting in the approved queue. want me to submit it?
[22:49] <osseta> hey does Ubuntu One work behind authenticated proxies? I'm trying to set it up and haveing a bitch of a time.
[22:50] <osseta> I have a launchpad account setup and can log into one.ubuntu.com and see my account and empty file list
[22:50] <osseta> but the client gnome panel doesn't seem to do anything and gives no errors, warning, hints
[22:55] <verterok> osseta: there is no proxy support in the client yet
[22:56] <osseta> eventually I guess that was the case, I tried it via proxychains (which I use for other apps that don't support proxies) but that didn;t work either
[22:57] <osseta> does the client use HTTP only, or are there other protocols invloved?
[22:59] <statik> osseta: it uses a custom protocol, written in google protocol buffers and published in the ubuntuone-storage-protocol package. we run over port 443 masquerading as https, so proxies should not interfere. I think the main problem with proxy support at the moment is that it's busted in the core python stdlib(or it was until very recently)
[23:00] <osseta> are, yes, I remember protocol-buffers being mentioned on the FLOSS podcast now
[23:08] <MrKeuner> am I not supposed to run two computers at the same time for healthy synchronization using UbuntuOne?
[23:09] <statik> MrKeuner, running 2 or 5 computers at the same time should be fine
[23:10] <MrKeuner> I just updated ubuntuOne files and second system's ubuntuone tray icon tells me that my files are up to date. Bu they are not
[23:13] <MrKeuner> Oh, OK there was a conflict
[23:13] <MrKeuner> it did not let me know though
[23:13] <MrKeuner> how can I resolve conflicts?