[06:39] <hackel> On my netbook, my Ubuntu One directory is 7M and the client says it is up to date.  On the website it shows I am using 3.6M.  On my laptop, it is 4.7M and also says it is up to date!  Where do I even begin to figure out where things are going wrong?
[10:40] <rocky|raccoon> hello :> i uploaded a file in ubuntu one's web interface and it's showing up there, but it's not showing up in my local "ubuntu one" folder and the update applet says that my files are up to date... :/ is there something i'm missing?
[14:10]  * rtagger wonders why my ISP drops the connection right when I post something to #ubuntuone...
[14:17] <rtagger> joshuahoover: what do the checkmarks on folders and files really mean?
[14:19] <dobey> rtagger: it's supposed to mean "synchronized"
[14:19] <joshuahoover> rtagger: good question
[14:19] <joshuahoover> yeah, what dobey said
[14:20] <joshuahoover> keywords: "supposed to"
[14:20] <dobey> well
[14:20] <dobey> "directory is synchronized" doesn't mean what you probably think it means, also
[14:20] <rtagger> dobey: mmm... Even if I have no connection to u1, I get the checkmarks. I remember there were clouds with arrows some time in the beginning.
[14:21] <dobey> rtagger: yes, clouds with arrows mean it's transferring the file
[14:22] <rtagger> dobey: just created 'new file' in the U1 directory, got 'Ubuntu is updating your files', the file immediately got the checkmark, but... it is not even connected to the server :)
[14:24] <dobey> rtagger: i never said it was perfect :)
[14:25] <rtagger> okay, next question - does syncdaemon know all about server-side files before it starts downloading them?
[14:28] <dobey> well it knows their filename, and if they are newer or not
[14:29] <aquarius> dobey, clouds with arrows? orly?
[14:30] <dobey> aquarius: well, clouds with arrows if you're using !Humanity for your icon theme
[14:30] <aquarius> hm.
[14:30] <dobey> or maybe it's hard disk with arrows now
[14:30] <dobey> but either way, there are definitely arrows :)
[14:30] <aquarius> icon theme: Humanity Dark
[14:31] <dobey> aquarius: yeah i don't remember what the icons for that in humanity are exactly
[14:32] <dobey> just green/blue arrows it seems
[14:32] <aquarius> I was using U1 the other day and I thought: hey, it'd be really neat if we had "this file is currently syncing" emblems. So, since we already *do* have that, clearly it ain't working for me :P
[14:35]  * rtagger :'-( No check marks in "Human" theme...
[14:37] <dobey> aquarius: then clearly there is a bug :)
[14:38] <rtagger> dobey: no arrows on Humanity Dark, Humanity, Human. But I clearly remember seeing them some time before Karmic release
[14:39] <dobey> rtagger: ok. it's a bug, and i think there might be a bug filed already
[14:41] <dobey> thomastp: hi. so i'm not entirely sure what to do about the bdist_rpm issue. i can't find any clear documentation of what it needs/does, and whether or not sdist is supposed to put MANIFEST in the tarball... and haven't had time to try and parse the source for distutils
[14:42] <thomastp> dobey: ok, I'll see if I can look into it some more when I hve some time
[14:42] <dobey> but i'd really like to get a new tarball built/released with the fix :)
[14:47] <dobey> aquarius: you had to copy the flash plugin somewhere else? it Just Works (TM) for me...
[14:47] <dobey> aquarius: with the ubuntu packages
[14:47] <aquarius> dobey, doesn't just work for me
[14:47] <dobey> weird
[14:47] <aquarius> ah!
[14:47] <dobey> apt-get install adobe-flashplugin :)
[14:48] <aquarius> I wonder if I have a private copy of libflashplayer.so in .mozilla/* rather than the ubutnu package
[14:48] <dobey> probably
[14:48] <aquarius> although I am toying with the idea of seeing if I can do without flash. So far the only thing that's annoyed me is lack of youtube and grooveshark, and I don't mind running grooveshark in firefox since it's a whole window app anyway
[14:48] <dobey> well
[14:48] <aquarius> and that will all be fixed once I have all my music in U1 anyway, which is the project for this week ;)
[14:48] <dobey> you can youtube in totem
[14:49] <dobey> although it would be nice if totem had ui for comments/etc... on youtube too
[14:49] <aquarius> yeah, I mean embedded youtube widgets in pages, not browsing it myself :P
[14:50] <dobey> i wonder if there is some easy way to muck about with the DOM of a page in chrome
[14:50] <dobey> you could fix it to behave more like mobile
[14:50] <dobey> so that you can click on embedded videos and have them play in totem
[14:51] <aquarius> I could do that if I knew how to make an extension fire up a desktop app, which I do not know how to do ;(
[14:51] <dobey> also, the multimedia keys handling is kind of crappy still
[14:51] <aquarius> rly? work perfectly for me, multimedia keys
[14:52] <dobey> aquarius: register a URL handler and make a custom youtube: link and open it :)
[14:52] <aquarius> yeah, but I don't know how to teach chromium about new url handlers
[14:52] <dobey> aquarius: well, except i can't pause rhythmbox after i've opened totem
[14:52] <aquarius> I need that to handle feed: urls too so I can pass them to liferea
[14:52] <dobey> i have to use a mouse to do so
[14:52] <aquarius> oh. heh.
[14:52] <aquarius> I haven't tried that :)
[14:52] <aquarius> also, I need to sync rhythmbox's db into desktopcouch, which will be entertaining
[14:53] <dobey> *shudder*
[14:53] <dobey> eventually, i will just have a storage server in my house that serves up music/videos with upnp :)
[14:54] <aquarius> I have that, but I don't like it because when I go away here's no music on my laptop
[14:54] <aquarius> 'cept it's not upnp :)
[14:55] <dobey> well, that's what tunnelling is for :)
[14:56] <dobey> or vpn
[14:56] <dobey> i might set it up a vpn too eventually
[14:56] <dobey> but i have about 4343095839403 projects i need to do right now :)
[15:01] <vds> me me me! :)
[15:02] <dobey> vds: i would LOVE your help on these projects! feel free to come on over and help out :)
[15:02] <aquarius> me
[15:02] <aquarius> oh, hang on
[15:02] <dobey> this is why i love the ubuntuone team. you guys are so helpful!
[15:02] <aquarius> MEETING BEGINS: Desktop+ standup meeting. Say "me" if you're in it.
[15:03] <vds> dobey: sorry but that me was for a different purpose :)
[15:03] <vds> me
[15:03] <aquarius> vds, dobey, teknico, urbanape, cardinalfang, jblount, statik, Chipaca
[15:03] <aquarius> me
[15:03] <teknico> me
[15:03] <Chipaca> me
[15:04] <dobey> me
[15:04] <aquarius> er, rodrigo_1 too :)
[15:04] <rodrigo_1> me
[15:05] <teknico> chad is not here, and jblount is away
[15:05] <aquarius> then go, vds
[15:05] <vds> DONE: worked ont the two branches to port funambol v8 in sourcedeps with teknico, discussed about scaling, monitoring and testing funambol with statik, thinking about funambol/couchdb timestamps, discussed several issues with the funambol support
[15:05] <vds> TODO: land the two branches, continue the discussion about timestamps, go trough the funambol docs about scaling
[15:05] <vds> BLOCKED: nope
[15:05] <teknico> urbanape and statik stikk missing
[15:06] <teknico> still, even
[15:06] <vds> aquarius: please
[15:06] <aquarius> ⚀ DONE: more work on desktopcouch developer docs; discussions about music store
[15:06] <aquarius> ⚁ TODO: look at oauth-enabling twisted; make tomboy first-sync experience nicer; continue work on desktopcouch developer docs; write up things learned at UDS/sprint; work with rodrigo on Music Store
[15:06] <aquarius> ⚂ BLOCKED:
[15:06] <aquarius> teknico in da hizouse
[15:06] <teknico> DONE: landed the branch to fix Funambol config and tests for v. 8.0 (#403435), continued updating the Funambol code to v. 8.0 in our code (#403435), some reviews
[15:06] <teknico> TODO: more reviews, finish updating the Funambol code to v. 8.0 in our code (#403435)
[15:06] <teknico> BLOCK: none
[15:06] <teknico> next: Chipaca
[15:06] <Chipaca> DONE: went to the foundations+ sprint, talked with the guys there, talked about sso, had a bbq dinner. TODO: planning, prodding, pleading. That means you will be getting a ping from me :). NEXT: dobey
[15:06] <teknico> Chipaca, you seem to enjoy your new job ;-)
[15:07] <Chipaca> teknico: not yet, but soon
[15:07] <dobey> ☺ DONE: Work estimation, Triage, MANIFEST dist debugging
[15:07] <dobey> ☹ TODO: Backport fixes to stable-1-0, Prepare stable releases/SRUs
[15:07] <dobey> ☹ BLCK: None.
[15:07] <dobey> rodrigo_: take your protein pills, and put your helmet on :)
[15:07] <rodrigo_> • DONE: More music store widget work. Triaged some evo-couchdb bugs
[15:07] <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 couchdb-glib documentation to xdg page (http://www.freedesktop.org/wiki/Specifications/desktopcouch/Documentation/SimpleGuide)
[15:07] <rodrigo_> • BLOCKED: no
[15:07] <rodrigo_> nobody next, right?
[15:07] <dobey> si
[15:08] <dobey> hrmm
[15:10] <dobey> not sure oauth token per app really makes sense
[15:11] <aquarius> dobey, the problem with the current method is that you can't revoke one app
[15:12] <dobey> aquarius: yes, but we were talking of doing 2-legged, so getting a token is going to be a seamless operation for all our 'official' apps anyway
[15:13] <aquarius> mm, true
[15:13] <aquarius> woah, u1 is going mad and updating loads of my files. I wonder what it's doing?
[15:14] <dobey> no idea
[15:14] <dobey> syncing your files? :)
[15:14] <aquarius> I haven't changed anything, though :)
[15:15] <mandel> aquarius: hello, wanted to talk?
[15:15] <aquarius> mandel, teknico; teknico, mandel.
[15:16] <mandel> teknico: hello
[15:17] <teknico> mandel, hi :-)
[15:18] <teknico> mandel, I was reviewing your desktopcouch/contacts-wrapper branch, wanna field some questions? :-)
[15:18] <mandel> teknico: sure
[15:18] <teknico> mandel, very nice work, by the way
[15:18] <mandel> teknico: thanks!
[15:19] <mandel> teknico: I want to improve it more, I do not like the coupling between record and contact, but I'll do that later on
[15:19] <teknico> mandel, fyi, I'm a KDE user, long time python programmer, and in dire need of a decent contacts program myself :-)
[15:20] <mandel> teknico: hehe we are plenty then :P
[15:20] <dobey> a decent contacts program is one you don't really ever need to use :)
[15:20] <teknico> dobey, the zen channel is over there, thanks ;-P
[15:21] <teknico> mandel, I have not had much time to get into the architecture of the code
[15:22] <teknico> but have a couple questions about how it's written
[15:22] <mandel> teknico: sure
[15:22] <mandel> teknico: shoot
[15:22] <teknico> I was wondering about the enum module
[15:23] <teknico> we'll need to add it as a dependency, unless there's something comparable in the stdlib that can be used
[15:23] <mandel> teknico: yes, I hate that too
[15:23] <teknico> named tuples would seem good enough, but they're only in Python 2.6 unfortunately
[15:23] <dobey> teknico: i pretty much never use contacts applications, except to add detailed information/new contacts
[15:24] <mandel> teknico: but python has not enum and I did not wanted to use a tuple mere int or strings
[15:24] <teknico> dobey, so you're talking about seamless integration with contacts-using applications, agreed
[15:24] <teknico> mandel, right
[15:24] <mandel> teknico: but you are right, enum is ugly...
[15:24] <teknico> mandel, it's not really bad, I was just wondering if it was necessary
[15:25] <dobey> just define a dict, or FOO_TYPE_UNO=1, FOO_TYPE_DOS=2, ...
[15:26] <mandel> teknico: I don't think is fundamental, it can be easily changed to be a tuple
[15:26] <teknico> dobey, mandel's using enums with attribute syntax, and also for checking valida values
[15:26] <teknico> mandel, ok, not a big problem anyway
[15:27] <teknico> mandel, another thing
[15:27] <teknico> can you take one underscore away from the private methods names? :-)
[15:28] <mandel> teknico: sure ;)
[15:28] <mandel> teknico: and I can actually clean the mapping process to be better
[15:28] <teknico> mandel, I mean, double underscores at the start of private methods are a nuisance with not much benefit
[15:28] <teknico> mandel, but you know that :-)
[15:29] <mandel> teknico: I know hehe I'll do so, no big deal, I get carried away by __ sometime
[15:29] <mandel> I blame working on c# at the office
[15:30] <dobey> _foo aren't private
[15:30] <dobey> __foo are :)
[15:30] <teknico> dobey, they're private enough :-P
[15:30] <dobey> except for the fact that we call them everywhere, sure :)
[15:31] <teknico> mandel, I had suspected java ;-)
[15:31] <teknico> dobey, that's our problem :-)
[15:32] <teknico> mandel, there also seem to be some redundancy and boilerplate among object classes
[15:32] <teknico> I wonder if we can unify their code some more
[15:32] <dobey> teknico: yes. yes it is :)
[15:32] <teknico> mandel, that's more of an intuition than anything else though
[15:33] <mandel> teknico: you are actually right, I have been thinking about it most of the day, I should be able to get code to be smaller and nicer
[15:33] <teknico> mandel, that would be great
[15:33] <mandel> teknico: mainly for the serialization and attributes, it should be easy
[15:33] <teknico> mandel, oh, and the big "if" in telephone_number.get_number has to go :-)
[15:33] <mandel> teknico: haha  I knew that one would come back and bite me in the ass
[15:34] <mandel> teknico: is horrible, I'll get rid of it
[15:35] <teknico> about the TestContact class, you probably meant to fill it up later, but right now it's useless
[15:35] <teknico> the tests too look like they could be shortened quite a bit
[15:35] <mandel> teknico: yes, I was but I ended focusing more in the Repository
[15:36] <mandel> teknico: about the tests, I dont know.... maybe just testing one attribute should be more than enough since it uses the same code path
[15:36] <mandel> teknico: and there is no reason to test a path more than once
[15:36] <teknico> mandel, I don't mean reducing the actual tests, no
[15:36] <mandel> teknico: what do u mean then?
[15:36] <teknico> just expressing them with less lines of very similar instructions
[15:37] <teknico> more data-based, if you catch my drift
[15:38] <teknico> like, instead of having 20 similar asserts, you could loop over a tuple with 20 items, and have one assert line inside the loop
[15:39] <teknico> and end up with 4 lines instead of 20, or something
[15:39] <teknico> it's just that I find big, uniform blocks of code unpleasant :-)
[15:40] <mandel> teknico: me too, but those block are the ones I'm talking about
[15:40] <mandel> that is actually testing that all attributes of the contact have been correctly retrieved, and it has so many that it look like big chuck of code
[15:41] <mandel> when I add mroe than one contact I use a loop
[15:41] <teknico> ok, let me whip up something quickly
[15:46] <teknico> mandel, https://pastebin.canonical.com/25036/
[15:47] <teknico> mandel, anyway, clear code is important, but don't let my whining put you off, you're doing great stuff :-)
[15:48] <mandel> teknico: don't worry, I really like people to complain about my code, get me to be better
[15:48] <mandel> teknico: as long as it is useful whining it is good, and I have been thinking of it during the day and I did not like it was not small
[15:49] <mandel> I'll have something nicer by 2night :P
[15:50] <teknico> mandel, later I'll have a go at your code at https://launchpad.net/macaco-contacts
[15:51] <mandel> teknico: that one is outdated hehe
[15:51] <teknico> where is it now?
[15:51] <mandel> better give me 'til the weeked to see the new impl and complain then
[15:51] <mandel> it is a branch in the same project
[15:52] <mandel> currently here p:~mandel/macaco/harmony
[15:52] <mandel> I went to UDS and the designers changed all my UI... and I decided to port my code and push it to desktopcouch, specially 'cause other people wanted to work with contacts
[15:52] <dobey> apachelogger: more around now?
[15:53] <teknico> mandel, it's great that you're so open to having your work hacked down to pieces :-)
[15:54] <teknico> I hope to also contribute with code, not just whining :-)
[15:54] <teknico> even if I have no desktop toolkit programming experience, mainly web
[15:54] <teknico> but Qt is a good investment anyway :-)
[15:55] <mandel> teknico: sure, I'll be contributing as much as the everyday work and girlfriend let me, and I will do a wt version, not just gtk
[15:56] <mandel> I meant qt, stupid azerty keyboards..
[15:56] <teknico> oh, it's gtk now?
[15:58] <mandel> teknico: the initial part yes 'cause it would be nice to push it to universe, but I prefer qt (I'll not get in a blame war)
[15:58] <chewit> dobey, not wanting to be a pain, but whats the latest on the web interface, you were going to ask the devs today about it
[15:58] <teknico> a blame war sounds more interesting than a flame one ;-)
[15:58] <mandel> teknico: anyway I sleep few hours a day, so both should be ready at at the same time
[15:58] <dobey> hi chewit
[15:59] <dobey> teknico: it's all your fault :)
[15:59] <dobey> chewit: they haven't pinged me back yet :)
[15:59] <teknico> dobey, which one of the many? :-)
[15:59] <dobey> chewit: but am pinging some more
[15:59] <dobey> teknico: yes :)
[15:59] <chewit> thanks :D
[16:00] <apachelogger> dobey: aye
[16:00] <apachelogger> dobey: ahoy
[16:00] <teknico> mandel, anyway, I have no stake in the race, no problems with gtk
[16:01] <teknico> mandel, so let me know when you feel like it's a good time to have a look at your code
[16:01] <dobey> apachelogger: hey!
[16:01] <mandel> teknico: will do!
[16:01] <teknico> mandel, and thanks again :-)
[16:01] <dobey> apachelogger: so are you planning on getting your kde code integrated in our tree? or what?
[16:02] <apachelogger> dobey: yes, though I'd first like to have it somewhat mature, it is considerable easier to get help from KDE developers via gitorious
[16:03] <dobey> apachelogger: ok, good. :)
[16:06] <dobey> apachelogger: we're going to also be getting rid of the applet for the most part, for Lucid... i'm not sure if you've seen any of the new u1 login/settings ui wireframes or not
[16:06] <dobey> apachelogger: hopefully deaign team will have the final bits of that UI spec done soon, though, so i can start working on it next week (tomrorow/fri are holidays for pretty much everyone in the US)
[16:09] <aquarius> mandel, did you see my comment about the copyright stuff?
[16:10] <mandel> aquarius: yes, I was going to send it, right away, but I think I did already 'cause I already push some patches, I'll do it again just in case
[16:12] <dobey> apachelogger: so your code might be better spent on the konqeror/dolphin/whateverkdeusesnow extension to share a folder and do emblems for status and such right now?
[16:13] <apachelogger> dobey: I suppose there is still some status indiciation?
[16:13] <dobey> apachelogger: i'm glad you're working on the kde stuff. but i don't want it to be stuck behind what the gnome bits will be :)
[16:13] <dobey> apachelogger: there will be the notifications we have now still, yes, and the emblems on files in nautilus.
[16:13] <apachelogger> because that is really what the applet is, the icon is nothing but a legacy representation of the noveau statusnotifieritem
[16:14] <dobey> apachelogger: we should have arbitrary folder syncing for lucid as well
[16:16] <apachelogger> oh, cool
[16:17] <dobey> so "open in file manager" won't really make much sense as an action either, since you might have several random folders synced :)
[16:18] <apachelogger> dobey: you underestimate the superior powers of kde :P
[16:18] <apachelogger> or limitations ... in a way ...
[16:19] <apachelogger> the only way to implement visible feedback is to derive a new file/folder presentation class, known as kio slave
[16:20] <apachelogger> so you'd go to ubuntuone:/ and could technically get everything listed that is to be synced
[16:20] <apachelogger> the disadvantage is that by default I dont see how one would squeeze the emblembs etc. into the regular file presentation
[16:20] <apachelogger> I need to consult with the dolphin devs on that matter
[16:21] <dobey> yeah, i'm not sure how kde does the emblem stuff
[16:21] <dobey> but if you have to write a kioslave, i presume you can specify custom icons in it
[16:22] <apachelogger> yes
[16:23] <apachelogger> I expect no problems there ... but say the user navigates in regular file view to a folder that is watched by ubuntuone, then the user would get no visual feedback of this
[16:23] <dobey> yeah, it would be nice to not have the kioslave and just do stuff in the regular view
[16:23] <dobey> custom views are generally fail :)
[16:23] <apachelogger> though IMHO this is a larger problem with dolphin and kioslaves, konqueror in KDE 3 supported on-demand kioslave switching, so if you navigated to a folder with .svn you'd get a slave for svn interaction
[16:24] <dobey> cool
[16:33] <rtagger> joshuahoover: does syncdaemon "see" the files that are put to Ubuntu One directores while it is busy uploading the other files? got 40 files synced, but 18 are still w/o server-hash, therefore not uploaded.
[16:35] <rtagger> joshuahoover: never the less, it just notified me that sync is complete. SYS_STATE_CHANGED, args:(), kw:{'state': <NonActiveConnectedSDState IDLE>} and I don't see any syncdaemon desire to continue uploading - state: IDLE
[16:45] <joshuahoover> rtagger: hmmm...i'm not sure the answer to that...the guys who know the most about the syncdaemon are all meeting together this week so they're focusing on fixing a lot of things right now
[16:47] <rtagger> joshuahoover: 'cause as a casual user I tend to believe what the notification told me. If my files are synced, then I may turn off the PC and access them remotely... Oops, not all files are synced. But I believed you!..
[16:47] <dobey> nice. my gnome-power-manager fix got uploaded to lucid
[16:47] <joshuahoover> rtagger: right, agreed :)
[16:50] <rtagger> hmm, looks like bug #432569
[17:00] <rtagger> yep, these files are simply invisible to syncdaemon until syncdaemon is restarted. Looks like syncdaemon does not rescan directories and the signals about files being placed to the directories are simply lost. Time for debugging...
[17:02] <dobey> rtagger: i wonder if maybe it's an issue with pyinotify?
[17:03] <rtagger> dobey: could be
[17:06] <dobey> hrmm, i REALLY do not want to have to hack on chromium to get it to open stuff in a new window
[17:08] <aquarius> dobey, right-click, open in new window?
[17:08] <dobey> aquarius: chromium %u
[17:09] <aquarius> oh, right, from the command line
[17:09] <dobey> opening from external places should really always open new windows. i don't know why they have it open a new tab in a current window, and have no way to change it
[17:11] <dobey> anyway, bbl :)
[17:58] <mandel> teknico: if you still want to take a look at the code today, I already made the changes :D
[18:03] <teknico> mandel, you're too fast! :-)
[18:04] <mandel> teknico: hehe well, hopefully it was fast and meets your standards, just fast is not good enough :P
[18:04] <teknico> mandel, that's also true ;-)
[18:04] <teknico> still in the harmony branch?
[18:05] <mandel> teknico: not, they are at the lp:~mandel/desktopcouch/contacts-wrapper
[18:06] <mandel> teknico: I think I removed around more than 400 lines to do the same work
[18:06] <teknico> mandel, oh, right, I thought you were talking about the whole of macaco code
[18:07] <teknico> I was a bit surprised :-)
[18:07] <teknico> mandel, less code is more good ;-)
[18:07] <mandel> teknico: I wish! for that I have to do the ui, and working with glade is slow...
[18:28] <mandel> laters, coffee is needed
[18:54] <tcole> what's the procedure to authorize a machine again?
[18:59] <rtagger> tcole: Remove keyring info from seahorse - Ubuntu One token for https://ubuntuone.com and desktopcouch secrets, I believe.
[19:01] <tcole> hm, in this case I have a user for whom the keyring info is already removed
[19:01] <tcole> it isn't automatically re-adding it
[19:16] <rtagger> tcole: and this is really interesting, we have a bug report about this. Do you have an access to the machine experiencing this at the moment?
[19:16] <tcole> unfortunately no
[19:16] <tcole> I'm just triaging bugs at the moment
[19:18] <rtagger> tcole: :) I guess we are talking about the same bug. I remember someone here reported about this issue and we made it work by running ubuntuone-client-applet from terminal. It magically started the browser and he re-added his computer
[19:18] <tcole> hmm
[19:19] <rtagger> tcole: browser... was there some issue with a browser not being the firefox by default?...
[19:19] <tcole> not sure
[19:20] <tcole> it should be using xdg-open
[19:20] <tcole> rather than firefox specifically
[19:20] <rtagger> tcole: yes, but I clearly remember someone was talking here about opera not being cooperative
[19:20] <tcole> ahh
[19:21] <rtagger> tcole: because as far I have examined the code, the only thing that is possible if no token is found in keyring is to start the browser and navigate to u1 login screen.
[19:23] <rtagger> tcole: Bug #419292
[19:23] <tcole> ah, interesting
[19:24] <rtagger> joshuahoover: what kind of problem opera was/is having with ubuntuone client?
[19:25] <rtagger> tcole: ah, this is bug #385399 - localhost port access is not granted
[19:25] <tcole> hm, that doesn't seem to be what they're experiencing
[19:29] <dobey> hrmm
[19:30] <dobey> tcole: no error dialog or anything?
[19:30] <tcole> dobey: evidently not
[19:30] <tcole> errors from no token in the keyring, and it all ends there
[19:30] <dobey> tcole: applet icon turns to "connected" and then back to "disconnected" in a few seconds?
[19:32] <dobey> tcole: there have been some bug reports that suggest ~/.config/ubuntuone/syncdaemon.conf having bw throttling set to on = True might be related
[19:32] <tcole> hm
[19:32] <dobey> with the limits set to -1 or 0
[19:33] <dobey> (0 would be bad, -1 is supposed to be infinite)
[19:36]  * rtgz is new for rtagger, rtagger is a tagging application, while I am known everywhere as rtg :)
[19:37]  * rtgz added z so that it is somehow different from other rtg. So, Empathy crashed and left me with...
[19:37] <rtgz> tcole: are you talking about bug #452589
[19:38] <tcole> ahh
[19:38] <tcole> sounds like it
[19:38] <tcole> launchpad's bug search is pretty useless
[19:38] <rtgz> tcole, then this transcript might be interesting http://irclogs.ubuntu.com/2009/11/20/%23ubuntuone.html#t00:22
[19:40] <tcole> hmm
[19:43] <rtgz> tcole, and the worst thing here is that it resolved by itself
[19:43] <tcole> :/
[19:45] <dobey> weird
[20:10] <rtgz> okay, it may be nice to set up DEBUG log for oauth modules and then ask people to reproduce.
[20:11] <rtgz> unfortunately this will require source code modification but they may be ok with this given the exact instructions or better yet, a script to handle this for them.
[20:13] <joshuahoover> rtgz: are you referring to bug 452589?
[20:13] <dobey> rtgz: i'm wondering if there is a difference in behavior between 1.0.2 and 1.1 with that
[20:13] <rtgz> joshuahoover, yep
[20:13] <dobey> rtgz: ie... is the bug already fixed in trunk...
[20:14] <rtgz> e.g. sed -ie 's/logging.INFO/logging.DEBUG/' /usr/share/pyshared/ubuntuone/oauthdesktop/logger.py
[20:14] <joshuahoover> rtgz: that particular bug looks like it was related to a temporary error on the server side based on what is in the oauth-login.log
[20:15] <rtgz> joshuahoover, yep, this is right, I've already tried reproducing that and one has to have server really broken for this thing to reproduce, but the other issue is that token does not re-appear.
[20:16] <rtgz> joshuahoover, e.g. the token got killed and it does not want to be created again. I posted the link to the discussion here of a live example of such behavior
[20:16] <rtgz> urbanape, hi, may I ping you a little bit ?
[20:18] <rtgz> tcole, you said you were triaging bugs and found the one, what was the bug #?
[20:19] <tcole> I think I ultimately ended up marking it a duplicate
[20:19] <tcole> let me see if I can find it again
[20:19] <joshuahoover> rtgz: but the token did get reproduced in that case, no?
[20:20] <rtgz> joshuahoover, in that case, yes, after he killed the applet and restarted it. But it has not been working for him for 2 weeks or so, so I don't really see what was special with that kill.
[20:20] <joshuahoover> rtgz: true
[20:20] <rtgz> tcole, I guess I found, 3 duplicates, right?
[20:21] <rtgz> joshuahoover, I guess I'll file a bug so that oauthdesktop/logger.py will take log level from some file/setting, not hard-coded.
[20:21] <joshuahoover> rtgz: yes, that would be good! thank you!
[20:23] <tcole> was a duplicate of bug #452589
[20:23] <tcole> I forget which one now
[20:25] <rtgz> joshuahoover, I am constantly thinking about the utility that will examine the logs on the client and provide some useful info for the person, i.e. Throttling = on/ min/max=-1 - fix it. Permission denied - fix it. Token missing - gather more info and paste it somewhere, etc...
[20:26] <rtgz> joshuahoover, something like ubuntuone-client-diagnostics :)
[20:26] <joshuahoover> rtgz: you and me both!
[20:30] <dobey> i don't think we should have a script to fix problems with config/etc... such problems should be fixed with updates
[20:35] <rtgz> dobey, yes, but there are "known bugs" for the release, we know they exist, they are fixed, say, in PPA. But not in karmic-updates and sometimes there are workarounds that are easy to implement. Or analyze logs on client side by finding known patterns and advice about solutions...
[20:36] <dobey> rtgz: if it's fixed in the PPA and a known bug, I don't necessarily know that.
[20:36] <dobey> rtgz: but if it's fixed in the PPA already, and a known bug, and it keeps affecting people, then we need to escalate the importance, and get it backported to stable and in an SRU for karmic
[20:37]  * rtgz was a developer of update system once, not yum update, apt-get update, but.. Look in this file, set this line to 2 if that line was 1. Detect the running configuration fix the configs, etc...
[20:37] <joshuahoover> dobey, rtgz: an example would be bug 455544
[20:38] <joshuahoover> dobey, rtgz: this is one we're going to release in an sru but it's not there yet for karmic so there are workarounds right now
[20:38] <dobey> joshuahoover: i think that's actually a twisted bug...
[20:38] <rtgz> joshuahoover, yep, even I got caught by this bug, like what? "Protocol version error"??? DivisionByZero o_O
[20:39] <joshuahoover> dobey: that is ultimately caused by bandwidth throttling settings
[20:39] <joshuahoover> dobey: at least that's what verterok has told me :)
[20:39] <dobey> joshuahoover: well, that's the visible cause, yes.
[20:40] <dobey> joshuahoover: -1 is supposed to be infinite, according to what verterok told me when he wrote the feature :)
[20:40] <dobey> joshuahoover: which means that on with values of -1 should be the same as off
[20:40] <joshuahoover> dobey: hahaha!!! i've told verterok and rmcbride i hate those settings...-1 and 0...confusing
[20:40] <rtgz> dobey, he should have used ∞ instead *offtopic*
[20:40] <dobey> joshuahoover: of course, that is apparently not the case, so i think maybe something is wrong in twisted
[20:41] <dobey> rtgz: i don't think that is a valid integer
[20:41] <joshuahoover> dobey: there's bigger problems with bandwidth throttling anyway, where you can set it too low and break the client because storage protocol can't get through
[20:42] <rtgz> dobey, later on there is a division by zero triggered by something that set -1 => 0.
[20:42] <dobey> ah
[20:42] <dobey> right
[20:43] <dobey> joshuahoover: well "0" is too low, yes
[20:43] <dobey> joshuahoover: but in the UI, 1 is "1 KB/s"
[20:43] <dobey> connecting shouldn't require more than 1K/s
[20:44] <dobey> if it does, then protobuf is insane :)
[20:44] <rtgz> dobey,  there was such line "throttle_time = (float(self.writtenThisSecond) / self.writeLimit)", with
[20:44] <dobey> rtgz: yes sure
[20:46] <rtgz> Additionaly I am scared about the following - what if I connect via my cell phone and don't really think that my 100Mb of the limit are worth being used by ubuntuone at the moment, like in emergency... But this is definitely another subject...
[20:47]  * rtgz is generating more questions than writeLimit...
[20:47] <dobey> it'll get used by gwibber instead :)
[20:47] <joshuahoover> heh
[20:48] <dobey> or something. connecting to the internet via gprs/3g/whatever is very suboptimal in the desktop right now
[20:48] <dobey> not just with ubuntuone, but with pretty much anything
[20:49] <rtgz> dobey, I spent 2 months on GPRS this winter...
[21:41] <maurizio__> Hi
[21:41] <maurizio__> I've installe karmic koala, I've yet an ubuntuone account, but I don't know how to configure the client to make it work
[21:47] <dobey> maurizio__: opening Internet->Ubuntu One should take you through the set up process to get your computer connected
[21:48] <maurizio__> dobey in preferences I can't do anything.
[21:49] <dobey> maurizio__: if preferences opens, then the applet is already running, and you should see the icon on your panel
[21:49] <maurizio__> it has an x on it
[21:49] <dobey> maurizio__: but i fear you may be running into rtgz's new favorite bug
[21:49] <dobey> maurizio__: click on it and choose "Connect"
[21:50] <maurizio__> already done. nothing happens
[21:50] <rtgz> dobey, woo-hoo, reproducible!
[21:51] <rtgz> maurizio__, hi, before it fixes itself, may we take some 5 minutes of your attention? We need to get some more debugging info.
[21:51] <maurizio__> ok
[21:51] <maurizio__> ask me?
[21:51] <maurizio__> ask me
[21:51] <rtgz> maurizio__, 1. sudo  sed -ie 's/logging.INFO/logging.DEBUG/' /usr/share/pyshared/ubuntuone/oauthdesktop/logger.py
[21:52] <rtgz> maurizio__, 2. killall ubuntuone-client-applet
[21:52] <maurizio__> done
[21:52] <dobey> you could just right click and choose quit.. you don't have to tell people to type killall
[21:53] <dobey> and killall is a bad idea for Solaris users :)
[21:53] <rtgz> maurizio__, 3. start ubuntuone by  Applications->Internet->Ubuntu One
[21:53] <maurizio__> Don't worry I don't use solaris
[21:53] <dobey> sure. just general knowledge :)
[21:54] <maurizio__> done
[21:54] <rtgz> dobey, okay, just sometimes the applet hides itself, so just to make sure. Okay, will be more applet-friendly next time :)
[21:55] <dobey> rtgz: yes, but it only hides while it is connected. disconnected state always shows the icon :)
[21:55] <rtgz> dobey, Frankly speaking I don't like the default behavior of hiding if no activity is performed. It is not that obvious that I have to do something for my cloud to appear
[21:55] <maurizio__> rtgz it doesn't work anyway
[21:55] <dobey> rtgz: you can discuss that with design team
[21:55] <dobey> rtgz: of course, it won't matter in lucid, as the applet is going away anyway :)
[21:56] <rtgz> maurizio__, cool, this is what we are trying to prove. Now could you paste ~/.cache/ubuntuone/log/oauth-login.log to http://paste.ubuntu.com?
[21:56] <maurizio__> ok
[21:57] <maurizio__> http://paste.ubuntu.com/328030/
[21:58] <rtgz> dobey, is NetworkManager dependency gone for 1.0.2 ?
[21:58] <maurizio__> rtgz http://paste.ubuntu.com/328030/
[21:59] <rtgz> maurizio__, I suspect you are not using NetworkManager, right? And you were not able to use ubuntuone previously, right?
[21:59] <maurizio__> I'm using wicd
[22:01] <rtgz> maurizio__, then this is a known bug and I advise you update to PPA version where it is fixed: https://launchpad.net/~ubuntuone/+archive/beta
[22:01] <rtgz> maurizio__, just a moment, will make sure it IS fixed...
[22:01] <maurizio__> ok
[22:01] <maurizio__> a second
[22:02] <rtgz> maurizio__, yep, we will assume that we are connected if NetworkManager is not found so you will be able to use ubuntuone with wicd
[22:03] <rtgz> maurizio__, so feel free to update to PPA version.
[22:06] <dobey> rtgz: the NM fix is only in trunk
[22:06] <joshuahoover> dobey: it's in the ppa beta release i believe
[22:06] <dobey> yes
[22:06] <dobey> but it's not in the stable branch yet
[22:06] <dobey> it's one of the many things i need to backport for SRU
[22:06] <joshuahoover> dobey: right
[22:07] <dobey> and hopefully i will be able to get that out early next week
[22:07] <dobey> until then, i am off to try and ejoy the holiday :)
[22:07] <joshuahoover> enjoy!
[22:08] <rtgz> dobey, erm, it is definitely in beta, since I am using the beta and the code is there :)
[22:08] <rtgz> dobey, yep, enjoy the holiday :)
[22:08] <joshuahoover> rtgz: i think he meant it's not there for karmic
[22:09] <dobey> i meant it's in trunk, and beta PPA is built from trunk, not stable branch
[22:09] <dobey> bye :)
[22:10] <rtgz> joshuahoover, btw, this is one more bug that could be easily diagnosed on the client side by grepping 'Unable to contact NetworkManager'.
[22:11] <joshuahoover> rtgz: sigh...if only it were that simple...i've seen a number of bug reports where they weren't using networkmanager but we didn't get any specific networkmanager errors
[22:12] <rtgz> joshuahoover, but being on the client we can probe for networkmanager as well. Okay, since I still need more python practice - will see what I can do tomorrow with the errors listed on UbuntuOne/Bugs page.
[22:12] <joshuahoover> rtgz: cool...thanks again for all your help! we greatly appreciate it!
[22:13] <rtgz> okay, time to sleep. Good night to everyone running UbuntuOne :)
[22:13] <joshuahoover> goodnight rtgz :)
[23:16] <maurizio__> rtgz done
[23:16] <maurizio__> thankyou
[23:16] <maurizio__> but it is possible to indicate different folders to sync with ubuntu one?
[23:16] <maurizio__> is it possible?