[01:31] Hey, I just got an error called capabilities mismatch [01:32] I am running 9.10 on a 20" imac [01:32] I just got an error called capabilities mismatch [01:59] ikelewis: run the update-manager and make sure you're up to date. it sounds like you don't have the latest updates for 9.10 installed [03:21] home sweet home [03:22] woops [03:30] is there documentation for the desktopcouch.records python api somewhere? === cpg is now known as cpg|brb === cpg|brb is now known as cpg === cpg is now known as cpg|brb === cpg|brb is now known as cpg [11:37] Good morning! [11:52] duanedesign, mornings! [11:58] rye: next UDS announced! How far is Brussels from Kiev? [11:59] the last UDS was 4hrs by car from where i live. Unfortunately by Grandpa passed away the day before it started :( [11:59] duanedesign, http://maps.google.com/maps?f=d&source=s_d&saddr=Brussels&daddr=Kiev&hl=en&geocode=&mra=ls&sll=50.400284,30.624805&sspn=0.006825,0.01929&ie=UTF8&t=h&z=5 [12:07] duanedesign, that's pretty far from here, plus visa. need to apply for that... [12:07] duanedesign, btw, do you use gwibber? [12:07] duanedesign, ah, you are not on lucid, right? [12:07] it looks like there is something really wrong with couchdb in lucid lynx [12:11] rye: Yes i use Gwibber. I only have Lucid in a VM [12:11] rye: i have been really wanting to upgrade to Lucid [12:12] hm [12:12] dc guys, is desktopcouch-service detects desktopcouch port using the dbus call to itself? aquarius, CardinalFang ? [12:13] can't it be in erm... recursive call? [12:16] rye, hi. The service passes the PID of the couchdb it cares about, which makes it go down another code path and get the port directly. [12:18] CardinalFang, hm... something is definitely wrong somewhere - I can't get to the stage where replication is actually starting. Additionally, getPort() stops doing things [12:18] leading to broken communication with couchdb [12:18] which is happily running [12:19] rye, hrm. is that a package or trunk? [12:19] CardinalFang, package... [12:20] rye, what version? [12:21] CardinalFang, Installed: 0.6.2-0ubuntu1 [12:21] rye, Okay, there was an error in building that package. Do you know how to build a package from source? [12:22] rye, $ mkdir /tmp/a; cd /tmp/a; bzr branch lp:~statik/ubuntu/lucid/desktopcouch/fix-bad-upload [12:22] CardinalFang, hm, an error in building? is there a bug report? 'cause otherwise it looks like it does something... but not completely... [12:23] $ cd fix-bad-upload; dpkg-buildpackage -rfakeroot [12:23] CardinalFang, ah desktopcouch/records/server_base.py and desktopcouch/replication.py [12:24] just what I needed :) [12:24] Yes. [12:25] CardinalFang, are there any plans to release the patched version as lucid package ? [12:26] rye, we're not leaving it broken. :) [12:26] CardinalFang, ok, thanks :) [12:26] rye, you see something hours old. wait until st-tik wakes up. [12:26] 'cause while this is a honor to be in the privileged set of those who have it working in Lucid that's not that great :) [12:29] CardinalFang, one more - is it ok that desktopcouch-stop does not stop desktopcouch-service ? [12:30] rye, it's not well-named, but it was supposed to stop couchdb only. [12:31] rye, really, we shouldn't redistribute it. [12:33] CardinalFang, only for tests, right. Is there any spec re the database names that are allowed [12:33] CardinalFang, how did the DC changes go? Is there anything you need from me? [12:33] rye, database names? What? [12:33] CardinalFang, i.e. i created "rtg/test" database and I get ServerError: (500, ('case_clause', '{error,{{{case_clause,{error,illegal_database_name}} :) during replication [12:33] aquarius, ^ [12:34] aquarius, I am looking for problems with DC now. I don't need anything, except maybe some eyeballs to make sure it's working there too. [12:41] CardinalFang, sweet. good work. How did you sort out the d-bus problem? [12:54] Whew! Freshly showered now. [12:55] aquarius, I don't know that I did. Let's back up; I don't know what you are starting from. The packages from 24 h ago were bad. [12:55] This, lp:~statik/ubuntu/lucid/desktopcouch/fix-bad-upload , has better. [12:56] kenvandin- said those were much better, but said there maybe something still screwy, just before we both passed out last night. [12:57] aquarius, so, I'm starting this morning by looking for problems. [12:57] CardinalFang, ah, what I was talkign about was the set-an-envar thing so that get_port knows that you're desktopcouch-service, where you were looking for a better way to do it, and d-bus was complicating matters [12:58] ah, right. I retracted that merge proposal. At least most of that problem was bad packaging. [12:58] aquarius, so, lp:~statik/ubuntu/lucid/desktopcouch/fix-bad-upload . dpkg-buildpackage! === teknico is now known as teknico_away [13:05] CardinalFang, something is wrong with this dc as well [13:06] aha! [13:07] my rtg/test database breaks it, and replication service exits completely [13:10] Segmentation fault (core dumped) [13:10] desktopcouch-service [13:11] CardinalFang, aquarius - [15812.548833] desktopcouch-se[23026]: segfault at 4 ip 00b7fa96 sp bfa9ff58 error 6 in libdbus-1.so.3.4.0[b5a000+37000] [13:13] rye, Hrm. [13:13] gdbing it.... [13:13] If that works, I want to know how. [13:14] segv above "Python_eval()" doesn't count. :) [13:19] CardinalFang, http://paste.ubuntu.com/387592/ [13:28] pdb is of no use - it segfaults :) [13:28] is there any possibility in python to trace the execution statements, like set -x in bash ? [13:31] The problem cannot be reported This is not a genuine Ubuntu package...grrr [13:35] rye, not that I know of. pdb does have some dark dusty corners that may conceal tricks like that. [13:35] bt did not show the presence of dbus [13:35] but the dmesg says we crashed in dbus [13:35] libdbus === teknico_away is now known as teknico [13:40] CardinalFang, http://paste.ubuntu.com/387600/ - are you sure it goes through a different path to find port? [13:42] python: malloc.c:4628: _int_malloc: Assertion `(unsigned long)(size) >= (unsigned long)(nb)' failed. [13:42] Aborted (core dumped) [13:42] wow [13:42] rye "dbus_io.get_seen_paired_hosts()" is old code. === vds1 is now known as vds [13:44] CardinalFang, hm, I am running latest dc that I've built from st-tik's branch [13:44] rye, first 8 or so are old. line 255 has first new problem. [13:44] aquarius, https://code.edge.launchpad.net/~cmiller/desktopcouch/fix-replication-getport-nouri/+merge/20546 [13:45] CardinalFang, I don't understand why that works [13:45] CardinalFang, ah, that's old and new code writing to the log,... [13:46] aquarius, I should write better commit messages, then. [13:46] aquarius, if we know the port, we don't try to look for it. The service knows the port. Everything it calls should be given the port. [13:47] In this case, it's in uri/url form. [13:47] CardinalFang, oh, so because dc-srv knows the port already, it tells dc.records to *explicitly* use the port it knows about, rather than implicitly working it out? [13:53] directhex: speak to aquarius [13:53] * directhex speaks to aquarius [13:53] heya, directhex :) [13:54] oh, THAT aquarius. ello [13:54] speak to me about what? [13:54] * aquarius grins [13:54] aquarius, i think i might have a working u1 (therefore u1ms) mono binding. is there a magic way to let my testu1ms.exe see something more than just "coming soon"? [13:54] i.e. for banshee [13:54] assuming this works without exploding [13:55] directhex, ya, there is. You're running lucid, yes? [13:55] directhex, I'll send you the Secret Beta Test Instructions Email :) [13:55] aquarius, testing in a lucid pbuilder or vm, but let's say yes [13:55] ta! [14:00] aquarius, exactly. [14:01] aquarius, evidently, it can't work it out. Thread A is serving the port information. Thread B wants port information. They don't play nice together. [14:01] So, they both use the same info now. [14:02] CardinalFang, cool [14:28] pfibiger, your question about my phone service made me check last night. [14:29] what'd you learn? [14:31] pfibiger, I found out a way to save almost $500 per year! (Yay http://www.billshrink.com/ .) [14:31] pfibiger, so I immediately blew it all on a Nexus One. [14:38] hey muffinresearch! [14:45] hm, nessita, hi, may i ask you whether the bugs #494218 and #487257 are related ? [14:45] Launchpad bug 494218 in ubuntuone-client "Fix ActionQueue and ActionQueueProtocol disconnect and connectionLost features" [Medium,In progress] https://launchpad.net/bugs/494218 [14:45] Error: Could not parse data returned by Launchpad: The read operation timed out (https://launchpad.net/bugs/487257) [14:46] cool [14:46] ubottu, bug 487257 [14:46] Launchpad bug 487257 in ubuntuone-client "The client gets stuck into STANDOFF state" [High,Triaged] https://launchpad.net/bugs/487257 [14:46] ubottu, thanks [14:46] You're welcome! But keep in mind I'm just a bot ;-) [14:47] rye: yes, totally related [14:47] so, nessita - the STANDOFF bug seems to arise because the disconnect is not detected properly, whereas the bug title for "Fix ActionQueue ... connectionLost features looks promising" [14:47] rye: and you made me remembert o update status [14:47] rye: indeed [14:47] rye: facundobatista and I are right now working on that [14:47] * rye wants to mark the STANDOFF-bug as a duplicate of the first one... [14:49] joshuahoover, bugs are related and one is the result of another, should I consider them as separate ones or I can make one as a master bug report for another one ? ^ [14:49] rye: is not... [14:50] nessita, erm... yeah, one is created by some issues that are the target of fixing for the another bug report. Is that correct? [14:50] rye: yes [14:50] ^_^ [14:55] rye: hmmm... [14:56] rye: i wouldn't want to mark a bug as a dupe if it's related but not truly a dupe...i guess it's up to you and whoever would be working on it...if it makes sense to dupe it to you and nessita then go for it, otherwise i wouldn't [14:57] joshuahoover, ok, not flagging as duplicate. will just tag them somehow... [14:58] hi folks :) [15:01] is there kde support too till now [15:01] or is the gnome client fine on kde [15:10] jms@osc-bigmac:~/Projects/libubuntuone/bindings/mono$ make && ls *.dll [15:10] ubuntuone-sharp.dll [15:16] aquarius, did you review that patch? I'm eager to re-release. [15:16] thisfred, https://code.edge.launchpad.net/~cmiller/desktopcouch/fix-replication-getport-nouri/+merge/20546 [15:17] CardinalFang: I's on it [15:17] CardinalFang, approved. I'm still a bit worried about this; specifically, it feels inelegant and that there are now loads of little special-case hacks all over the codebase :( [15:17] convince me that I'm wrong [15:18] aquarius, you're wrong. [15:20] con·vince [kuhn-vins] –verb (used with object),-vinced, -vinc·ing. to move by argument or evidence to belief, agreement, consent, or a course of action [15:20] aquarius, the only special case stuff is in replication, which is called from a single place, and in find_port, to avoid calling DBus methods when we have a couchdb PID in mind in particular, which may not be the dbus-created couchdb. [15:21] aquarius, all replication now never tries to discover the port through DBus. [15:22] ...because that doesn't work from the same process as the server. [15:22] because replication is *always* being run from dc-srv, because no-one else ever calls it? [15:22] Yes. [15:22] and find_port is special-cased so that it knows that it's being called from the dc-srv setup process and therefore doesn't call back through d-bus? [15:23] aquarius, Er, we shouldn't be calling find_port at all from replication. [15:24] aquarius, I was referring to tests, which start their own couchdbs and know the PIDs and DBus can't help them. [15:24] but find_port is called from dc-srv when you call over d-bus, right? [15:26] aquarius, dc-srv should start couchdb once, calling find_pid and then find_port(PID), and never call it again. [15:26] k, makes more sense now. Thanks for explaining! [15:27] CardinalFang: I get dbus errors on that branch :( [15:28] perhaps they're in trunk already [15:29] CardinalFang: hmm, not on trunk, and a second run does not result in errors [15:29] CardinalFang: yay indeterminacy in tests [15:29] joshuahoover, should I send the list of the bugs that I feel are important via mail? [15:29] e [15:29] rye: sure, maybe just reply-all to the email i sent yesterday [15:30] thisfred, yay. [15:30] CardinalFang: looks like it may be a dbus time out [15:30] is there a way to up that? [15:30] or deal with it in d-c? [15:30] "nxdomain hijacked" ha ha [15:30] CardinalFang: yeah, go comcast [15:32] or is that not what that print means? They say they let you turn it off, but it does not seem to work in FF [15:32] thisfred, hrm, I think that should not happen. It shouldn't use DBus at all. [15:33] CardinalFang: ok, I pasted the full test report, if you think it's unrelated, I'll approve but then we should file a bug [15:34] CardinalFang: running the tests as follows: PYTHONPATH=. trial desktopcouch [15:34] $ host eric.example.com [15:34] Host eric.example.com not found: 3(NXDOMAIN) [15:36] CardinalFang: don't know if you saw james w's second branch, that may help make (some of) the tests more unit and less functional [15:36] thisfred, no, I didn't. Where? [15:37] in the review queue [15:37] CardinalFang: basically he factored out the authentication and the keyring parts [15:39] I see it. It looks like he's not using it yet. [15:47] ok, me definitely needs to go away to grab some fresh air, terribly slow now. I am leaving the client connected so feel free to send messages especially regarding the bug reports :) [16:08] Ok, I just deleted my db from couch in futon. I started my app once, and it didn't show any data like it should, but on the next start of the app all the data is somehow back. === beuno is now known as beuno-lunch [16:46] aquarius, my binding kinda works, except i can't work out all the autofoo semantics [16:47] directhex, ok... [16:47] directhex, what can I do to help with that? [16:48] aquarius, i just need to find someone who can help me get an "install" rule which does something in proper autofoo style [16:48] rodrigo_, ^ [16:48] rodrigo_ actually set all that stuff up, so he should know :) [16:48] i'm hacking at lp:~directhex/+junk/ubuntuone-monobinding [16:49] right, it's gohometime now [16:56] directhex, is that a binding for libu1? [17:31] * rye is back and ready to think... somehow [17:54] it is, but rodrigo's escaped [18:04] hyperair, superstar that he is, has fixed most of the autofoo issues [18:07] nice! [18:08] directhex, so, do you still need rodrigo assistance to get it built? [18:08] grrr... the power went off causing the ISP router and my router to go down. The power returned after a second or so, the router got wifi started pretty quickly so there was no state change from NM. And now my syncdaemon is in START_STANDOFF_WAITING_WITH_NETWORK_WITH_CONTQ [18:09] aquarius, no, hyperair's a real whizz at this stuff, as well as banshee related things (hint hint if you have any control over uds sponsorship) [18:09] * aquarius grins. I don't, really :) [18:09] okay let me test build this... [18:11] yep, looks smooth to me [18:11] aquarius, revision 53 on https://code.launchpad.net/~hyperair/+junk/ubuntuone-monobinding contains all the good stuff [18:15] grrr, now I can't get it to reconnect except of quitting it... ggrrrrr === beuno-lunch is now known as beuno [18:23] out of interest, is there any restriction on u1ms regarding use in other distros? can upstream apps use u1ms as their default music store on non-ubuntu? [18:23] directhex, on a phone call [18:34] rye_: gonna look around, but i thought i would check with you to see if you had seen this exception.log before. bug 530803 [18:34] Launchpad bug 530803 in ubuntuone-client "Was logging out of system. " [Undecided,New] https://launchpad.net/bugs/530803 [18:51] directhex, hey [18:51] directhex, restrictions in other distros: no [18:51] duanedesign, that probably means that the user has upgraded to new version of metadata but returned to previous one [18:51] directhex, if someone does that, no problem [18:52] what's the latest ppa? [18:52] hm [18:52] aquarius, okay, cool [18:53] duanedesign, hm.. but what kind of metadata he has now... [18:54] directhex, questions like that are mattgriffin sorts of questions, mind, so he gets to override anything I say ont he matter. :) [18:56] directhex: we would love to work with other distros and upstream apps to use the Ubuntu One Music Store. we've already received approval to embed the store within a handful of apps (I think they're listed on the FAQ) but we can extend that list. [19:01] duanedesign, bug 517505 [19:01] Launchpad bug 517505 in ubuntuone-client "Right click on applet at computer startup made syncdaemon crash" [Medium,Invalid] https://launchpad.net/bugs/517505 [19:02] duanedesign, AttributeError: 'MetadataUpgrader' object has no attribute '_upgrade_metadata_6' [19:05] directhex, ping re your tweet? [19:05] hm? [19:05] directhex, I do non-root installs of libubuntuone to arbitrary folders about 60 times a day. [19:05] works for me [19:05] what's not working for you [19:05] ? [19:06] directhex, ./autogen.sh --prefix=/home/aquarius/code/libu1 ; make ; make install [19:06] aquarius, was spitting out errors when trying to write to system site-packages dir... on reflection, perhaps something's not getting updated at some point [19:07] you may have to touch /home/aquarius/code/libu1/lib/python2.6/site-packages/ubuntuone/__init__.py [19:07] directhex, weird if it doesn't work for you. I didn't do anything special to make it work. Are you doing the autogen step? [19:07] aquarius, yeah, but it might be related to a dirty build dir. maybe [19:18] guys, bug #531273 is a show-stopper for me... It does not start uploading anything until it scans all metadata... [19:18] Launchpad bug 531273 in ubuntuone-client "Local metadata rescan takes ages: 28 minutes for 1457 objects" [Undecided,New] https://launchpad.net/bugs/531273 [19:22] verterok: ^ [19:23] and I believe that will raise huuuuge complaints from our users when they put files and the files only start uploading after an hour or so.... === rye_ is now known as rye [19:23] adding a file does a whole local rescan? [19:24] hm... [19:24] * verterok keep saying isn't the metadata rescan... [19:25] rye: this is when the client it's started, right? [19:25] rye: also, 28 min? that's a lot! [19:25] dobey, ah... the files _do_ get uploaded.. but that happens at 1/2 of regular speed because every upload is happening in between of server query [19:26] i have enough problems to deal with right now anyway :) [19:26] verterok, sorry for 'metadata rescan' but I can't find out what it actually doing in query() starting, query() running [19:28] rye: is quering the server for changes [19:28] verterok, ok, i put libgnome-2.28 to the dir @ 21:08:13, it is now 21:28 and it is still has 177 files to go... [19:29] verterok, and the upload rate is 10Mbit! [19:29] rye: so, the bug is: upload is slow :) [19:29] verterok, the bug is the download is even slower when there is a query for server-side changes [19:30] verterok, it just picks the file, asks the server about it, receives the response and goes to another file? [19:30] rye: sorry, I don't understand [19:32] verterok, as I understand, there is a queue of operations, so when I have only uploads pending it is 1x slow. But when the server query is being berformed it is 2x slow because the calls to the server for every single file status interleave with the upload action. [19:33] rye, yes. also the meta queue commands are executed with higher priority than the content queue [19:35] :'-( [19:38] verterok, ... can we stuff them into some kind of bigger container, e.g. 100 requests in a row, then send it as a big parcel? [19:40] rye: 100 request of what? :) [19:40] http://www2.apebox.org/wordpress/wp-content/gallery/00-single/banshee-u1ms.png [19:40] ^^^ [19:40] ^^ [19:40] ^ [19:41] verterok, ok, let me give out the info on how I believe the syncdaemon behaves [19:41] rye: while you'r uploading a file, other client might be changing stuff in the same directory, e.g: uploading a file you have in the queue for upload [19:42] verterok, syncdaemon starts and eats the hard drive while it performs local rescan by means of os.walk [19:42] verterok, then it connects to the server (--connect) and performs server rescan - btw, what does that rescan mean? [19:43] verterok, then it loads info about all its files and starts to query server about each and every file system object it encounters. [19:43] verterok, then it becomes IDLE [19:44] rye: server rescan is "ask the server if something changed" [19:45] rye: point 3 is wrong, it doesn't check for every file [19:45] rye: just for the ones the server said that changed [19:46] directhex, sweet! [19:47] directhex, wanna talk over what the banshee plugin needs to do? a phone call might be quickest? [19:47] aquarius, totally not ready for human consumption yet. and i think there's a bug in the binding in hyperair's branch. but still, i'm pleased with it for a couple of hours' work total === teknico is now known as teknico_away [19:48] verterok, then it says that every server file has changed... hmmmm... coudl that mean that there is something wrong with UDF folders? [19:48] UD folders [19:49] rye: no idea, seems to work ok here. could you attach debug logs in a private bug? [19:49] verterok, ok, sure! [19:49] rye: in case there is info you don't want to be public in your file names :) [19:50] verterok, no idea... there are so many of them :) [19:50] heh [19:50] will mark my 'takes ages' report as private and will update it with a syncdaemon log of enormous size :) [19:50] rye: btw, local rescan is a bit more complex than os.walk, it's done incrementally and don't block scanning the directories [19:51] :) [19:52] verterok, I know only two terms, local rescan and metadata loading. I tend to use them whenever possible to describe that there is something being done by syncdaemon :( [19:52] rye: :( [19:52] * rye needs more terms [19:59] OKAY. revision 55 of lp:~directhex/+junk/ubuntuone-monobinding contains a happy, functional ubuntuone-sharp.dll with all the trimmings [20:07] verterok, it took 1h2m to upload 311 files (totalling 8.5M) to UbuntuOne during server query... [20:08] rye: s/during server query// :) [20:08] rye: something isn't good, do you have full debug logs of that run? [20:09] verterok, you believe that does not affect the upload? [20:09] verterok, I have all logs :) [20:09] rye: I mean, there is such thing as upload during server query [20:10] rye: a query to the server is a meta-command, and it's executed before content-commands (up/download) [20:15] verterok, I have added the logs to bug 531273, my syncdaemon is now in IDLE state [20:15] Bug 531273 on http://launchpad.net/bugs/531273 is private [20:16] rye: cool, thanks! [20:16] no, that's me thanking you :) [20:16] directhex, cool; when you get to the stage of actually hooking stuff into banshee, come talk to me and I'll explain what the rhythmbox stuff does, and see if there's anything else that you'd find useful [20:17] i don't know if it's a banshee bug or u1ms widget bug, but if i do a search, and scroll down, the text for entries i've scrolled past appears on top of the navigation bar (the thumbnails correctly go behind) [20:19] actually, hang on... [20:20] nope, not my fault afaik === jamalta-afk is now known as jamalta [20:31] OK, I ran updates, and restarted, and I am still getting the Capabilities Mismatch error. Any Ideas? [20:34] ikelewis, what's the output of apt-cache policy ubuntuone-client ? [20:34] OK just a sec [20:34] does ubuntu one support symlinks? [20:35] ubuntuone-client: Installed: 1.0.2-0ubuntu1 Candidate: 1.0.3-0ubuntu1 Version table: 1.0.3-0ubuntu1 0 500 http://us.archive.ubuntu.com karmic-updates/main Packages *** 1.0.2-0ubuntu1 0 500 http://us.archive.ubuntu.com karmic/main Packages 100 /var/lib/dpkg/status [20:35] JoshuaL: no, symlinks are ignored. but it will be possible to sync arbitrary folders soon [20:35] that is the result, sorry it is kinda messy [20:35] well mostly arbitrary [20:35] dobey, ok, since i hate having duplicates on my filesystem :) [20:36] ikelewis, you have 1.0.2 installed, you need to get 1.0.3, as indicated by 'Candidate' [20:36] so how do i do that, just sudo apt-get install ubuntu-one [20:36] ? [20:37] run update-manager [20:38] ok got it... thanks [20:40] ok, updated it, and then quit the ubuntu one icon in the status bar, and then opened it again from the internet menu of the main menu. Still has the error... [20:44] ok restarted it again, looks like it is working now... let me double check... [20:48] OK, figured it out... #1 - I did not have the right version, so I updated it. #2 - the checkbox, Limit Bandwith usage in the preferences was checked, with the defaults set to 0 and 0, so basically it was stopping all the network connections. I unchecked the box, and now it works. Thanks for all your help, dobey... [20:56] facundobatista_, is states-refactor ready for testing? I.e. just to see how that new state machine works (and look at svg diagram :) ) [20:56] rye, yes! https://code.edge.launchpad.net/~facundo/ubuntuone-client/states-refactor/+merge/20565 [20:57] rye, three SVGs, in docs/ [20:57] rye, one SVG for each manager (see states.py) [20:57] facundobatista_, yes, I saw the diff with svg file, can't render svgs in my mind at the moment :) [20:57] rye, oh, shame of you! [20:57] rye, :) [21:04] rye: to fix the issue with the new metadata aon an old client deleting ~/.local/share/ubuntuone will fix that? [21:05] duanedesign, that will create lots of conflicts - https://bugs.launchpad.net/ubuntu/+source/ubuntuone-client/+bug/414453 [21:05] Launchpad bug 414453 in ubuntuone-client "Removing .cache/ubuntuone creates useless conflicts" [High,Triaged] [21:05] rye: glad i asked [21:05] duanedesign, so he will need to move the folder to another place, perform the cleanup and then move files back, I believe... [21:07] rye: ok cp ~/Ubuntu\ One ~/Ubuntu\ One.bak ; remove the contents of ~/Ubuntu\ One ; start the client. Let it do its thing. Then put the contents back in ~/Ubuntu\ One [21:08] remove .local/share/ubuntuone before starting ubuntuone client [21:10] rye: ahh, ok. Thank you! === teknico_away is now known as teknico [21:21] wow [21:22] LOCAL_RESCAN state, me loves taht [21:22] Connection: Not User With Network :) [21:29] woo-hoo, no more crazy IDLE/non-IDLE... states. [21:29] I am testing states-refactor branch, FWIW [21:31] rye: are the nightlies being built again. I noticed there has been a few succeful builds lately [21:32] duanedesign, hm, I have to admit that I don't know where nightlies are... rmcbride or rmcbride_ ^ ? [21:42] rye: hmm, could be a coincedence but i have never seen this error before (_upgrade_metadata_6) and now two on the same day [21:43] rye: dear rye [21:43] rye: you have a reaily available karmic vm thing to test something in? [21:43] duanedesign, I usually use the trunk, i.e. something that has not been built yet [21:44] dobey, karmic vm is starting... [21:45] hmmm [21:45] guest kernelpanicks... [21:45] dobey, pls wait, changing CPU [21:46] bug 531597 [21:46] Launchpad bug 531597 in ubuntuone-client "Ubuntu one crash at login" [Undecided,New] https://launchpad.net/bugs/531597 [21:46] rye: please open the u1 folder and right click on the shared with me and let me know if it says "Share on Ubuntu One..." in the menu [21:47] dobey, hm, no, all my vms die [21:47] after some upgrade [21:48] eh ok [21:48] anyway i am 9000043434399% sure it is broken [21:49] in kvm_leave_lazy_mmu ... [21:49] hmmm [21:49] dobey, can you tell me what you want to do ? [21:50] rye: read up a few lines. i already did :) [21:52] lol, for a bug description i got: Robert 61 years old Male. Sounds like a singles Ad [21:53] one moment, I'll have the nightlies loc [21:54] duanedesign: https://edge.launchpad.net/~ubuntuone/+archive/nightlies [21:54] rmcbride_: oh wow! [21:54] thank you. Fun stuff [21:54] duanedesign: and I did resume nightlies recently. [21:55] rmcbride_: i noticed some recent succesful builds and was hoping that was the case :) [21:55] I can't start KVM VMs... aaa!!! I depend on them... aaa! [21:55] duanedesign: no problem. I should have a new client in the Beta PPA shortly as well [21:55] duanedesign: yea I've been doing a lot of testing of things since we layed off of nightlies for about a month before I started copying stuff to Beta again [22:00] Hi, should the notes issue affect syncing between Tomboy and u1 if I've created notes in both separately? Ran tomboy --debug and see a call to https://one.ubuntu.com/notes/api/1.0/user/ that gives HTTP 500... red herring? [22:04] Actually, I think it is ... looking closer I have an error: [ERROR]: Synchronization failed with the following exception: a name did not start with a legal character 32 ( ) Line 6, position 107. [22:05] So I'll pare down my (u1) notes until I find the one with the illegal character. Ciao all. [22:05] glennji, I believe you have hit https://bugs.launchpad.net/ubuntuone-servers/+bug/527374 [22:05] Launchpad bug 527374 in ubuntuone-servers "XML entities are not escaped during note save - /notes/ oopses" [Critical,In progress] [22:05] that illegal character is "&" [22:05] Yeah, okay, sounds like I confused myself :-) [22:05] Thanks! [22:05] glennji, do you know where to look for notes? [22:06] glennji, they are in your local couchdb as well [22:06] glennji, and if you fix the HTML of those notes then they will be working properly... well, that's the best workaround we currently have... [22:07] Cool, I'll give it a go. [22:07] So the u1 notes should be in a local couchdb already? Or they're only added when parsed okay? [22:08] rmcbride_: i did a little work on adding a class to mago's mago/applications/ubuntu.py so i could create some tests [22:08] duanedesign: awesome! [22:08] glennji, couchdb does not care about their validity, it stores them anyway. [22:09] Cool, thanks rye. I'll read up on how to access couchdb now. [22:09] glennji, see http://blog.rtg.in.ua/2010/02/ubuntu-one-internals-notes.html [22:10] Perfect, thanks rye. [22:10] glennji, you are welcome [22:10] * rye will reboot now, will see whether updates have fixed my kvm... [22:10] rmcbride_: i got a little distracted with a couchdb book but i really want to get back to that. [22:11] duanedesign: That will be pretty helpful. I'm just doing some straight testdoc based stuff now, but I really want to get more stuff in lp:ubuntuone-testing [22:12] I haven't added to that in quite a while [22:16] rmcbride_: yes i have ben looking at the stuff on ubuntuone-testing. Oh i found a neat magazine, you have probablly heard of it. 'Software Testing Club' [22:16] duanedesign: the first rule of Software Testing Club... [22:17] duanedesign: actually I had not heard of that one [22:17] although I suppose you could also say a Software Testing Club is like a LART stick for coders [22:20] duanedesign: wow. Adding that to my RSS reader. looks like good stuff [22:23] rmcbride_: oh, i guess this was there first issue. [22:24] duanedesign: yea looks like. They've got a lot of content and feeds on line as well. interesting [22:25] duanedesign, http://paste.ubuntu.com/387882/ [22:26] duanedesign, finally, something that is somehow usable and not that flickering... but it is still flickering... [22:28] rmcbride_: yes i had not been to the site till just now. [22:28] rmcbride_: I think i heard about the magazine from Ara Pulido's blog. Ara wrote an article on pg. 32 [22:29] duanedesign: you know I might have seen it mentioned there too. It sounded familiar, like one of those things that was mentioned to me while I was working on something [22:29] and I think I've made those "Software Testing Club" jokes at least once before :) [22:30] rye: nice, thanks [22:31]