/srv/irclogs.ubuntu.com/2010/02/25/#ubuntuone.txt

olyhi, i am having problems syncing ubuntu one on one machine i have 09.10 and 10.04 on the other do the clients differ in these versions ?09:13
olycould this be the cause and if this is the case is there a ppa i can use to update 09.10 ?09:14
popeyoly: does u1sdtool -s show that both machines are connected?09:22
olyi know the 10.04 machines are connected, but the 9.10 machine is not09:23
olythat command does not work by the way no -s parameter09:24
olybut intresting to know was not aware of that command09:24
aquariusu1sdtool is significantly nicer in 10.0409:25
olyin nautilus you hit connect and it does not even attempt to connect looking at it09:25
olyi am on the 9.10 machine curently09:25
aquariusin 9.10, you need to look at the log file, ~/.local/share/ubuntuone/log/syncdaemon.log, and see if it contains things that look like errors09:25
olythe 10.04 machine at home are connected and syncing nicely09:25
olyclose file is actually located in ~/.cache/ubuntuone/log/syncdaemon.log on my machine09:27
olystate: STANDOFF_WAITING_WITH_NETWORK_WITH_BOTHQ09:27
olyhas lots of those states from this morning09:27
aquariuser, yeah, .cache, sorry :)09:29
aquariushm, if you're at standoff then (1) restarting the syncdaemon may help (2) you want to be talking to someone more knowledgeable about syncdaemon states than me :)09:29
aquariushttps://wiki.ubuntu.com/UbuntuOne/Bugs may be useful, if you've not already read it09:30
aquariusand rye knows about this stuff more than I do, so he may be able to help when he wakes up09:30
duanedesignoly: is there anything in ~/.cache/ubuntuone/log/syncdaemon-exceptions.log09:33
aquariusaha, or duanedesign :)09:34
olynope seems to be empty09:34
duanedesignhis ears must of been burning :)09:34
duanedesignhello rye09:34
ryeduanedesign, hello, i am about to upload "The art of unsuspending" :)09:34
duanedesignoly: could you pastebin the syncdaemon.log09:35
olyi just killed the sync deamon now the icon correctly shows not connected :p09:35
olyyeah09:35
duanedesignhttp://ubuntu.pastebin.com/09:35
ryeduanedesign, http://picasaweb.google.com/lh/photo/E9ukoCzOnTd8x0O5OJHqMw?feat=directlink09:35
duanedesignrye: I like the wallpaper :P09:36
ryeduanedesign, desktopcouch did not start eating my CPU after unsuspend but I got an awesome wallpaper, right.09:36
olyhttp://paste.ubuntu.com/383571/09:37
olythats part of it but its the same messages repeated09:37
ryebug 487257 ?09:38
ubottuLaunchpad bug 487257 in ubuntuone-client "The client gets stuck into STANDOFF state" [High,Triaged] https://launchpad.net/bugs/48725709:38
ryeduanedesign, oly - I believe there is disconnect info somewhere09:39
olyit looks like it may be fixing itself09:40
olyi killed the sync daemon and have reconnected09:40
olynow says syncing09:40
olyalthough the applet seems to get confused between if its connected or not09:40
duanedesignoly: can you run: dpkg -l ubuntuone-client09:44
olyhttp://paste.ubuntu.com/383575/09:44
duanedesigncheck whick version you have?09:44
olythat may be a more usfull error log as it has a division by zero error :)09:44
duanedesignaha09:44
oly1.0.3-0ubuntu109:44
duanedesignoly: check your preferences09:45
olyalready disabled the throttling09:45
olybecause of that error09:45
duanedesignok09:45
duanedesignoly: you might also want to delete ~/.config/ubuntuone/syncdaemon.conf09:47
olydone09:47
duanedesignbug 50974009:48
ubottuLaunchpad bug 509740 in ubuntuone-client "Maximum upload speed does not save consistently in client GUI" [Undecided,Confirmed] https://launchpad.net/bugs/50974009:48
duanedesignoly: you might of already saw that. But thats the bug related to the error09:49
olyyeah, now i recons i am connected and my files are in sync which they most certainly are not :/09:53
olyi shall have a look a bit later though09:53
duanedesignoly: you can run : u1sdtool --current-transfers09:55
duanedesignto check current transfers09:55
olysays 0 and 009:56
olyi am attempting the nirvana option09:57
ryeduanedesign, oly - could you please give me some info on what is happening?09:58
ryeplease :)09:58
duanedesignrye: he had 50974009:58
duanedesignhttp://paste.ubuntu.com/383575/09:58
duanedesignturned off 'limit bandwidth'09:59
ryeaha, float division due to 0 in one of the fields09:59
duanedesignand deleted syncdaemon.cnf09:59
duanedesignthats about where we are09:59
olyi also have 2 machines at home which are 10.04 and one here which is 09.10 that the newer ones are syncing fine09:59
olywas not sure if it was version differences10:00
popeyis u1 known to be flaky behind a proxy?10:00
olyu1sdtool -w recons my network connection is broken10:00
duanedesignoly: you might try quiting the client by r-clicking the applet. Running u1sdtool -q to quit the syncdaemon10:01
popeyoops, sorry, I'll wait until oly is finished10:01
olyFailure: dbus.exceptions.DBusException: org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.10:01
olythat was with the wait command10:01
olyi have quit using -q and tried again with -w and its given same error10:02
duanedesignok10:03
duanedesignoly: ps uaxxc | grep ubu10:04
duanedesignanythingcome back?10:04
ryeoly, the syncdaemon has a bit long startup before it can service dbus requests. Therefore dbus requests time out, then syncdaemon goes up completely, says "huh?" and continues working properly10:05
olyubuntuone-clien and ubuntuone-syncd10:05
olyi wonder if it could be todo with my msn problems as well10:05
olysometimes here msn will refuse to connect, i get arround it by sshing to a remote machine and forwarding pidgin from my machine at home10:06
olynever figured out what stops msn from connecting as everything else seems to work with out a hitch10:07
olybut i guess what ever stops msn connecting could be causing the same problem for ubuntu one10:08
duanedesignoly: are you getting the same: STANDOFF_WAITING_WITH_NETWORK_WITH_BOTHQ10:09
duanedesignin your syncdaemon.log ?10:09
olystate: READY_WITH_NETWORK_WITH_BOTHQ; queues: metadata: 54; content: 53; hash: 0, fsm-cache: hit=71823 miss=13760810:10
olycurenttly got that no sign of the other error10:10
ryeoly, this means that you needs to connect via applet10:13
duanedesignpopey: bug 38730810:16
ubottuLaunchpad bug 387308 in ubuntuone-client "[Wishlist] Proxy Support" [Wishlist,Confirmed] https://launchpad.net/bugs/38730810:17
popeyaw shame10:17
popeyi cant even open ubuntuone-preferences10:17
popeyit barfs10:17
popeyi have filed bug 527658 if you believe it's a dupe of 387308, so be it10:27
ubottuLaunchpad bug 527658 in ubuntuone-client "ubuntu-preferences wont start" [Undecided,New] https://launchpad.net/bugs/52765810:27
=== teknico is now known as teknico_away
* rye reboots, suspend broke video :(11:45
=== teknico_away is now known as teknico
vds1Desktop+ MEETING BEGINS aquarius beuno CardinalFang Chipaca dobey jblount teknico urbanape ?15:11
Chipacame15:11
teknicome15:12
beunome15:12
dobeymeh15:12
vds1me15:13
CardinalFangme15:15
urbanapememememe15:16
teknicothe meme of me15:16
dobeyyawn15:21
ChipacaDONE: looking at oauthdesktop and hmm'ing a lot. Also: removing note edition from web until we fix it. TODO: push note edition removal branch. Crack down on oauthdesktop. BLOCKED: no. NEXT: teknico.15:21
teknicoDONE: did more reviews; found a workaround for the funambol DNS/haproxy/exchange problem in production with the L0SAs, and phone sync finally worked! (#511316)15:21
teknicoTODO: fixing more deployment and web ui bugs; showing the web desktop+ guys around our phone sync code15:21
teknicoBLOCK: none15:21
tekniconext: beuno15:21
beunoDONE: Half-did a lot of things. Debugged with funambol support the duplicate checking, started a branch to show more information about your phone in the index15:22
beunoTODO: Continue pursuing dupe prevention, finish and land the phone details branch15:22
beunoBLOCKED: Yes, on funambol15:22
beunonext: dobey15:22
dobey☺ DONE: Removed connect button from Nautilus, Started adding Connect to CP15:23
dobey☹ TODO: Finish branch to add Connect to CP, Finish the rest of the CP work15:23
dobey☹ BLCK: None.15:23
dobeyvds1: roll15:23
vds1DONE: landed one_pwd_generation branch, rolled back patch_13, investigated on contact dups but stopped as local sync is broken, investigated on broken local sync and filed a bug, investigated on broken production sync, reviewed some bugs, started #49832415:23
vds1TODO: finish #498324 re-land patch_13, re-start invetigating on dups as soons as the problem can be reproduced15:23
vds1BLOCKED: no really15:23
vds1CardinalFang please15:23
CardinalFangaw, crap, not ready.  urbanape take my spot.15:24
urbanapeDONE: nailed most of the ajaxy goodness for funambol phone choice15:25
urbanapeTODO: little smoothy bits (spinners, &c) and CSS lovin'15:25
urbanapeBLOCK: None15:25
urbanapeCardinalFang, whenever you're ready15:25
CardinalFangDONE: worked on desktopcouch-service noticing when couchdb is listening on a different port, and change its zeroconf advertisements and replication.  tested get-port branch more.  fixed bug with dbus attempting communication before we are ready, with help from james_w.15:28
CardinalFangTODO: get reviews on get_port branch.  release to kenvandine.15:28
CardinalFangBLOCKED: None15:28
CardinalFangmaybe-TODO: help thisfred and mthaddon diagnose couchdb on server15:29
CardinalFangthat's all from me.15:29
popeyis it possible to start syncdaemon on the command line? I want to run it in foreground15:30
verterokpopey: /usr/lib/ubuntuone-client/ubuntuone-syncdaemon -h15:31
popeyi have looked at that15:31
verterokpopey: if you want debug logs to stdout: /usr/lib/ubuntuone-client/ubuntuone-syncdaemon --debug15:31
popeyok, thats non-obvious15:31
popeyi would have expected a --no-daemon or something15:32
popeythanks15:32
verterokpopey: the daemon itself doesn't fork, it's usually executed in background becasue it's started via dbus activation15:33
popeyI'm trying to run it behind a horrid proxy, and have made a magick network, but can't get syncdaemon to use it15:34
kklimondahmm.. how can I check sync progress in lucid?15:34
popeyooo, made it work15:35
popeyyay15:35
duanedesignkklimonda: u1sdtool --current-transfers15:37
=== aquarius_ is now known as aquarius
kklimondaduanedesign: thanks15:38
=== beuno is now known as beuno-lunch
bollullerahello everybody!17:40
duanedesignhello bollullera17:43
ryere: proxy. It is possible to run syncdaemon behind the proxy, but it requires some configuration changes that are not supported17:51
ryebut it is not possible to run desktopcouch with proxy for now, I haven't yet figured out how to make that work properly. It will definitely require source code changes17:52
ryeok17:52
ryeme is back and full of energy to do things17:52
* rye is going to vm fullscreen (I have only one display :( ) to test note sync XML/HTML issue again17:54
* CardinalFang tries landing a bunch of desktopcouch branches.18:05
CardinalFangmandel is a mad-man!18:06
=== beuno-lunch is now known as beuno
CardinalFangmandel!  Hi.  I was landing code and I have a complaint about one of your patches.  I want to make sure I'm sane.18:22
mandelCardinalFang, hehe I've noticed18:22
mandelCardinalFang, let me take a look18:23
CardinalFangmandel, thisfred, aquarius, re-review  https://code.edge.launchpad.net/~mandel/desktopcouch/fix_bug_517676/+merge/18708 ?  Previously approved.18:23
thisfredCardinalFang: ouch, I totally got sidetracked on your other review too18:23
thisfredCardinalFang: I think it may contain the _rev, actually. Admittedly that's because python-couchdb does something VERY unexpected, which exploded my brain elsewhere18:26
thisfredbasically:         self.db[record.record_id] = record._data18:26
thisfredwill modify record._data to contain the new _rev18:27
mandelthisfred, CardinalFang, indeed python couchdb will do that and that is why I filled the bug, it would be stupid to retrieve the code gain18:29
CardinalFangI am going to verify that.  If it does, put_record has been wrong forever and at least two other parts of d-c can be cleaned up.18:29
thisfredCardinalFang: that may well be, as I only learned of this behavior when it broke something on the server18:30
mandelCardinalFang, I'll take a look too, just o make sure I did not understood python-couch wrong18:30
thisfredmeanwhile reviewing that other branch18:30
* CardinalFang prepares to eat his hat.18:30
* thisfred hands CardinalFang salt and pepper shakers18:30
CardinalFangI suspect it's salty enough.18:31
mandel:D18:31
mandellook at line 323 at client.py18:31
mandelthisfred, CardinalFang, that is updating the _id and the _rev, right?18:32
CardinalFangIndeed.18:33
CardinalFangSide-effects on __setitem__.  I'd puke if it weren't so useful.18:34
CardinalFangOkay, I withdraw my second complaint.18:34
mandelCardinalFang, hehe I just discovered by chance18:34
mandelCardinalFang, give thisfred and mandel a complement in the comment ;)18:35
CardinalFangFirst complaint is that we've documented this behavior for a while.18:35
mandelIndeed but asking people to perform extra requests when they do not need to is very ugly18:35
thisfredCardinalFang: yes, but I can't think of a backward compatible way to fix it.18:35
thisfredand I think we do want to fix it asap18:36
thisfredrather than lucid + x18:36
thisfredI don't know if it merits a freeze exception18:36
CardinalFangI think we need another method.  put_record_and_update()  ?18:37
mandelCardinalFang, doing the same but returning the record??18:38
mandelCardinalFang, from a coding point of view I'd complain, on the other hand for backward compatibility... which apps will be affected? gwibber?18:39
thisfredCardinalFang: note that if someone passes a record into this one and keeps reference to it, it *already* updates it18:40
CardinalFangLiterally nothing that uses put_record() currently will continue to work.  We have to get the new record_id to do anything currently.  Everyone is using that.18:40
CardinalFangthisfred, Yeah, I was just noticing that.18:40
thisfredand always has done so18:41
thisfredSo, ok, I can live with moving this into a new method, I think that's a compromise that will keep everyone mildly unhappy18:41
thisfredmandel: ?18:41
thisfred(I have to go walk the dog now, bbiab)18:42
mandelthisfred, agreed, move it to a new method to keep backwards comp but put a warning in the old one18:42
mandelat least people will know there is a "better" way to do it18:42
mandelCardinalFang: ?18:42
CardinalFangA note in the comments, and changing the docs, but no programmatic warning yet.  Maybe next version.18:44
mandelCardinalFang, ok, but I'll keep an eye on that bug since is a pain in the ass in some cases18:45
aquarius*nod* No breaking backwards compatibility. Good decision, CardinalFang/thisfred18:46
aquarius(sorry I've not been available to be part of that discussion)18:47
mandelblah, I've been outnumbered :P18:48
mandelCardinalFang, aquarius, thisfred, what do you think about the following: http://pastebin.ca/181053119:05
mandelI've got something similar in my code and other people might find it useful19:06
aquariusthat's an implementation of schemas. I don't like schemas :(19:06
aquariusI certainly can see how it's useful, I just don't like the idea of baking support for it into desktopcouch, because it's against the philosophy. But I'm happy to be talked out of it :)19:06
mandelaquaris, I knew your were going to do that...19:06
aquariusmandel, well, yeah :)19:06
aquariuswe're at opposite ends of the spectrum on this ;)19:06
mandelaquarius, if you bring up AC/DC I'll go off line ;)19:07
mandelaquarius, anyway, I'm using it because is a pain in the ass to be accessing application annotations all the time with by code... ofcourse it has a problem, you need to provide the app name19:08
aquarius*nod*19:09
aquariusthat's the same thing as, say, calling D-Bus, though; you need to do it there, too.19:10
aquariusIf you wanted to be particularly magic, you could default appname to the actual name of the executable19:10
mandelindeed, but providing the app_name is certainly better19:11
ryereproduces note misbehavior, now with complete STR19:11
mandelaquarius, I'll put the code in a post out there in case people want to use it19:13
aquariusmandel, yeah, definitely -- I've certainly got no problem with someone (you?) building a wrapper library on top of DC which offers things like StrictRecord. People are likely to find that useful19:13
aquariusand that's a good way to show that I'm wrong and it should be in DC core, too -- if loads of people find the library useful, then it's clearly something we should think about including. that's what happened with python and urllib219:14
mandelaquarius, ok, you just convinced me, I'll put it somewhere as an add-on or something19:15
aquariusmandel, you ge tthe satisfaction of laughing at me when I'm wrong, which is not to be underestimated :)19:15
aquariusmandel, feel free to edit the desktopcouch documentation to point at docs for your wrapper library, too19:16
mandelaquirius, I'll think about it, but I doubt I'll point to my code from here without your agreement, I hate self-promotion :P19:18
aquariusmandel, you have my agreement :)19:19
aquarius__lucio__, *nice* work on the windows port19:19
__lucio__aquarius, thanks. its still missing some stuff: filesystem monitoring, dbus replacement and proper testing19:20
__lucio__but at least it downloads stuff19:20
__lucio__ill try to propose this as soon as i clean up the tests19:20
aquarius__lucio__, that's amazing. I almost wish I ran windows just so I could try it :)19:20
__lucio__aquarius, no, you dont :)19:21
aquariuswell, OK, that was a white lie, I admit it. I like Ubuntu :-)19:21
till__lucio__: not sure if you know, but dbus master works fine on Windows19:22
__lucio__till, but thats not very "windows", is it?19:22
tillWell, depends what your goal is.19:22
__lucio__till, im afraid that trying to use dbus from a file explorer extension may be a mess19:23
tillPossibly, yes, you can probaby judge that better than me, just thought I'd point out that it does, in fact, work fine :)19:24
thisfredmandel: re: schema's I'm with aquarius mostly. I'd use the mappings to transform from and to application data, and add more mappings where needed19:25
thisfredmandel: you know that the fieldmappings already put unrecognized fields in the correct application annonations namespace, right?19:25
mandelthisfred, nope, I did not know about those, where are they?19:26
thisfredmandel: look at the doctest in desktopcouch/doc/19:26
mandelthisfred, I remember now, you show them to me before, I completely forgot about them19:29
=== sandy__ is now known as sandy|lurk
thisfredmandel: they're basic, but they're in use on the server, and I'm happy to help expand them with whatever client developers need19:40
CardinalFangmandel, thisfred, aquarius, I think this is sufficient.  https://code.edge.launchpad.net/~cmiller/desktopcouch/put_record_documented_more/+merge/2016619:41
thisfredand they impose nothing on the client application, nor any new restrictions on the record format19:41
mandelthisfred, I'm actually thinking of using them in a special thing I'm writing right now, let me a couple of days and I'll bother you about them19:41
thisfredmandel: ok, great! :)19:42
ryeok. notes are definitely not converted during sync. Me is extremely sure and has performed a lot of time-consuming tests.19:47
ryeaquarius, is the method planned for dc to have replication done on dbus request, in order not to wait for 10 minutes for the miracle to happen?19:47
ryejoshuahoover, just verified note conversion during sync in tomboy - it is not happening. HTML nodes are not converted to XML if they are stored in HTML on server side19:49
aquariusrye, not sure -- CardinalFang will know19:49
joshuahooverrye: and you were able to do this following your steps for bug #527335 ?19:50
ubottuLaunchpad bug 527335 in ubuntuone-servers "Tomboy notes are not converted to XML on Tomboy sync" [High,New] https://launchpad.net/bugs/52733519:50
ryejoshuahoover, i have added a document and a link to pastebin that was used for testing19:50
CardinalFangrye, it will spawn replication after changes only, using the _changes interface.  Not very soon.19:51
CardinalFangIn Murderous Marmot or whaterver19:51
joshuahooverrye: cool...if these are the steps to reproduce, can you update the description so that it's clear these are the best steps to use to try to reproduce the bug?19:52
ryejoshuahoover, updating...19:52
joshuahooverrye: thanks!19:52
=== teknico is now known as teknico_away
ryejoshuahoover, update finished19:58
joshuahooverrye: perfect! thank you!19:59
ryeDid you know that you can use Bug:12345 format to specify a link to LP bug at wiki.ubuntu.com? No need to remember http://bugs.launchpad.net/bugs/12345 !20:00
ubottuLaunchpad bug 12345 in isdnutils "isdn does not work, fritz avm (pnp?)" [Medium,Fix released]20:00
* rye needs to pick a better random number next time20:00
CardinalFang1 is good20:00
CardinalFangBug:120:01
ryeCardinalFang, 1 is not that random.20:01
=== ersoy is now known as hersoy
CardinalFangNo specific number is random.20:01
* rye will use random bug number generator20:02
=== cpg is now known as cpg|brb
mandelCardinalFang, I'll take a look at the merge problems20:29
ryehm20:33
=== cpg|brb is now known as cpg
mandelCardinalFang, when a merge is not longer needed is it better to delete or to reject?20:46
CardinalFangSuperceded, I think.20:46
CardinalFangI don't know if that's an option.20:46
mandelCardinalFang, I don have the right, can you do that for: https://code.launchpad.net/~mandel/desktopcouch/fix_bug_519922/+merge/1902920:47
sanderqdjust to check: ubuntuone doesn't store the ~/Ubuntu One files in a couchdb database, right?20:49
CardinalFangNo, sanderqd.20:49
sanderqdok, thanks20:49
CardinalFangcouchdb synch and file synch are separate.20:49
mandelCardinalFang, fixed https://code.edge.launchpad.net/~mandel/desktopcouch/batch_update/+merge/17482 it should work ok now21:17
CardinalFangmandel, thanks!21:18
mandelCardinalFang, as soon as that is merged all fix fix_bug_519873/+merge/1901821:18
=== mandel_macaque is now known as mandel
espen77i am not able to connect to ubuntuone for some reason, have a fresh lucid a3 and neither files or tomboy can sync, but i can register the computer and it shows how much space on u1 used, any ideas?22:37
mandelCardinalFang, before I go, I revised https://code.edge.launchpad.net/~mandel/desktopcouch/fix_bug_519873/+merge/19018 as shoul dbe ok now, but since I use lucid I get the fails reported by John, good night22:40
=== wojtek_ is now known as wojtek

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!