[11:55] <duanedesign> morning all
[14:02] <mahen> jdobrien: hi :) It seems there is a fix regarding my bug report :) Please tell me if you want me to check whether it works or not (if that's available in a ppa or proposed repo maybe ? :)
[14:03] <jdobrien> mahen, what do you mean a fix?
[14:04] <jdobrien> mahen, I started a branch which will fix it when it's finished
[14:04] <mahen> well I don't know, I just saw this "** Branch linked: lp:~jdobrien/ubuntuone-client/fix-666942 " added to my bugreport, I assumed it was one fix :)
[14:04] <mahen> aah :)
[14:04] <mahen> Okay !
[14:04] <mahen> Sorry :)
[14:04] <mahen> Well, thanks for your work then :)
[14:04] <jdobrien> mahen, we found the issue..it's going to take longer for me to write tests for it than to fix it...as usual :)
[14:07] <mahen> jdobrien: okay :) Sorry for giving you work fixing stuff that will, if I understood correctly, be dropped anyway for 11.04
[14:07] <jdobrien> mahen, what will be dropped 11.04?
[14:08] <jdobrien> mahen, this is a real bug in the code actually :)
[14:08] <mahen> I thought the whole sync daemon was being rewritten in a completely different way to allow things like resume, better reliability/performance etc.
[14:08] <mahen> But I just caught some words here & there :)
[14:13] <rye> mahen, from performance point of view I can say that since it is highly modular the bits can be changed w/o rewriting all other code.
[14:14] <mahen> rye : good to know
[14:14] <duanedesign> hello all
[14:14] <rye> duanedesign, morning!
[14:15] <duanedesign> good day rye. Hope you had a nice weekend
[14:18] <mahen> thanks everyone & see you
[14:19] <rye> duanedesign, yes, pretty much! I can now be contacted about music store issues too :-). Played with moovida, ushare, gmediacenter and mediatomb, set up the network properly... well, a normal weekend you know :)
[14:19] <duanedesign> rye: do you know if their is an existing bug report about the Emblems in Lucid not refreshing automagically.
[14:20] <rye> duanedesign, in lucid?... let me have a look at the code
[14:20] <duanedesign> rye: the user is using the stable/PPA
[14:21] <duanedesign> bug 671380
[14:21] <ubot4> Launchpad bug 671380 in ubuntuone-client "Nautilus checkmarks not refreshing automatically (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/671380
[14:26] <rye> duanedesign, uh-huh, i understand the issue now...
[14:27] <rye> duanedesign, need to check how nautilus reacts when we .. hmmm
[14:29] <duanedesign> yeah i did not have a VM of lucid haandy to confirm it...
[14:29] <duanedesign> probably need to get one set-up
[14:30] <rye> duanedesign, confirming in maverick, moved file does not get any emblem (well, it just bears the same set of emblems it used to)
[14:32] <rye> duanedesign, it's like it does not have the info about this file :-/, ok, getting the source of ubuntuone-cilent-gnome and a hammer
[14:33] <duanedesign> :)
[16:05] <hallyn> jinkeys, my first:  next_actions.otl.u1conflict
[16:07] <duanedesign> ruh roh
[16:08] <hallyn> well one thing i finally accidentally confirmed:  ecryptfs won't write out the lower files until you umount - so i left for coffee shop with netbook, and ubuntuone had my old files :)
[19:18] <jdobrien> mahen, ping
[19:44] <mahen> jdobrien: yep ?
[19:45] <jdobrien> mahen, did you cancel your subscription?
[19:45] <jdobrien> mahen, I just noticed that it wasn't active
[19:46] <mahen> jdobrien: hm ? I thought I just downgraded it to 20 GB temporarily actually
[19:46] <jdobrien> mahen, let me check
[19:46] <mahen> as soon as it works properly I'll upgrade again of course :)
[19:46] <jdobrien> mahen, did someone do that for you?
[19:47] <mahen> jdobrien: no. Hmm, something definitely changed since last time I visited the one.ubuntu.com/account page
[19:47] <mahen> indeed it seems to be cancelled now
[19:47] <mahen> it wans't last time I tried the sync daemon
[19:47] <mahen> wasn't
[19:47] <jdobrien> when is the last time you checked?
[19:48] <jdobrien> mahen, ok im going to fix it...don't worry you won't get charged, but you'll get a silly welcome email
[19:48] <mahen> I'm not sure, maybe it was last wednesday (date of my latest debug log)
[19:48] <jdobrien> oh ok
[19:49] <mahen> jdobrien: no problem :)
[19:49] <mahen> thanks
[19:50] <jdobrien> mahen, ok...you should work now
[19:50] <tfi> Hi everybody ;-) . Don't know if this is the right place.. I'm a new Ubuntu One user and trying to create an account there.. I'm waiting for 10 minutes now for the confirmation code but no one arrives yet.. Is there a problem or should i simply wait?
[19:51] <jdobrien> mahen, I would like to find out what's going on with that Move retry thing
[19:51] <jdobrien> mahen, it's bewildering
[19:51] <mahen> jdobrien: OK, I have some spare time :)
[19:51] <mahen> please tell me what to do :)
[19:52] <jdobrien> mahen, is your client connected?
[19:52] <mahen> not yet
[19:52] <mahen> shall I ?
[19:52] <jdobrien> yeah
[19:52] <mahen> ok, it'll take a couple of minute (drive spinning)
[19:52] <mahen> +s
[19:52] <jdobrien> ok
[19:53] <mahen> (which is another issue I have btw, but I'll bother you later about this ;)
[19:55] <mahen> jdobrien: online
[19:55] <jdobrien> mahen, are you seeing exceptions?
[19:56] <mahen> yes (I have debug enabled also)
[19:57] <jdobrien> mahen, can you paste the exceptions?
[19:58] <jdobrien> mahen, hello?
[19:59] <mahen> yep
[19:59] <mahen> http://paste.ubuntu.com/528324/
[20:50] <toml_> echo
[20:50] <toml_> honk
[20:51] <rye> toml_, hi
[20:51] <toml_> hey rye
[20:51] <toml_> got a q that's not covered in the FAQ
[20:51] <rye> toml_, okay, i am listening
[20:51] <toml_> cool
[20:52] <toml_> two folders synced into Ubuntu One
[20:52] <toml_> across two computers
[20:52] <toml_> folders A and B
[20:52] <toml_> both are synced
[20:52] <toml_> let's add a third computer
[20:52] <toml_> can I just sync folder A to the third computer, and not folder B?
[20:53] <toml_> ie. so folder A synced across 3 computers, folder B synced across only two
[20:53] <rye> toml_, yes, but by default it will subscribe the folder so that it will start syncing
[20:53] <toml_> so how to override the default behaviour?
[20:53] <rye> toml_, config file
[20:53] <toml_> path?
[20:53] <rye> udf_autosubscribe = False in [__main__] section of ~/.config/ubuntuone/syncdaemon.conf
[20:54] <toml_> on the third computer?
[20:54] <rye> toml_, all available options are listed in /etc/xdg/ubuntuone/syncdaemon.conf
[20:54] <toml_> rye, k let me have a skinnt at it
[20:55] <toml_> rye, thanks
[20:55] <rye> toml_, yes, on the third computer will be not synchronizing anything (but ~/Ubuntu One/ folder) by default. To force subscription - u1sdtool --subscribe=$folder_uuid, which can be found by u1sdtool --list-folders
[20:56] <toml_> rye, so unset the default subscription, then on the third computer force sub to the folder with the uuid you want
[20:57] <rye> toml_, yes, by the way, what version of Ubuntu are you runnign?
[20:58] <toml_> rye, 10.04 netbook remix on 2 x boxes, and 10.10 on a third
[20:59] <rye> toml_, 10.10 has this option, while i am not sure about 10.04, but if third machine won't subscribe then it won't sync to it for sure
[20:59] <toml_> rye checking it now
[21:01] <toml_> rye, 10.4 Netbook has this option
[21:07] <rye> toml_, ok, then it should be ok
[21:07] <toml_> rye just checking it out now; thanks loads for the help with this.
[21:08] <rye> toml_, you are very welcome!
[21:27] <toml_> rye, slight problem
[21:27] <rye> toml_, yup?
[21:28] <toml_> rye, okay, on 10.4 possible to edit syncdaemon.conf and disable automatic subscription; however, after adding computer to my account
[21:28] <toml_> rye, it still automatically subscribes
[21:29] <toml_> rye, when i run the --unsubscribe-folder option, it ignores it
[21:29] <rye> toml_, run u1sdtool --disconnect
[21:29] <rye> toml_, in order for it to stop syncing
[21:29] <toml_> rye, -same as -d ?
[21:29] <rye> toml_, hm, let me double-check
[21:29] <rye> toml_, yes
[21:29] <toml_> rye, thanks
[21:31] <rye> toml_, wow, there's even a dbus method to disable udf autosubscribe in 10.10
[21:32] <toml_> rye, sounds... scary
[21:33] <toml_> rye,  ahaha I fecked up ... changed the wrong option in syncdaemon.conf
[21:34] <toml_> rye, will go back over it again ... think it may have been an error existing between brain and computer
[21:35] <rye> toml_, in order to prevent syncing of everything - just restart syncdaemon - u1sdtool --quit; u1sdtool --start, verify that udf is not subscribed and then --connect
[21:35] <toml_> rye, right
[21:36] <toml_> rye, hmm still subscribed to the folder
[21:36] <rye> toml_, before connecting - unsubscribe it
[21:39] <toml_> rye, nah,... -folder-unsubscribe=<id> isn't responding
[21:39] <toml_> rye, so I quit, start, list folders, try to unsub from the ID I want to. then quit, start, list... and it's still there
[21:39] <rye> toml_, u1sdtool --unsubscribe-folder=aaabbbccc... ?
[21:40] <toml_> rye, yup
[21:40] <rye> toml_, is u1stool --list-folders shows subscribed=False ?
[21:40] <toml_> rye, standby
[21:40] <rye> toml_, for the folder which you don't want to be sync locally
[21:41] <toml_> rye, no, actually it doesn't have a flag: it's id=asdsadafafas subscribed= path=/home/etcetcetec
[21:41] <toml_> rye, but it doesn't say =true
[21:41] <rye> toml_, ok, that's 10.04, False was not properly printed then
[21:42] <toml_> rye, okay, so I can safely delete the initial file structure that it laid down on the local machine before I told it not to sync?
[21:42] <rye> toml_, yes, to test - you can try keeping sd disconnected, remove the files/folders and then run u1sdtool --waiting-meta
[21:43] <toml_> rye, and that will then tell me whether it is will try to affect changes to that folder when i reconnect to the cloud?
[21:43] <rye> toml_, yes
[21:44] <toml_> rye, testing
[21:44] <toml_> rye, --waiting-meta ... nothing reported
[21:45] <rye> toml_, u1sdtool --status = ?
[21:46] <toml_> rye, connection: Not User With Network \ decription: ready to connect \ is_connected: false \ is_error: false \ is_online: false \ queues: IDLE
[21:47] <rye> toml_, awesome!, feel free to connect
[21:47] <toml_> rye, ta for walking me through it rye ... really appreciate the support
[21:47] <mahen> good night everyone ! thanks again jdobrien !
[21:47] <toml_> rye, also a good crash course in the tool ;)
[21:49] <rye> toml_, you are welcome!
[21:49] <toml_> rye, are you a dev for this tool, or just volunteering here?
[21:49] <rye> toml_, I am more like customer support
[21:49] <toml_> rye, aha, for canonical?
[21:50] <rye> toml_, canonical/ubuntuone-specific
[21:50] <jdobrien> rye, don't give away all our secrets on one day
[21:50] <toml_> rye, well, good on you man - clear instructions, really observant of what I was doing. Nice one.
[21:51] <jdobrien> rye, finally figured out the real problem behind the moves
[21:51] <rye> jdobrien, /etc/xdg/ubuntuone/syncdaemon.conf is full of them, unencrypted
[21:51] <rye> jdobrien, aaannd??? I am all ears!!!!
[21:51] <jdobrien> rye, :)
[21:51] <rye> well
[21:51] <rye> eyes
[21:51] <jdobrien> rye, well...of you disconnect the client
[21:52] <jdobrien> rye, then mv <oldname> <newname> && mv <newname> <oldname>
[21:52] <jdobrien> rye, on a huge directory...
[21:52] <jdobrien> rye, the stupid server tries to do it
[21:52] <jdobrien> rye, more specifically...my stupid code
[21:52] <jdobrien> rye, or...stupid me :)
[21:53] <rye> jdobrien, erm, okay, mv oldname newname - Move() oldname -> newname... and another Move() to bring newname into oldname... that's only one dir level change... right? node id is kept...
[21:54] <jdobrien> rye, bascially it results in a move that does nothing
[21:54] <jdobrien> rye, do it with the client disconnected to a large directory
[21:54] <jdobrien> rye, then reconnect the client and you're stuck
[21:55] <rye> jdobrien, but why large directory is a must?
[21:57] <jdobrien> rye, because the server tries to do a lot of work on moves... we're going to work on fixing that this cycle as well
[21:58] <jdobrien> rye, when you move a large directory structure, there are path renames and we need to check for shares etc....it's huge
[21:58] <rye> jdobrien, so TRY_AGAIN comes from db layer?
[21:58] <jdobrien> rye, usually yes
[21:59] <rye> jdobrien, so the query times out, then TRY_AGAIN is issued to the client which calls the server to perform the same operation which is locked which calls TRY_AGAIN and so on?
[21:59] <jdobrien> rye, it's not a locking issue as I once thought
[21:59] <jdobrien> rye, but you're close
[22:01] <rye> jdobrien, wow