[06:20] <kermiac> ty for the link regarding the new bug workflow joshuahoover :)
[10:41] <rye> evolution-data-server 188% CPU O_O
[10:42] <duanedesign> lol
[10:49] <duanedesign> rye: gonna run down the street and get some breakfast. If you get a minute could you look at http://paste.ubuntu.com/392426/
[10:49] <duanedesign> its the UseSettings() function that doesnt seem to return the DB. IF your busy no problem dont worry about it :)
[10:49] <rye> duanedesign, package unifier ?
[10:50] <rye> i mean make the same set of packages available on different sets of machines?
[10:52] <duanedesign> yes
[10:52] <rye> hm, did not create db for me
[10:52] <duanedesign> really:(
[10:55] <duanedesign> rye: it didnt create a DB called packages
[10:57] <rye> duanedesign, hm... something is really wrong with my futon
[10:57] <rye> duanedesign, now I have the packages db :-/
[10:58] <duanedesign> rye: you had me nervous,lol
[10:58] <rye> false alarm, package db is created... I believe my 'rtg+test, rtg$test and rtg/test' databases made futon a little bit worried about its future
[10:58] <duanedesign> :)
[10:59] <rye> duanedesign, your 'installed' view does not generate any documents - try it in futon
[11:00] <rye> function(doc) { emit(null,doc.packages) }
[11:00] <rye> try this
[11:01] <duanedesign> how do i try it in futon?
[11:03] <duanedesign> i tried it with curl and it didnt get anything
[11:03] <duanedesign> oh try the one you posted
[11:04] <rye> duanedesign, the futon has a UI to run the views - it is the select box in upper right corner
[11:04] <duanedesign> thats neat, will look
[11:09] <duanedesign> rye: lol, i never saw that function in Futon. That is awesome
[11:10] <rye> duanedesign, imagine me panicking when I switched to some view w/o realizing what that does and stopped being able to browse the records
[11:10] <rye> duanedesign, that's how I learned about it :)
[11:11] <duanedesign> ;)
[11:14] <rye> evolution-data-server-Message: Server up and running
[11:14] <rye> [New Thread 0xb7f8fb70 (LWP 4372)]
[11:14] <rye> impl_GNOME_Evolution_Addressbook_BookFactory_getBook
[11:14] <rye>  + couchdb://127.0.0.1
[11:14] <rye>  => 0x807a660
[11:14] <rye> impl_GNOME_Evolution_Addressbook_Book_open (0x807a660)
[11:14] <rye> Program received signal SIGSEGV, Segmentation fault.
[11:14] <rye> [Switching to Thread 0xb7f8fb70 (LWP 4372)]
[11:14] <rye> 0x00f01a96 in ?? () from /lib/libdbus-1.so.3
[11:14] <rye> ARGHHHHH!
[11:21] <rye> YES. bug #534469 - I am not alone now
[11:22] <rye> when was separating of /usr /var, /opt, etc... partition obsoleted? I have been happily running it on this machine since intrepid ibex and that was actually working :)
[11:22]  * rye switches to thunderbird as my e-d-s misbehaves today
[11:24]  * wgrant points U1 devs in the direction of https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/535651
[11:30]  * rye points wgrant to https://bugs.launchpad.net/ubuntuone-servers/+bug/535137
[11:31] <wgrant> Ah, that one is a dupe.
[11:32] <rye> wgrant, this is not a security vulnerability for now, since the notes cannot be shared between the users
[11:32] <wgrant> rye: Not directly, no.
[11:32] <rye> wgrant, so the worst that could happen is that the user breaks his own notes UI
[11:32] <wgrant> But see the bug I just filed.
[11:33] <wgrant> Leaving XSSes around is a seriously bad idea.
[11:33] <wgrant> Because they will come back and bite you when another vulnerability appears.
[11:34] <rye> wgrant, ideally this is what should happen when notes are entered via the web ui - all the tags that are not translated to tomboy xml subset are removed and since tomboy does not use attributes - they will be ignored
[11:35] <wgrant> rye: Yes -- I reported a bug to that effect 6 months ago.
[11:35] <rye> wgrant, however, at this time there is a bug #535137 which causes the notes that are properly escaped in the database not to be escaped properly via the web ui
[11:36] <rye> wgrant, let me try one trick...
[11:37] <wgrant> lolwut?
[11:37] <wgrant> Why is my bug a dupe of 535137?
[11:38] <rye> wgrant, because you were able to insert the html content and it was not properly escaped when it was displayed
[11:38] <wgrant> My new bug references my old bug which is related to 535137.
[11:38] <wgrant> But neither is a dupe.
[11:38] <rye> wgrant, ok, I might have acted too quick
[11:39] <rye> wgrant, so, you were able to inject the HTML code via the web editor and later on that code was sent to the browser unmodified, right?
[11:40] <wgrant> rye: Yes, but that is 452689, not 535651
[11:40] <rye> wgrant, I removed the dup status, sorry about that.
[11:41] <wgrant> rye: Thanks.
[11:41] <rye> wgrant, so, you mean the notes that are entered in the web UI are left as HTML?
[11:41] <wgrant> rye: The U1 web UI will render arbitrary HTML attributes that I give it.
[11:41] <wgrant> That is a bug.
[11:41] <rye> wgrant, what????
[11:42] <wgrant> rye: Have you read 452689?
[11:43] <rye> wgrant, ok, testing this on trunk. If that appears to be valid after all that was done to Notes during last 3 weeks then I will start killing one my note at a time until it is fixed
[11:43] <rye> wgrant, there is an explicit conversion for the notes format, if that is now broken then I am very mad
[11:43] <wgrant> rye: Ah, I haven't tested on edge.
[11:44]  * wgrant tests on edge.
[11:44] <rye> wgrant, no, that was already in production for a week or so
[11:44] <wgrant> rye: I tested it again about an hour ago.
[11:44] <rye> awesome
[11:44] <wgrant> It has been declared fixed something like three times now.
[11:44] <wgrant> Not once has it actually been fixed.
[11:44] <wgrant> And now I have found something else that piggy-backs on it.
[11:44] <wgrant> I am not amused.
[11:46] <rye> wgrant, there is a whitelist for conversion now...
[11:46] <rye> of course, if that works
[11:48] <wgrant> rye: Still broken on edge.
[11:48] <wgrant> Your whitelist does not work.
[11:49]  * rye checks deeper
[11:51] <rye> wgrant, yes, that is bug 535137 - the data is stored properly but it is not properly escaped on display
[11:51] <wgrant> I am confused.
[11:51] <wgrant> This isn't an escaping issue.
[11:51] <wgrant> The tags have to be there, to represent boldness for example.
[11:51] <wgrant> The bug is that it allows extra attributes.
[11:52] <wgrant> Some tags are already stripped.
[11:52] <wgrant> Bug #535651 is much more important.
[11:58] <rye> wgrant, were you creating the notes via the web ui or tomboy?
[11:58] <wgrant> rye: Web UI. The Tomboy sync crashed.
[12:00] <rye> wgrant, great
[12:00] <rye> wgrant, attributes are placed as is
[12:00] <rye> asesome
[12:00] <rye> desktop+, rodrigo_  - BUG!
[12:01] <wgrant> Yes, I talked to rodrigo_ about it earlier.
[12:01] <rodrigo_> yeah, already looking at it
[12:01] <wgrant> Thanks.
[12:03] <rye> wgrant, you see - the whitelist is there and it will happily replace <script> with nothing, since it does not know this item. But it will also make <b onclick="alert('test');">wow</b> into <bold onclick="alert('test');">wow</bold> for couchdb and the reverse conversion will restore <b> element with onclick attribute
[12:03] <wgrant> rye: Right.
[12:04] <rye> but... Tomboy does not use attributes...
[12:14] <rye> wgrant, and this is made extremely easy with bug 535137 - you don't even need to use the firebug for that...
[12:15] <duanedesign> thats an interesting NVIDIA bug report rye
[12:19] <rye> duanedesign, and this will affect people starting from -16 kernel where  nuve... grrr. nouveau module is available in the kernel
[12:21] <nperry> Where has ubuntuone-client-applet gone in lucid :(
[12:27] <rye> nperry, ubuntuone client applet is removed from lucid. The only UI that will be present is ubuntuone control panel
[12:27] <rye> nperry, I plan to write about that on the wiki once I find out how the actual startup will be handled
[12:27] <nperry> How do I sync/connect now then
[12:28] <nperry> Prefenances have all my account info there
[12:28] <rye> nperry, you can open ubuntuone-preferences via MeMenu
[12:29] <nperry> rye, File sync is ticked - But is greyed out, normal?
[12:29] <Chipaca> nperry: or you can run "u1sdtool -c" from a terminal, if that's where you roam
[12:29] <rye> nperry, yes, by default it is enabled
[12:29] <rye> Chipaca, I need to find someone who will tell how the startup will be handled :-)
[12:30] <Chipaca> rye: it won't be handled for this beta
[12:30] <rye> Chipaca, are there any documents I have probably missing that tell how that would be handled in the release?
[12:30] <Chipaca> rye: by next beta, syncdaemon will start and connect on login if you have network and have file sync ticked and have it set to start on login
[12:31] <Chipaca> rye: *however*
[12:31] <Chipaca> rye: you need to stop thinking of the syncdaemon as synonymous with ubuntu one :)
[12:32] <Chipaca> rye: it's never been the case, and while in the past the confusion was understandable (because desktopcouch, the only other part of ubuntu one, was very much behind the scenes), that is no longer the case
[12:32] <Chipaca> i.e. desktopcouch is much more prominent, and it is no longer the only other part of Ubuntu One
[12:32] <wgrant> (and the service was announced solely as a filesharing service)
[12:32] <Chipaca> wgrant: nope, it wasn't
[12:32] <rye> Chipaca, then I will need to think that bandwidth preferences are applied to couchdb replication and they are not, couchdb replication happens even if the Devices tab has the device not connected. So two of 3 tabs are applied to syncdaemon
[12:33] <Chipaca> wgrant: it was picked up by the media as such, unfortunately
[12:33] <wgrant> Chipaca: I never saw any references to concrete further extension ideas.
[12:33] <Chipaca> rye: you're looking at an *old* ubuntuone-preferences :)
[12:34] <rye> Chipaca, how old? I was using trunk... yesterday, there were more devices listed there ... and my phone that became a victim of a private bug :).
[12:34] <Chipaca> wgrant: where were you looking?
[12:35] <wgrant> Chipaca: I'm pretty sure I looked just about everywhere in the day or two after the announcement.
[12:36] <Chipaca> rye: ubuntuone-preferences has three tabs: the 'account' tab which is common to all services, the 'devices' tab which is mostly common to all services but has some controls specific to ubuntuone-syncdaemon, and the 'services' tab which lets you turn on/off services
[12:36] <rye> true
[12:37] <Chipaca> rye: granted we still need to fix the devices tab so the controls specific to syncdaemon are more prominently so, but we haven't found an effective way of saying this to the lay user
[12:37] <rye> Chipaca, btw, is scrolling planned for the devices tab - with my 10 devices it was not fitting the screen
[12:38] <rye> brrr... what we have started with?
[12:39] <rye> ah, will desktopcouch replication be controlled via ubuntuone preferences (i.e. connect/disconnect) ?
[12:39]  * wgrant has had to firewall desktop-couch.
[12:39] <wgrant> It was continuously syncing gwibber_messages, eating lots of bandwidth.
[12:40] <rye> i believe if I move to mobile service with per-mb fee then I will need to shut down dc completely...
[12:40] <rye> wgrant, it is possible to omit gwibber_messages from replication
[12:40] <wgrant> rye: Ooh. How?
[12:40] <rye> wgrant, let me double check that
[12:41] <Chipaca> wgrant: yes, we'll be shipping gwibber with syncing off by default :)
[12:42] <wgrant> Chipaca: Syncing accounts is OK, but messages is a bit strange.
[12:42] <rye> wgrant, http://paste.ubuntu.com/392499/
[12:43] <wgrant> rye: Thanks!
[12:43] <rye> Chipaca, will that be via management db/excluded-names?
[12:43] <wgrant> And off-peak is aqbout to start, so I can safely turn it back on for a while.
[12:43]  * wgrant tries.
[12:43] <rye> wgrant, I haven't tested that, going to do that now
[12:43] <Chipaca> rye: hey, I'm a manager, I don't know technical mumbo-jumbo
[12:43]  * Chipaca lies!
[12:43] <Chipaca> rye: I don't know, actually. I know we need to fix the services tab so it works (there's a branch somewhere for this)
[12:44] <Chipaca> rye: dobey is/was working on that, but he's under a lot of pressure from too many fronts, so I might need to ask for a freeze exception so we don't lose him :)
[12:46] <rye> wgrant, it is excluded_names, not excluded-names
[12:49] <rye> wgrant, you might also want to remove the database from the server
[12:49] <rye> wgrant, http://www.freedesktop.org/wiki/Specifications/desktopcouch/Documentation/Troubleshooting
[12:49] <rye> wgrant, ubuntuone-couchdb-query --http-method=DELETE 'gwibber_messages'
[12:50] <rye> i did that, waiting to see what happens :)
[12:50] <rye> wgrant, from u1 server
[12:50] <wgrant> rye: Ah, I was wondering how to manipulate the server DB.
[12:50] <wgrant> Hm, I do not have such a command.
[12:51] <rye> wgrant, the Troubleshooting page has that command
[12:51] <rye> *the script
[12:52] <rye> something is wrong with my "thinking-typing" connection today
[12:55] <wgrant> rye: Oh, oops, missed that.
[13:58] <urbanape> morning, oh wild and wooly ubuntuonians
[14:24] <dobey> eh, gwibber really should be storing credentials in the keyring, not in couchdb
[14:34] <kklimonda> yup
[14:38] <rye> dobey, i wonder what happens when 2 gwibber instances are running at the same time - do messages get duplicated?
[14:39] <dobey> rye: they shouldn't, but they might i guess, no idea. i'm not using it because i don't want my passwords stored in couchdb
[14:43] <rye> in plaintext... righ
[14:43] <rye> t
[14:44] <rodrigo_> yeah :(
[15:03] <rye> great, with fullscreen terminal my train in LS is moving with a reduced frame rate :-/
[15:03] <rye> something became wrong here...
[15:44] <rye> guys, I was trying to do this on my previous work place and now I just want to ask it here - why are <a> nodes used for buttons that do not actually have the normal href attribute?
[15:45] <rye> desktop+ ?
[15:46] <alevine> rye, you asked me to ping you yesterday about note synchronization failing. It looks like the bug got marked a duplicate of another, but I'm still having the issue.
[15:58] <dobey> rye: they are anchors for the css magic to draw the special buttons
[15:58] <rye> alevine, hi!
[15:58] <alevine> rye, how's it going?
[15:58] <rye> alevine, do you have any recent OOPS for this?
[15:59] <rye> alevine,  it looks like there are logs that were not synced from production service. And it happens to be fixed 15 hours ago
[16:00] <rye> dobey, i just don't like href="#" part when this is actually not working w/o js at all
[16:01] <dobey> rye: eh? href="#" works fine
[16:02] <rye> dobey, # is not - in case the page is scrolled, that forces the browser to scroll it to the top. onclick="return false" will help but that might be an issue with event handlers
[16:03] <dobey> eh
[16:03]  * rye never tried mixing these two
[16:03]  * dobey considers the web to be defective by design
[16:03] <rye> dobey, yes, UI in HTML is
[16:03] <rye> dobey, but we don't have much choice at the moment, XUL ?
[16:04] <dobey> xul is not a choice, it's an abomination
[16:04] <dobey> and no the answer isn't to use xul
[16:05] <dobey> i don't know why forms don't use proper standard form elements any more
[16:05] <alevine> rye, I will get you a new OOPS-id now
[16:05] <rye> alevine, ok, thanks
[16:06] <alevine> rye, can I make one right now? or do you need an old one but not that old? :)
[16:07] <rye> alevine, if it has happened an hour ago - that would be great. But if not, then a new one would do, but we will need to wait for an hour for it to become accessible
[16:07] <alevine> rye, none an hour ago. OOPS-ID-1530appserver70605 is one I generated right now
[16:11] <alevine> will try to remind you at noon
[16:49]  * rye reboots
[17:21] <rye> test
[17:21] <rye> ubottu, test
[17:21] <rye> ubottu, thanks
[17:24] <alevine> rye, it's been an hour ;) OOPS-ID-1530appserver70605
[17:31] <rye> alevine, while I am waiting for a meaningful answer from the the team, could you please check whether the same happens on https://edge.one.ubuntu.com ?
[17:31] <alevine> rye, you mean change my notes server to that?
[17:32] <rye> alevine, ah, you are syncing tomboy, right... hm, no, then this is better not to do since that will reset the sync state
[17:32] <alevine> rye, I will be away from my desk for the next hour or so. I will ping you when I'm back if you're still here. Thanks!
[17:32] <rye> alevine, thanks!
[18:16] <dumpa> hi
[18:17] <rye> dumpa, hi!
[18:17] <dumpa> anyone having trouble with notes sync
[18:17] <dumpa> ?
[18:18] <dumpa> I get [ERROR]: Synchronization failed with the following exception: Unexpected character '<' at [1:1]
[18:18] <rye> dumpa, are you able to open the notes via the web ui?
[18:19] <alevine> rye, I'm back if you need any more info
[18:20] <rye> alevine, the team is now investigating why oopses are not properly generated on the production machines
[18:20] <dumpa> rye, yes I just did
[18:20] <rye> alevine, but... are you able to access your notes at https://one.ubuntu.com/notes/ ?
[18:20] <alevine> rye, yes
[18:20] <rye> dumpa, ok, how many notes do you have, rough estimate
[18:21] <dumpa> hundreds I guess
[18:23] <dumpa> more than 500
[18:23] <alevine> rye, I have 31 notes
[18:23] <rye> dumpa, ok, i am thinking how to actually check what data.... Unexpected character, '<'...
[18:24] <rye> dumpa, could you please start tomboy from the terminal with --debug ?
[18:24] <dumpa> ok
[18:24] <alevine> rye, dumpa unexpected character is what I get from the JSON parser
[18:24] <alevine> in my case it is because HTML is being returned
[18:24] <alevine> (the 500 error from the server)
[18:27] <dumpa> alevine, I'm sorry i don't get you, what does that mean?
[18:28] <rye> dumpa, the server returns the error response, not the data Tomboy expected, the first character of the response is <, from <html>
[18:28] <dumpa> ok
[18:28] <alevine> dumpa, and < is not a valid first character for JSON, so the parser throws that error
[18:29] <dumpa> ok, so is this a known bug?
[18:29] <fax> is this cloud computing?
[18:31] <rye> dumpa, this is the bug that might be caused by unexpected service restarts, but I feel that it is strange that the same code works fine for web ui and oopses during the outputting the data to tomboy
[18:34] <dumpa> rye, what do you mean by same code?
[18:35]  * rye has an idea about what can cause that...
[20:24]  * wgrant points at https://bugs.edge.launchpad.net/ubuntuone-servers/+bug/535651 again.
[20:24] <dobey> haha
[20:25] <wgrant> Not really, no.
[20:25] <dobey> not the bug, no
[20:25] <dobey> ubottu's "list index out of range"
[20:26] <dobey> certainly wouldn't think of that
[20:26] <wgrant> Heh.
[23:24] <duanedesign>   bug 531709
[23:26] <Omar> Help meee
[23:26] <Omar>  wiht ubuntu one
[23:26] <Omar> ayuda
[23:26] <Omar> intento suscribirme y al continuar con la suscripcion me bota este error Password must be at least 8 characters long, and must contain at least one number and an upper case letter.
[23:29] <dobey> duanedesign: the logs on there don't line up right. he'll need to attach a new copy of oauth-login.log that has times close to the syncdaemon.log entries
[23:30] <duanedesign> thank you dobey
[23:34] <dobey> duanedesign: but my guess is that syncdaemon is just taking forevor to start
[23:41] <duanedesign> dobey: ahh, and that would be easily confirmed with the oauth-login.log DBus Error
[23:42] <dobey> duanedesign: yes, if the dates lined up it would be pretty easy to confirm :)
[23:49] <duanedesign> I am just about done with my Python script that saves dpkg selections to a Couchdb and then uses that database to set the selections on other computers (after U1 syncs the DB of course :))