13:01 <popey> #startmeeting Trojita meeting
13:01 <meetingology> Meeting started Fri May 16 13:01:42 2014 UTC.  The chair is popey. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
13:01 <meetingology> 
13:01 <meetingology> Available commands: action commands idea info link nick
13:01 <popey> Busy!
13:01 <DanChapman> :-D
13:01 <popey> We're getting prepared to go to Malta next week for a sprint
13:02 <popey> Which means we'll all be working in the same timezone for a week which is nice!
13:02 <DanChapman> Ahh yes i remember jono mentioning when i spoke with him, sounds fun :-)
13:02 <mhall119> o/
13:02 <popey> hey mhall119
13:03 <popey> DanChapman: I think dholbach has resolved the recipe building issue in launchpad
13:03 <popey> (and found another issue)
13:04 <DanChapman> oh great, I did see one of the builds had failed on all arch's. What's the new problem?
13:04 <DanChapman> hey boren
13:04 <dholbach> hiya
13:04 <boren> hey DanChapman
13:04 <popey> hi boren
13:05 <DanChapman> hey dholbach o/
13:05 <mhall119> he had an issue with the git commit hash not being ordered, but that should be easily fixable
13:05 <dholbach> hey DanChapman
13:05 <popey> dholbach: so last week there was a concern about the build system, is that all resolved now?
13:05 <dholbach> it's fixed
13:05 <popey> Super!
13:05 <DanChapman> awesome
13:05 <popey> Your work here is done.
13:05 <popey> ☻
13:05 <dholbach> I just pushed a fix for a fix
13:05 <dholbach> but unfortunately we have to wait for a new commit in git to be imported, etc
13:06 <dholbach> but it should work now
13:06 <dholbach> keep up the good work everyone! :)
13:06 <DanChapman> There should be a couple new commits landing soon so not long to wait on that one
13:06 <mhall119> thanks dholbach
13:06 <dholbach> DanChapman, awesome
13:07 <DanChapman> popey did (was it mardy?) get back to you about the password storage blueprint
13:07 <DanChapman> I scanned scrollback but didn't find anything
13:08 <popey> no, i dont think so.
13:08 <popey> do you have the link handy?
13:09 <DanChapman> I have somewhere 2 secs
13:10 <DanChapman> popey: https://blueprints.launchpad.net/ubuntu/+spec/mobile-s-password-storage
13:10 <popey> ta
13:11 * popey summons mardy
13:12 <DanChapman> popey the one that would really be great to have is gnome keyring since trojita already has support for QtKeychain which uses Gnome keyring if available
13:12 <DanChapman> but hey anything would be great tbh
13:13 <davmor2> popey: burns the wrong fragranced stick for summoning again
13:13 <popey> indeed
13:14 <popey> I'll poke him again later.
13:14 <DanChapman> boren did you have any more luck with the message view? I wan't to see where it is at before sending a message out to webkit ML
13:14 <popey> DanChapman: is that blocking you?
13:15 <DanChapman> popey ok great thanks
13:15 <DanChapman> popey no not blocking, i've just finished implementing the password watcher but atm it stores passwords in cleartext
13:15 <DanChapman> ideally we don't want that
13:16 <boren> Danchapman, nothing.
13:16 <popey> \o/ My magic still works.
13:16 <mardy> :-)
13:16 <popey> mardy: we're discussing the Trojita email client.
13:17 <mhall119> welcome mardy :)
13:17 <popey> It's an upstream project which we're contributing to.
13:17 <popey> DanChapman: has a question regarding https://blueprints.launchpad.net/ubuntu/+spec/mobile-s-password-storage
13:17 <DanChapman> boren ok cool. Lets hope we get a response from someone :-D
13:17 <popey> and what the status is
13:17 <popey> as we'd like secure storage for email creds on the device of course
13:17 <DanChapman> mardy, hey o/ yes what popey said ^
13:18 <mardy> DanChapman: so, AFAIK the plan is to use the GNOME keyring on the phone as well, but it's just not there yet
13:18 <popey> Oh that's good news.
13:18 <mardy> need to find a password for it
13:18 <DanChapman> mardy, awesome that's what i was wanting to hear :-D
13:19 <popey> mardy: multi-user is coming soon I thought?
13:19 <mardy> however, I'd really recommend you to store the passwords via Online Accounts
13:20 <mardy> at least for those providers which are supported by OA, such as gmail :-)
13:21 <popey> yeah, thats gonna be tricky I think.
13:21 <popey> especially for $random email provider
13:22 <mhall119> we could probably make a Generic IMAP provider for OA
13:22 <popey> that would indeed be one way
13:23 <mhall119> I assume the provider can store server/port/encryption info
13:24 <DanChapman> is that something we would need to do? I can't see any docs for creating an OA provider
13:26 <mhall119> we could get somebody else to make that
13:26 <mhall119> maybe kenvandine
13:26 <DanChapman> mhall119: that would be awesome
13:26 <DanChapman> hey jkt :-)
13:27 <kenvandine> not sure how easy a generic imap provider would be
13:27 <jkt> hi there
13:27 <jkt> sorry, got another meeting at work
13:28 <popey> np
13:28 <popey> I'm not inclined to want to go down that road right now
13:29 <popey> jkt: for context we were talking about the gnome keyring for credential storage, then mardy suggested we craft a generic IMAP online accounts plugin
13:30 <jkt> yeah, well, gnome-keyring is already handled by the qtkeychain, which means that there's no work but packaging for this one
13:30 <popey> indeed
13:32 <mhall119> DanChapman: jkt: were you guys able to get the sorting/old email issue resolved?
13:32 <jkt> dholbach: fyi, that won't really help with exporting the git hash into bzr, but here's how the OBS builds were changed so that the packages are now properly versioned, http://quickgit.kde.org/?p=trojita.git&a=commitdiff&h=aefd064d575eed09b9a3973d0da57667fa272c85
13:32 <jkt> mhall119: nope
13:33 <mhall119> :( that's an annoying one, it makes the app pretty unusable at the moment
13:33 <jkt> mhall119: I didn't have any time to work on that, unfortunately, and the next week is gonna suck as well
13:33 <dholbach> jkt, sorry - what won't help?
13:33 <dholbach> jkt, I generate src/trojita-version like you told me to in the last meeting
13:33 <DanChapman> mhall119: I've been wrestling with a temp fix for that this morning, using the rowsInserted and modelReset to try and force it to sort but when ever the model resets it jumps to the top of the view which is annoying when scrolling. THe best i have is toggling the sort order from the toolbar, which seems to hold it's state from then on
13:33 <jkt> mhall119: seems that a patch which "just" adds an explicit call to setting the sort prefs should do the trick, though
13:34 <jkt> DanChapman: don't reset the model, just set the sorting prefs
13:34 <jkt> dholbach: then it means that everything is fixed, sorry for noise -- I forgot about that one
13:34 <dholbach> ok cool
13:34 <dholbach> thanks
13:36 <mhall119> DanChapman: can you try jkt's suggestion and see if that works?
13:36 <popey> DanChapman: is the trello up to date?
13:37 <DanChapman> jkt that's what i've been trying, but no luck with it so far
13:38 <mhall119> jkt: is there anyone else who's been working on Trojita with you that could work with DanChapman on this?
13:38 <DanChapman> popey yes AFAIK not sure what the status is with gcollura's branch, i'll check that now
13:38 <mhall119> paalsteek for instance
13:39 <popey> thanks
13:39 <popey> mhall119: any progress on click builds?
13:39 <popey> let me rephrase.
13:39 <popey> Any progress on easier click builds? ☻
13:39 <jkt> mhall119: people can ask on the ML or on IRC
13:40 <mhall119> bzoltan yesterday told me to as set(CLICK_MODE on) to tell QtCreator to treat it like a click project, I've yet to try it though
13:40 <mhall119> jkt: ok
13:40 <mhall119> popey: I also need to make the CMAKE copy the click files out of the packaging folder into the build root
13:41 <gcollura_> hello everyone, sorry for being late
13:41 <popey> mhall119: reckon we can nail that in malta next week?
13:41 <popey> hi gcollura_
13:41 <mhall119> popey: unless I get time enough today to try it :)
13:42 <DanChapman> mhall119: but we don't have CLICK_MODE setup like the core-apps do in Cmakelists
13:42 <mhall119> DanChapman: that's what I'm going to try adding
13:43 <DanChapman> mhall119: ahhh sorry misunderstood
13:43 <DanChapman> :-)
13:45 <mhall119> hey gcollura_
13:45 <mhall119> so what's the status of sending email?
13:46 <popey> Yeah, i can't wait to start spamming people from my phone with a signature...
13:46 <mhall119> IIRC, that's the last thing we wanted before publishing the early versions to the store
13:46 <popey> Sent from Trojita on my Ubuntu device
13:46 <popey> ☻
13:46 <DanChapman> hey gcollura_ didn't see you arrive
13:46 <mhall119> +1
13:48 <gcollura_> hi DanChapman, this weekend I should have more time, this week I've been very busy. I requested jkt the review of the first piece of code, I hope to fulfill his requests and meet his requirements :)
13:49 <DanChapman> sounds great :-)
13:50 <mhall119> thanks gcollura_
13:50 <popey> super.
13:51 <popey> DanChapman: Are there any other blockers you have or anything we can help with?
13:52 <gcollura_> jkt: do you have an idea for me to solve this one? https://git.reviewboard.kde.org/r/118030/#comment40134
13:52 <jkt> well, I could use someone to smack the QML developers about how they implemented enums
13:52 <popey> heh
13:52 <DanChapman> jkt +1
13:52 <jkt> basically anything which is outside of a QObject cannot be used
13:52 * mhall119 volunteers for smacking duty
13:52 <jkt> which kinda sucks if you have an existing project
13:52 <jkt> go ahead, then
13:53 <DanChapman> popey just the webview displaying messages is the biggest issue atm, im composing an email for that now
13:53 <jkt> mhall119: http://lists.qt-project.org/pipermail/interest/2014-May/012217.html
13:53 <DanChapman> hopefully get a response
13:53 <jkt> oh yeah, and do that while you're at it as well
13:53 <jkt> DanChapman: can you block http requests using the custom URL scheme delegates?
13:54 <jkt> if not, then it means that Trojita would have serious privacy issue by automatically confirming e-mail opening to a hostile 3rd party
13:55 <DanChapman> jkt, ok well the documentation on the experimental parts is next to nothing, I will look into it this afternoon
13:56 <mhall119> jkt: DanChapman: I think we can overrridde onNavigationRequested for that
13:56 <DanChapman> mhall119: what's the future of QtWebkit being on device once oxide is on par with it?
13:57 <mhall119> DanChapman: it may be "on device" but we already don't recommend it
13:57 <jkt> mhall119: nope; this is abuot fetching stuff like "background CSS"
13:57 <mhall119> unfortunately UbuntuWebView doesn't have the experimental features we're using
13:57 <jkt> or inlined images
13:57 <DanChapman> https://bugs.launchpad.net/oxide/+bug/1260016 would be needed for a start
13:57 <mhall119> jkt: onNavigationRequested?
13:58 <mhall119> I think I used that in the past to intercept link clicks on anothe rapp
13:58 <jkt> mhall119: suppose I send you a mail whose html contains <img src="trackme.com/transparent.png?mail=mhal119@example.org"/>
13:58 <mhall119> oh, I see what you mean now
13:59 <jkt> mhall119: DesktopGui intercepts that, and shows a button, empowering the user to load these and to let the net know that the user has read the mail
13:59 <mhall119> I thought yo uwere talking about when the user clicked alink
13:59 <jkt> so, what exactly is the UbuntuWebView?
14:00 <mhall119> jkt: a wrapper around the Oxide component, which is itself an Chromium embedding component
14:00 <mhall119> it basically wrapps the webbrowser-app's renderer as a reusable component
14:01 <jkt> mhall119: OK, so there are essentially two requirements for any web rendered used within Trojita
14:01 <jkt> mhall119: first of all, it must be embeddable into QML, and it must be possible to configure it so that it will *not* handle scrolling and resizing etc on its own
14:01 <jkt> mhall119: second, it must handle any sort of network IO whatsoever, including HTTP, to Trojita
14:01 <mhall119> the first is true for UbuntuWebView, I'm not so sure about the second
14:02 <mhall119> that last one I know it doesn't do
14:02 <jkt> then it cannot be used, period
14:02 <popey> or these need to be tracked as items it needs to be able to do
14:02 <jkt> unless you do crazy stuff like injecting a proxy into the path, which would be a creative, yet rather inefficient hack
14:03 <jkt> sure, I'm discussing technical requirements
14:03 <jkt> I also acknowledge that it's a PITA to work with WebView
14:03 <jkt> gcollura_: sorry for ignoring you
14:03 <mhall119> I think we should stick with webview for now, even if it means we have to ship it as part of our click package in the future
14:04 <popey> +1
14:04 <gcollura_> jkt: don't worry :)
14:04 <DanChapman> +1
14:04 <jkt> gcollura_: about these macros, I see a couple of identical functions which differ only in name of the member they access, and in its type
14:05 <jkt> gcollura_: perhaps you could do something like PROPERTY_IMPL(TYPE, NAME, NAME_UPPERCASE)
14:05 <gcollura_> ok I know how to do that, perfect :)
14:06 <jkt> gcollura_: I hate macros as much as anybody else, but in this case they are probably reasonable
14:06 <jkt> too bad qt4's moc doesn't do MEMBER, as Thomas Lübking suggested
14:07 <jkt> gcollura_: I would also like to have some reasonable error reporting, i.e. something which reaches the user instead of just ending up in console.log
14:08 <gcollura_> I'd like to show some sort of popup notification, I don't know if there are some sort of design guidelines or I can use my own fantasy
14:08 <jkt> and I know that I still owe you some comment about how much I think this wrapper is needed, compared to "just" extending the Composer::MessageComposer with properties and invokable methods
14:09 <boren> DanChapman: would you mind letting me look into blocking http requests?
14:09 <DanChapman> boren go for it :-D
14:09 <popey> thanks boren
14:09 <gcollura_> I'll see if extending MessageComposer doesn't bring any mess
14:09 <mhall119> is there anything else to discuss in here rather than #trojita? we're already past the hour mark
14:10 <DanChapman> Nope i'm good for now, thanks guys
14:10 <boren> thanks
14:10 <gcollura_> thanks
14:10 <popey> Great, thanks chaps.
14:11 <popey> Have a great weekend, and if you need anything you know where we are.
14:11 <popey> #endmeeting