13:05 <popey> #startmeeting trojita
13:05 <meetingology> Meeting started Fri Jun 13 13:05:27 2014 UTC.  The chair is popey. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
13:05 <meetingology> 
13:05 <meetingology> Available commands: action commands idea info link nick
13:05 <DanChapman> yeah not too bad, i've got to shoot off early today meeting to attend so i will catch up on the backlog when i get back
13:05 <popey> Hows this week been for you guys?
13:05 <DanChapman> hey popey
13:05 <DanChapman> not too bad, quite a bit of head scratching but making progress
13:06 <DanChapman> boren nice job on the progress indicator
13:06 <popey> hi jkt
13:06 <DanChapman> hey jkt
13:07 <boren> Hi jkt
13:07 <jkt> hi
13:07 <dholbach> hiya
13:08 <DanChapman> howdy dholbach
13:08 <popey> I think we had a problem building earlier in the day, right dholbach ?
13:08 <dholbach> hey DanChapman
13:08 <dholbach> popey, we didn't notice for a short time
13:08 <popey> https://trello.com/b/73rFyG6n/ubuntu-email-client is the trello board for reference
13:08 <popey> ah right
13:08 <dholbach> one problem was that a build-dep had to be moved from python3-minimal to python3 (easy fix)
13:08 <dholbach> another problem was another test which failed on armhf
13:09 <jkt> dholbach: what fails?
13:09 <dholbach> hang on, I need to dig it out again
13:09 <dholbach> it might have been a timeout again
13:09 <dholbach> I have to check
13:10 <dholbach> ah no, here it is: https://launchpadlibrarian.net/177048562/buildlog_ubuntu-utopic-armhf.trojita_0.4.1%2B2-0~201406051548.git4a74770~ubuntu14.10.1_FAILEDTOBUILD.txt.gz
13:10 <dholbach> 6/24 Test  #6: test_Imap_MsgPartNetAccessManager ...........***Exception: SegFault  0.19 sec
13:10 <dholbach> ********* Start testing of ImapMsgPartNetAccessManagerTest *********
13:10 <dholbach> Config: Using QtTest library 5.2.1, Qt 5.2.1
13:10 <dholbach> PASS   : ImapMsgPartNetAccessManagerTest::initTestCase()
13:10 <dholbach> qemu: uncaught target signal 11 (Segmentation fault) - core dumped
13:10 <DanChapman> guys i've got to run, i'll catch up with you all as soon as i get back, any questions for me just leave them in here and i'll catch them when i get back. Sorry i have to go
13:10 <jkt> that doesn't look like a false positive
13:10 <popey> DanChapman: no problem
13:11 <jkt> dholbach: could you please run it through valgrind?
13:11 <dholbach> and there was the question if Qt5Keychain was something we should look into enablgin
13:11 <dholbach> popey, can we find somebody to check why this test is failing?
13:11 <dholbach> I'm behind on a number of things
13:11 <popey> someone with what skills? trojita or something else?
13:12 <dholbach> running the build and particularly the test_Imap_MsgPartNetAccessManager under valgrind on armhf
13:13 <dholbach> there was also the question if Qt5Keychain was something we wanted to start using(?)
13:13 <jkt> just that test, not the build itself
13:13 <popey> well, DanChapman has a device, perhaps he's best placed to do that?
13:14 <popey> I'll add a task to trello for him so he doesn't miss it.
13:14 <dholbach> I can add it to my TODO list, but I know that it'll be stay on there for a while :/
13:14 <popey> its fine, DanChapman isn't here to say no.
13:15 <dholbach> ok, let me know if that's going to work
13:15 <dholbach> Qt5Keychain was the last question I had
13:16 <jkt> for whom is that question?
13:17 <dholbach> I don't know
13:17 <dholbach> I just wasn't sure if using Qt5Keychain in the build was something we wanted to do
13:18 <jkt> from my point of view it makes sense, but it also means shipping that library either as a system one (which is good), or bundling it with Trojita (which is completely unsupported by me)
13:19 <jkt> without that library, there's no support for remembering password
13:19 <popey> Right. We already use gnome-keyring I believe on the device
13:21 <popey> So I'm not sure what the impact of having both qtkeychain and gnome-keyring on the device would be
13:21 <dholbach> ahhh ok
13:21 <dholbach> I'll leave the decision up to somebody else then
13:22 <popey> yeah, need to figure that out. I'll take a task for that.
13:22 <dholbach> awesome!
13:22 <dholbach> thanks guys
13:22 <popey> thanks dholbach
13:23 <jkt> popey: this is a Qt wrapper on top of gnome-keyring, essentially
13:23 <popey> so in the trello I see two things in "waiting for review"
13:23 <popey> "this" being Qt5Keychain ?
13:23 <jkt> yes
13:23 <popey> oh
13:24 <jkt> it will talk to gnome-keyring if it finds one
13:24 <popey> well that should be less stress to get on the image then.
13:24 <popey> I didn't realise that, thanks
13:24 <jkt> it does not offer its own storage or anything like that
13:24 <jkt> it just wraps what the platform provides
13:24 <popey> ah, i assumed it was an alternative
13:24 <popey> libqt5keychain0 looks to be the package in ubuntu
13:24 <jkt> it has adaptors for kde wallet, windows' whatever, and gnome-keyring as well
13:24 <popey> neat!
13:25 <popey> any particular version you need?
13:25 <popey> we have 0.3.0-2 in the ubuntu utopic archive
13:25 <jkt> let me check
13:26 <jkt> seems to be fine
13:26 <jkt> I see a commit in git after the 0.3.0 tag was made
13:26 <jkt> Detect GNOME keychain daemons presence using official API
13:26 <jkt> 
13:26 <jkt> The previous approach was hacky and broke often, as
13:26 <jkt> variables kept changing again and again.
13:26 <popey> ok, I'll see about getting that in the image, we'll also need the debian packaging updated to ensure the deb is built with that as a dependency
13:28 <popey> so back to trello, there's a couple of waiting for review items
13:28 <popey> one for markup plaintext messages as html, and a sync now function.
13:29 <popey> no links in the trello board so I cant see what they relate to directly.
13:29 <boren> wait a minute
13:30 <popey> do you have the links boren ?
13:30 <boren> yes
13:30 <boren> It's there
13:31 <jkt> I pinged gilles yesterday about the status of the "sync now" thingy
13:31 <jkt> seems that he updated the git branch, but did not update the reviewboard
13:32 <jkt> paalsteek reminded him what to do today
13:32 <jkt> another fact is that I'm still waiting for feedback from DanChapman on that review request
13:32 <jkt> beacuse I need to confirm that it's indeed a gmail screwup before I merge a patch which is pointless on any well-behaving IMAP server
13:33 <popey> Understood.
13:33 <jkt> re the plaintext -> HTML highlighting, I got mad at seeing a patch done exactly the wrong way, just as a result of Ctrl-C + Ctrl-V
13:34 <jkt> that's something I would prefer not to see from people who already know what development method I disliek in particular
13:34 <boren> jkt: Sorry. I thought copy would make it safe at first.
13:35 <jkt> boren: well, please understand that this is a huge waste of my time
13:36 <jkt> boren: to your question -- yes, moving stuff from SimpelPartWidget/whatever into UiUtils seems to be a reasonable thing
13:37 <boren> jkt: Sorry, next time I will say "literally copy" in my description and title.
13:37 <jkt> nope, that would be similarly wrong
13:37 <jkt> in general, I dislike to see *any* duplication of code, with no exceptions
13:38 <jkt> it is perfectly fine to e.g. make one commit where you refactor the existing DesktopGui code by moving a method from, say, the SimplePartWidget.cpp into a corresponding place in UiUtils/whatever.cpp
13:38 <jkt> while adjusting the DesktopGui code in the process
13:38 <boren> jkt, got it, sorry about wasting your time.
13:38 <jkt> then you can make another commit where you use that code from the Ubuntu's code
13:39 <jkt> I'm sorry for being mad here, but I was under impression that we already went through this discussion at least once, hence my surprise
13:39 <jkt> and it is important for me to use this as an opportunity to make sure that you guys know what approach I vastly prefer
13:41 <popey> Ok.
13:41 <popey> Are there any other issues or blockers we need to discuss?
13:42 <boren> For the webview's height been to large. I find that markingup html still wouldn't solve it as Danchapman suggested before.
13:43 <boren> I am think of using javascript to get the hight of element within <body
13:43 <popey> is that something you can easily prototype to see if that'll work?
13:45 <jkt> I'm afraid this won't work with pure HTML mail
13:45 <boren> I have made it work. It may need enabling javascript as the height might change for marked up html.
13:45 <jkt> right
13:45 <jkt> the current plaintext-to-html interactively hides quoted blocks using pure CSS, without JS
13:45 <jkt> so the height is indeed dynamic
13:46 <jkt> making JS temporarily enabled is ugly, IMHO, but at least not blatantly insecure
13:46 <jkt> I'm afraid that enabling JS on incoming HTML mail is behind the line, though
13:46 <jkt> so I fail to see how would this handle HTML mail
13:47 <boren> OK. I just want to find a temporary fix for now. It is really annoying.
13:48 <jkt> I share your goals, I'm just pointing out to what I see as problems
13:49 <jkt> isn't there some signal when the page height changes?
13:50 <boren> right, maybe that will work
13:50 <boren> contentHeight change.
13:52 <popey> Ok. Shall we wrap up and move further discussion to #trojita?
13:52 <popey> I'll catch up with DanChapman when he's back on line later.
13:53 <popey> Thanks jkt boren DanChapman
13:53 <popey> #endmeeting