16:02 <popey> #startmeeting Email client meeting
16:02 <meetingology> Meeting started Fri Mar 14 16:02:11 2014 UTC.  The chair is popey. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
16:02 <meetingology> 
16:02 <meetingology> Available commands: action commands idea info link nick
16:02 <popey> An excellent plan!
16:02 <popey> thanks for coming guys. unfortunately randomcpp couldn't make it, he's in a lecture at this time.
16:02 <popey> so the first thing I'd ask is, is this a good time for you guys?
16:03 <popey> if we do this every week is that okay or should we move it?
16:03 <DanChapman> good for me
16:03 <boren> yes
16:03 <mzanetti> o/
16:03 <DanChapman> weekly sounds like a plan :-)
16:03 <DanChapman> mzanetti: o/
16:03 <mzanetti> popey: if its easily possible I'd prefer earlier
16:04 <mzanetti> popey: but no big deal if not
16:04 <popey> ok. so is 2 hours earlier possible for you guys?
16:04 <boren> yes
16:04 <mhall119> works for me
16:04 <popey> Be nice to get as many people as possible here, but it's difficult when everyone is all over the world.
16:04 <DanChapman> works for me 2
16:04 <popey> plus work, family etc
16:04 <popey> hi jkt
16:04 <jkt> hi there
16:05 <popey> great timing, we're just starting.
16:05 <DanChapman> hey jkt
16:06 <mzanetti> hi jkt
16:06 <popey> Ok, so first thing. Thank you! I'm so happy that we have a bunch of enthusiastic people who want to help us build a great email experience for ubuntu
16:07 <mhall119> and a personal thanks to boren, who's port I've been using for a couple weeks now on my phone to *finally* access my work email :)
16:07 <popey> I'd like us to talk about the goals, but first I wonder, jkt would you have a moment to set out your house-rules as our upstream?
16:07 * DanChapman needs to try it out
16:07 <popey> Anything in particular we should ensure we care about, and processes you'd like us to follow.
16:07 <popey> We're keen to work happily with upstream.
16:08 <jkt> hi there again
16:08 <mhall119> also, I'd like to stress that this project isn't a fork, it's an *upstream* project, and we need to make sure we are working as closely as possible with jkt
16:08 * jkt had a bit of scheduling problem
16:08 <jkt> so basically, I would like to help as my time permits
16:08 <jkt> I also prefer working in a single repo as much as possible
16:09 <jkt> it's fine to try out stuff in private, of course, but I've seen it numerous times, and I'm referring to experience outside of the ubuntu community
16:09 <jkt> that when people don't approach upstream early, it takes a ton of time to get everything integrated properly
16:09 * mzanetti is confirming that experience
16:09 <boren> That is so correct
16:10 <jkt> I've seen an example of that in one of the ubuntu ports, too, unfortunately
16:10 <jkt> so I would like to avoid this in future
16:10 <mhall119> right, so first off we need to make sure that new work is done against a fresh clone of the upstream repo
16:10 * mzanetti needs to catch up. Where's the upstream repo?
16:10 <jkt> there's a Harmattan port during which we went through quite some trouble
16:11 <jkt> mhall119: http://quickgit.kde.org/?p=trojita.git
16:11 <jkt> so let's not go through all the pain again
16:11 <jkt> that's all I have for now :)
16:11 <mhall119> jkt: where can we push proposed changes? (/me isn't familiar enough with git)
16:11 <jkt> ah, well, this is going to be a bit inconvenient
16:11 <jkt> there's no easy setup like gerrit under KDE, unfortunately :(
16:11 <mzanetti> oh... in kde. so I guess we should use reviewboard
16:12 <mhall119> popey: can you give me an action item to update https://wiki.ubuntu.com/Touch/CoreApps/EmailClient
16:12 <jkt> KDE people love reviewboard, and there's a community of developers who are happy to comment on them and point out all the spell problems in commit messages :), that happens to me every time
16:12 <mzanetti> true...
16:13 <jkt> but RB's workflow kind of sucks when it comes to passing patches around, so please always accompany the review request with a pointer to a repo one could pull from
16:13 <jkt> sorry for that -- we're working on something better, but the community is around the RB for now
16:13 <mhall119> so a github repo would be okay ?
16:13 <jkt> yes, any repo, github included
16:13 <mhall119> ok
16:13 <jkt> the point is, if it's just on github, nobody sees it
16:13 <mhall119> as long as it's a clone of upstream, so they share revision history
16:14 <mhall119> right?
16:14 <jkt> if it's accompanied by a RB request, we notice
16:14 <jkt> yes
16:14 <popey> mhall119: ya
16:14 <mzanetti> ack. reviewboard it is
16:14 <jkt> "a repo" as in "have common history" :)
16:14 <mhall119> ok
16:14 <popey> #action mhall119 Update https://wiki.ubuntu.com/Touch/CoreApps/EmailClient
16:14 * meetingology mhall119 Update https://wiki.ubuntu.com/Touch/CoreApps/EmailClient
16:14 <jkt> oh, one more thing
16:14 <jkt> the smaller the commits are, the better
16:14 <jkt> makes them easier and more pleasant to review
16:15 <jkt> also, please no regressions if possible, especially build failures
16:15 <jkt> mades bisecting harder
16:15 <mhall119> so point #2 that jkt mentioned, the port should build off the Harmattan setup where cmake flags build a binary that uses a separate UI frontend
16:15 <mhall119> jkt: is there a test suite we can/should run before making a pull request?
16:15 <jkt> `make test`
16:16 <jkt> in fact, `time (make -j6 && ctest -j 666 --output-on-failure)` is a handy one-liner I use all the time
16:16 <jkt> kills debug noise from the tests
16:17 <jkt> there's also a CI setup at KDE which checks whatever gets pushed to KDE's repos
16:17 <mhall119> ok
16:17 <mhall119> jkt: mzanetti: can you link me to the RB process so I can document it in our wiki page?
16:17 <jkt> and an automated setup via github and travis-CI which builds both qt5 and qt4 versions on Ubuntu 12.04
16:18 <mzanetti> mhall119: https://git.reviewboard.kde.org
16:18 <mhall119> jkt: what version of Qt does Trojita support? is 5.2 okay?
16:18 <mzanetti> mhall119: on the workflow... there's many. as jkt said, its not the easiest one... I usually upload pathcsets I created with "git format-patch"
16:18 <popey> We landed qt5.2 in the archive *today*
16:18 <mhall119> thanks mzanetti, looks like I need to create another login account on another system :(
16:19 <mzanetti> yep
16:19 <DanChapman> mhall119: same here
16:19 <jkt> perhaps this could be extended to check the touch version as well
16:19 <mzanetti> not sure if it integrates with identity.kde.org
16:19 <mhall119> mzanetti: sounds like I'll be pinging you again as I write this down on the wiki
16:19 <jkt> mhall119: 5.2 is fine
16:19 <mzanetti> ack
16:19 <jkt> it's even required if it's qt5 due to some QUrl bugfixes
16:19 <mhall119> ok
16:19 <jkt> mzanetti: it does
16:20 <jkt> the KDE identity login is enough
16:20 <jkt> there's that post-review script which is pretty handy
16:20 <jkt> http://techbase.kde.org/Development/Review_Board#Using_post-review_to_post_changes_for_review
16:20 <jkt> makes working with RB tolerable
16:21 <mhall119> jkt: currently our click store doesn't support group ownership of an upload, is it okay by you if one of us is uploading new releases of hte Ubuntu port to the store?
16:21 <jkt> mhall119: sure
16:21 <jkt> license-wise, iirc the SMTP library we use is gpl-2 only
16:22 <jkt> most of the code is dual-licensed under gpl-2 and gpl-3, with a provision to have the KDE e.v. (and org behind KDE) the power to relicense to upcoming GPL releases, too
16:22 <jkt> I would prefer if the contributions were under this license, or anything more permissive
16:22 <jkt> BSD is fine
16:22 <popey> mhall119: I'd recommend you upload and I test and approve, to separate responsibility.
16:22 <mhall119> I think using the same license is best
16:22 <popey> +1
16:23 * mzanetti is fine with anything in this case
16:23 <jkt> is this OK with you, both regarding your contributions, and also about distributing stuff?
16:23 <mzanetti> anything free and open, that is
16:23 <mhall119> popey: works for me, once I get cross-compiling working, mzanetti can you help me with that one?
16:23 <jkt> some HW vendors are nuts with GPLv3, for example
16:24 <mhall119> if it's dual-licensed gplv2 and v3, I think it's okay
16:24 <jkt> I recall seeing some gpl-3 only pieces in some port I saw in the recent days, but I might be mistaken
16:24 <mhall119> an Ubuntu port?
16:24 <mzanetti> mhall119: no. I didn't manage to cross compile anything yet. click chroot fails in every possible way here every time I try
16:24 <jkt> that's a situation I would like to avoid
16:24 <jkt> mhall119: yes
16:25 <mhall119> mzanetti: my click chroot seems to be okay (after some package tweaking), but uic is failing to build some of the UI files for addressbook
16:25 <mzanetti> hmm... not even sure we support that, given we don't support qmake
16:26 <jkt> how come it's trying to build .ui stuff?
16:26 <mzanetti> but in any case, we shouldn't need to compile any QWidget stuff soonish
16:26 <jkt> that's for the desktop version
16:26 <mhall119> jkt: we can ask that contributions to this joint project are all the same license
16:26 <mhall119> mzanetti: it's cmake now
16:26 <mzanetti> mhall119: yeah, but uic is a tool coming with qmake
16:26 <mhall119> jkt: no idea, but it does
16:26 <jkt> must be some cmake screwup -- perhaps it's enabling parts of the desktop build for some reason
16:26 <mhall119> maybe my cmake flags aren't being transmitted over to the chroot
16:26 <jkt> like by -DWITH_QT5=on
16:27 <mhall119> I used -DWITH_QT5=ON -DWITH_UBUNTUTOUCH=ON
16:27 <mhall119> should I not use -DWITH_QT5?
16:27 <mhall119> I thought without that it would use Qt4, which isn't compatible with the Ubuntu UI toolkit
16:27 <jkt> I don't know how the code of that patched version is set up, sorry
16:28 <mhall119> ok, I'll try and figure that out
16:28 <jkt> where does "the code" live now?
16:28 <mhall119> depends on which port you're talking about :)
16:28 <mzanetti> so where should I start? from trojita master or some other branch?
16:29 <mhall119> alright, so who here has started an Ubuntu port?  I know boren has, anyone else?
16:29 <mhall119> kenvandine did some work on cmake to get it compiling with the UbuntuTouch UI (same way Harmattan did)
16:29 <DanChapman> mhall119: I've only played around with it, and not really invested anything since we last spoke
16:31 <popey> mhall119: randomcpp did, but he's not here
16:31 <mhall119> can you guys open https://docs.google.com/document/d/1MlGUefZxlYxJzwvIlD0LzhrUA8T4ulVtHA7Fb78RPyY/edit?usp=sharing
16:31 <DanChapman> boren could i have a link to your port?
16:32 <mhall119> heh, Google docs give anonymous people animal pictures now
16:32 <boren> https://github.com/bobo1993324/qmlTrojita. Don't use my port, it not anything close to Harmattan
16:34 <DanChapman> mhall119: so is the development process as it's not quite the 'normal' lp way be outlined in the wiki
16:34 <mhall119> boren: but it's also one of the most functional UIs for Ubuntu, we just need to convert it to be built like the Harmattan code
16:34 <mhall119> DanChapman: not yet, but it will be
16:34 <DanChapman> boren thanks, i'll take a look.
16:36 <mhall119> who can act as our resident Git expert and mentor?
16:36 <jkt> btw, one more thing which I should have said at the very beginning -- I'm really happy that this is happening :)
16:36 <popey> Me too.
16:36 <mhall119> me too :)
16:36 <popey> Like you wouldn't believe!
16:36 <jkt> sorry for random topic hijacking, I'm just taking the liberty of IRC :)
16:36 <mzanetti> mhall119: I guess I can
16:36 <mhall119> thanks mzanetti
16:37 <popey> jkt: just realised before you arrived we decided to move the meeting back two hours to 14:00 instead of the 16:00 start time it currently is, is that still okay with you?
16:37 <popey> (starting next week)
16:37 <mhall119> does anybody want to take on the first step, or providing a fresh git branch with CMake modifications and Ubuntu UI directory?  If not, I can try and tackle that with kenvanine's branch
16:37 <DanChapman> popey 14.00 is perfect for me just before school run
16:37 <DanChapman> :-)
16:38 <popey> cool.
16:38 <jkt> popey: so 14.00 UTC, every friday?
16:38 <mhall119> do we want to hold future meetings in here or in #trojita?
16:38 <popey> yeah.
16:38 <jkt> fine with me
16:38 <popey> great.
16:39 <mzanetti> ack
16:39 <jkt> I cannot promise I'll be always available, of course -- $dayjob, etc
16:39 <popey> of course
16:39 <jkt> I'll definitely be offline next Friday
16:39 <popey> the benefit of being in here with meetingology is the bot for logging and quietness
16:39 <popey> but I don't really care where the meeting is ☻
16:39 <jkt> if it's going to be Ubuntu-specific, I'm fine with being here
16:39 <jkt> I would prefer to move generic Trojita talk to #trojita, though
16:40 * mzanetti doesn't mind
16:40 <mhall119> works for me
16:40 <popey> ok. here for meetings, there for general discussion outside meeting hours. Done.
16:40 <mhall119> meetings here, everything else in #trojita
16:40 <popey> \o/
16:40 <mzanetti> ack
16:40 <DanChapman> mhall119: could i work with you on the Cmake stuff, i'm still trying to get to grips with it?
16:40 <mhall119> DanChapman: I'd be happy for any help, I'm learning it as I go :)
16:41 <DanChapman> cool :-)
16:41 <mhall119> popey: are we going ot have a blueprint for this so we can track progress on status.ubuntu.com?
16:41 <popey> I'd rather not tbh.
16:41 <mzanetti> :D
16:41 <popey> it's not technically an ubuntu project. we're "just" enabling community people.
16:41 * mzanetti is still unsure how to start
16:42 <mhall119> so where are we going to track work items, upstream bugzilla?
16:42 <popey> https://projects.flaska.net/projects/trojita/roadmap ?
16:42 <jkt> that would require one more account for you
16:43 <jkt> bugzilla is my preference
16:43 <mhall119> what's the bugzilla URL?
16:45 <jkt> https://bugs.kde.org/
16:45 <mhall119> thanks, I'll get that on the wiki page too
16:46 <mhall119> is everybody subscribed to the trojita mailing list?
16:46 <DanChapman> not yet, i'll do that today
16:47 <mzanetti> no, not yet
16:47 <mhall119> please do, I'll email that once I have a working cmake for the Ubuntu port
16:47 <mhall119> then others can start actually building the port
16:47 * popey subscribes
16:48 <mhall119> boren: how hard do you think it will be to convert your current UI code to work like the Harmattan port?
16:48 <mhall119> is it worth tryingn to convert it, or would it be easier to re-implement it?
16:48 <boren> mhall119
16:48 <boren> mhall119
16:48 <boren> I would say re-implement
16:49 <mhall119> ok
16:49 <mhall119> popey: can I give you a work item to put out a call for designs?
16:50 <popey> yes
16:50 <mhall119> ok, I think that's all the "first steps" that need to be done before all of the "just implement it" stuff, did I miss anything?
16:51 <popey> mhall119: will you update the wiki page with links to bugzilla etc?
16:51 <popey> mailing list etc
16:51 <mhall119> popey: yup
16:51 <popey> sweet
16:51 <mhall119> and upstream repos, and process for submitting patches, etc
16:51 <boren> mhall119: or I can try port my UI to use Harmattan backend and see if it works.
16:52 <mhall119> boren: not the Harmattan backend, but the Trojita components that Harmattan uses
16:52 <mhall119> like ImapAccess
16:52 <jkt> yes, that's right
16:53 <boren> mhall119: Sorry, that is what I mean.
16:53 <jkt> I don't care that much about the JS and QML bits themselves; I do accept that they will need quite some love
16:53 <mhall119> it looked like you were taking a similar approach, but you treated the Trojita code like a plugin, and your QML as the main app, rather than the Harmattan approach of treating Trojita as the main app and Harmattan as a UI that is dropped on top
16:53 <jkt> one more thing -- don't waste time with porting the QNAMWebView
16:53 <mzanetti> so did we agree to start from current master?
16:53 <mhall119> mzanetti: yes
16:53 <mzanetti> ack
16:53 <jkt> the Qt5's WebView is very different, and one should use URL handlers for the trojita-specific schemes
16:53 <mhall119> mzanetti: I've just cloned it, and I'll start by applying kenvandine's cmake changes
16:53 <jkt> I believe mhall119 has some code for that
16:54 <mhall119> I do
16:54 <mhall119> we might also use the new UbuntuWebView if that gives us additional advantages
16:54 <mzanetti> we probably should
16:55 <mhall119> then again, whatever boren did was also very nice, especially with inline quoted parts
16:56 <mzanetti> yeah... I don't know the new one yet. but for consistency with the browser we should at least evaluate the possibility to use it
16:56 <mhall119> but that's all implementation, we'll cross that bridge when we get to it
16:56 <mzanetti> yep. so how do we avoid conflicts?
16:56 <mhall119> with what?
16:56 <mzanetti> anyone creating a bunch of work items?
16:56 <mhall119> mzanetti: https://docs.google.com/a/canonical.com/document/d/1MlGUefZxlYxJzwvIlD0LzhrUA8T4ulVtHA7Fb78RPyY/edit#
16:57 <mzanetti> ah ok, cool
16:57 <mhall119> popey: anything else you had in store for this meeting?
16:58 <popey> Not really. I envisaged it mostly as a "meet and greet" and "bootstrap"
16:58 <popey> Making sure we covered jkt's concerns.
16:58 <popey> [DONE]
16:58 <popey> and gathered documentation on what next, and where the team can find out where everything is.
16:59 <popey> I'll ping a mail out to everyone with the links after the meeting ends.
16:59 <popey> Next meeting, friday 21st at 14:00 UTC
16:59 <popey> Anyone else got anything?
17:00 <DanChapman> nope, i'm happy so far
17:00 * jkt is happy, too
17:01 * mhall119 is good
17:01 <popey> \o/
17:01 <popey> in fact this is a pom-poms type of happy. *\o/*
17:01 <popey> Thanks everyone. I'll ping a mail out shortly.
17:01 <mhall119> thanks everyone
17:01 <popey> Have a great weekend.
17:02 <popey> #endmeeting