[webkit-qt] Minutes from the Status Meeting in #qtwebkit on Freenode IRC network

[Luciano Wolf]

Hi Simon,

After sending my status report I was informed by Alexis and Jesus that
you were working on the same stuff as me. I was waiting to talk to you
tomorrow at IRC (in the meanwhile I was reading about CI system). I'll
comment each of your question/comments below.

2012/1/9  <simon.hausmann at nokia.com>:
> Luciano,
>
> what config files are you generating?
>
On my first attempt to compile QtWebkit I was using a vanilla tree.
Qt5 was compiled without knowing about QtWebkit so everything worked
fine. As I did another attempt to get a single "make" compiling
everything I realized that I should modify some file elsewhere to
avoid "qt5/qtbase/mkspecs/modules/qt_webkit.pri" being
copied/generated. This file was causing some confusion as the qtwebkit
build was failing. I don't know much about QMake/QtWebkit buildsystem,
so I was investigating before asking.

>
> I sat down on Friday and wrote a quick script to import a webkit snapshot into a Git repo and submitted
> it today into qt5.git ( http://codereview.qt-project.org/#change,12415 ). I also uploaded the script for the
> import into http://gitorious.org/qtwebkit/tools . Note that webkit in qt5.git is currently not CI tested, but
> I think we can get that working soon.
>
Very good! I did the same with phonon module. I don't know about
qtwebkit-examples-and-demos module and the "fix-webkit" script. Are
they necessary/up to date?

Are you using a plain configure && make or Ossy's build script to
test? I'm trying a new build using your version to see if I get that
problem with .pri file again.

> One thing I noticed today is we do have a circular dependency: WebKit uses uitools (from qttools) if available, and
> qttools needs WebKit for assistant. From what I can tell we use QUiLoader for a layout test that related to widget
> embedding, and QtTestBrowser exposes it, too, for widget creation. I propose we remove the latter and change
> the former (in DumpRenderTree) to manually create the widgets the layout test needs, in order to eliminate our
> dependency to uitools.
>
I think someone with more know-how can comment on this one, but it
seems right to me.

> Simon
> ________________________________________
> From: webkit-qt-bounces at lists.webkit.org [webkit-qt-bounces at lists.webkit.org] on behalf of ext Luciano Wolf [luciano.wolf at openbossa.org]
> Sent: Monday, January 09, 2012 18:30
> To: Qt WebKit StatusBot
> Cc: webkit-qt at lists.webkit.org
> Subject: Re: [webkit-qt] Minutes from the Status Meeting in #qtwebkit on Freenode IRC network
>
> * luck status: back from vacation. Taking care of qtwebkit inside Qt5
> build system - my first attempts to compile were successful because I
> was using a vanilla qtwebkit. Now I'm dealing with config files to
> avoid invalid files to be copied/generated inside Qt5 tree.
>
> On Mon, Jan 9, 2012 at 11:16 AM, Qt WebKit StatusBot
> <qtwebkit-statusbot at openbossa.org> wrote:
>> Updates:
>>  * jprvita status: still looking on minibrowser bugs
>>  * elproxy status: seeing progress on the rawfont front, saving a lot of time thanks to tronical's black magic :)
>>  * bbandix status: fixed event delivery in tests for the pinch zoom patch and re-landed it, updating and completing flickable patch to be able to drop qscroller/widgets dependency
>>  * ahf status Waiting for bug 75807 to land from the CQ, working on porting the InverseMouseArea QML type (bug 75536).
>>  * azbest_hu status: n9 scratchbox build problem with current sources
>>  * ossy status: weekly meeting, try to reproduce a build-webkit bug (clean build feature)
>>  * reni status: trying to create the proper document hierarchy related to the recursive external resources (SVG)
>>  * torarne status: got harmattan components working with qt5 on device, solution for crasher and text rendering issues. now to fix the integration issues like orientation etc
>>  * mibrunin status debugging weird webprocess crash
>>  * kkristof status: work on parallel http tests running with nrwt
>>  * kling status: chilling
>>  * zalbisser status: looking into key event handling / fixing shortcuts in MiniBrowser on linux&mac.
>>  * jeez_ status: webcontexting. Having issues to "move" custom scheme stuff to componentComplete init mode. Will upload WIP patches soon to get some feedback.
>>  * kbalazs status: testfont bug
>>  * kenneth_ status: still fighting with how to do the popup/dismiss vkb change in webkit in an acceptable upstreamable manner...
>>  * tronical status: observing black magic in the oslo office lately, continuing the v8-out-of-qtbase move, integrated webkit snapshot successfully into qt5.git (yay)
>>  * cmarcelo status: snowshoeing, janitoring (and learning a lot!), now looking for things in the garden to fix.
>>  * jturcotte status: reading emails, trying to remember what I was doing.
>>  * lmoura status: fighting uic crashes blocking the n9 packages. fighing "random" wtr crashes on my machine 洵
>>  * darktears status :  hacked on https://bugs.webkit.org/show_bug.cgi?id=75714 (Remove inheritance of CSSValue from RefCounted.) to free some bits for upcoming border-radius patch ; it happens to be a not very good idea, then I uploaded https://bugs.webkit.org/show_bug.cgi?id=75841 (Extend CSSValueList to allow slash separated lists.) which was r+'ed so now I can finish the border-radius patch, which will be more trivial now. That would lead to close
>>  * carewolf status: Scrolling integration
>>
>> Missing updates from: Ossy, Smith, TwistO, Zoltan, andris88, hugopl, igoroliveira, laknudse, loki04, luck, mulvad, rafaelbrandao, setanta, tczene, zalan, zherczeg
>>
>> _______________________________________________
>> webkit-qt mailing list
>> webkit-qt at lists.webkit.org
>> http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt
>>
> _______________________________________________
> webkit-qt mailing list
> webkit-qt at lists.webkit.org
> http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt