- webkit-efl - lists.webkit.org

Re: [webkit-efl] Dropping support for the libcurl network backend
by Gyuyoung Kim 02 Feb '12

02 Feb '12

Hello ryuan, I'm sorry. I don't understand well what is your point. Anyway, if #if USE(SOUP) macro is problem, we can distinguish libsoup / libcurl file for each ewk file. For example, ewk_cookie_soup.cpp ewk_cookie_curl.cpp and so on. AS you know, ewk_auth already use this. We can select which file is built by make file. How do you think about it ? - Gyuyoung. ------- Original Message ------- Sender : ryuan Choi<ryuan.choi(a)gmail.com> Date : 2012-02-02 11:14 (GMT+09:00) Title : Re: [webkit-efl] Dropping support for the libcurl network backend Hi, guys. I'm not sure whether we can drop libcurl when libsoup is better. If I am right, ecore have options to support glib. I think that It means that glib is not mandatory on Efl world. Can we declare that webkit/efl is only available when ecore is compiled with glib? -- Ryuan Choi 2012년 2월 2일 오전 10:38, Gyuyoung Kim <gyuyoung.kim(a)samsung.com>님의 말: Hello Kubo, I agree with your opinion that soup macro clutter is growing. However, we don't test/profile which one is better yet. If libcurl's performance or features are better than soup, application may want to use libcurl instead of libsoup. I heard my co-workers have a plan to profile both libcurl and libsoup. Why don't we decide if we drop libcurl after the profiling ? -- Gyuyoung. ------- Original Message ------- Sender : Raphael Kubo da Costa<kubo(a)profusion.mobi> Date : 2012-02-02 01:44 (GMT+09:00) Title : [webkit-efl] Dropping support for the libcurl network backend Hi there. We've supported building webkit-efl with both the curl and libsoup network backends for quite some time. By default, the port builds with the libsoup backend, the layout tests (and the skipped list) assume the libsoup backend is used, and there is code in ewk (ewk_cookies, ewk_network and ewk_auth, for example) which only works in the soup backend. Given all that, the fact that the #if USE(SOUP) clutter is growing and the way these #ifdefs are causing bug 77341 to be more troublesome to integrate than neeed, I'm inclined to drop support for the curl backend. The only current drawback I see is that we'd force the dependency on unstable GNOME libraries for things to work. Thoughts? Does anyone desperately need the libcurl backend? -- Raphael Kubo da Costa ProFUSION embedded systems http://profusion.mobi _______________________________________________ webkit-efl mailing list webkit-efl(a)lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-efl _______________________________________________ webkit-efl mailing list webkit-efl(a)lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-efl

2 1

Dropping support for the libcurl network backend
by Raphael Kubo da Costa 01 Feb '12

01 Feb '12

Hi there. We've supported building webkit-efl with both the curl and libsoup network backends for quite some time. By default, the port builds with the libsoup backend, the layout tests (and the skipped list) assume the libsoup backend is used, and there is code in ewk (ewk_cookies, ewk_network and ewk_auth, for example) which only works in the soup backend. Given all that, the fact that the #if USE(SOUP) clutter is growing and the way these #ifdefs are causing bug 77341 to be more troublesome to integrate than neeed, I'm inclined to drop support for the curl backend. The only current drawback I see is that we'd force the dependency on unstable GNOME libraries for things to work. Thoughts? Does anyone desperately need the libcurl backend? -- Raphael Kubo da Costa ProFUSION embedded systems http://profusion.mobi

1 0

Re: [webkit-efl] Updating the layout test baselines to match Ubuntu 11.10's output
by Gyuyoung Kim 01 Feb '12

01 Feb '12

>> Right, one needs to have a web server and some other dependencies for all the tests to pass. >> GTK's list should be close enough to ours (we don't need xvfb, for example). Yes, let's define dependencies packages for running EFL layout test. :-) >> Do you mean you have manually built nettle, GnuTLS, glib, >> glib-networking, libsoup from git and/or recent tarballs to keep things >> working? I'm using some packages via tarball and other packages via git. I mean we need to mention minimum version for each packages in order to run EFL layout test in EFL layout test wiki page. >> P.S.: There is no need to CC me when replying, I'm subscribed to this list >> and can read the mails sent to it just fine. Ok, I will do that. -- Gyuyoung. ------- Original Message ------- Sender : Raphael Kubo da Costa<kubo(a)profusion.mobi> Date : 2012-02-01 02:28 (GMT+09:00) Title : Re: [webkit-efl] Updating the layout test baselines to match Ubuntu 11.10's output Gyuyoung Kim writes: >>> DumpRenderTree and the layout tests bring no additional dependencies, as >>> they are built with the rest of the port, so I don't think I understood >>> what you meant here. > > As you know, we have to install apache package to test http tests. In > addition, I installed more packages you let me know our private layout > test document. Right, one needs to have a web server and some other dependencies for all the tests to pass. GTK's list should be close enough to ours (we don't need xvfb, for example). >>> My original question still stands: are we OK with defining Ubuntu 11.10 >>> instead of Ubuntu 11.04 as the base platform to generate the test >>> baseline images (and for running the tests)? I am asking this because >>> the images are very likely to vary from distro to distro (and from >>> version to version) due to the way dependencies such as cairo or >>> fontconfig are compiled and the versions which are used. > > Personally, I 'm using Ubuntu 11.04 now. Do you mean you have manually built nettle, GnuTLS, glib, glib-networking, libsoup from git and/or recent tarballs to keep things working? > I wonder if other ports's layout test are optimized for Ubuntu > 11.10. Do you know ? I don't, but I don't think that matters. Could you elaborate? P.S.: There is no need to CC me when replying, I'm subscribed to this list and can read the mails sent to it just fine. -- Raphael Kubo da Costa ProFUSION embedded systems http://profusion.mobi _______________________________________________ webkit-efl mailing list webkit-efl(a)lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-efl

1 0

Re: [webkit-efl] Updating the layout test baselines to match Ubuntu 11.10's output
by Gyuyoung Kim 31 Jan '12

31 Jan '12

Hello Kubo, Thank you for your updating for http://trac.webkit.org/wiki/WebKitEFLLayoutTest wiki page. :-) >> DumpRenderTree and the layout tests bring no additional dependencies, as >> they are built with the rest of the port, so I don't think I understood >> what you meant here. As you know, we have to install apache package to test http tests. In addition, I installed more packages you let me know our private layout test document. If there are no need to install additional packages, I'm ok. But, I think we should verify it if layout test on EFL port can be run with only packages descripted in http://trac.webkit.org/wiki/EFLWebKit. As you know, GTK port needs to install additional packages in order to run layout test on GTK port as below, http://trac.webkit.org/wiki/WebKitGtkLayoutTests >> My original question still stands: are we OK with defining Ubuntu 11.10 >> instead of Ubuntu 11.04 as the base platform to generate the test >> baseline images (and for running the tests)? I am asking this because >> the images are very likely to vary from distro to distro (and from >> version to version) due to the way dependencies such as cairo or >> fontconfig are compiled and the versions which are used. Personally, I 'm using Ubuntu 11.04 now. I don't know how many people are working on Ubuntu 11.10. I wonder if other ports's layout test are optimized for Ubuntu 11.10. Do you know ? -- Gyuyoung ------- Original Message ------- Sender : Raphael Kubo da Costa<kubo(a)profusion.mobi> Date : 2012-01-30 21:11 (GMT+09:00) Title : Re: [webkit-efl] Updating the layout test baselines to match Ubuntu 11.10's output Gyuyoung Kim writes: > I made a wiki page for layout test of WebKit EFL. But, that page is not > completed yet. > (http://trac.webkit.org/wiki/WebKitEFLLayoutTest) I've improved the formatting and the text a little. > In my opinion, we need to define dependency list in the page with correct > version. DumpRenderTree and the layout tests bring no additional dependencies, as they are built with the rest of the port, so I don't think I understood what you meant here. My original question still stands: are we OK with defining Ubuntu 11.10 instead of Ubuntu 11.04 as the base platform to generate the test baseline images (and for running the tests)? I am asking this because the images are very likely to vary from distro to distro (and from version to version) due to the way dependencies such as cairo or fontconfig are compiled and the versions which are used. -- Raphael Kubo da Costa ProFUSION embedded systems http://profusion.mobi _______________________________________________ webkit-efl mailing list webkit-efl(a)lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-efl

2 1

Updating the layout test baselines to match Ubuntu 11.10's output
by Raphael Kubo da Costa 30 Jan '12

30 Jan '12

Hi there. Our current set of layout test baseline images was generated in a Ubuntu 11.04 machine (Ubuntu 11.10 was released while we were working on DRT, so we didn't switch at the time), and they haven't been updated very much lately. Building the port on a Ubuntu 11.04 machine is a bit too troublesome these days: besides building the unstable GNOME dependencies for the libsoup backend, one needs to build the dependencies' dependencies before that, such as GnuTLS, libnettle etc; this probably means people working on the port have already moved on to more recent distributions. I'd like to check with you guys if we adopt Ubuntu 11.10 (Oneiric) as the base platform for future baselines for now; if so, I'll spend some time next week working on bringing the current baselines and skipped list up to date. Note that in the future we may need to go for the nuclear option the webkit-gtk folks have adopted and just require a certain version of cairo, freetype and other libraries to make the layout test results be the same in any distro. -- Raphael Kubo da Costa ProFUSION embedded systems http://profusion.mobi

2 2

Radio buttons and checkboxes - surfaces do not render properly
by Michal Ciepielski 26 Jan '12

26 Jan '12

Hi, I've submitted a new bug to webkit bugzilla (https://bugs.webkit.org/show_bug.cgi?id=77101) Below is a description: I have found a bug in the current EFL Webkit port. I am linking to EFL 1.2.0 (checkout from svn r67333 19.01.2012). Radio buttons and checkboxes do not render properly. For radio buttons a whole group is checked at the same time and they cannot be unchecked. Checkboxes cannot be unchecked. I've found that the reason of this strange behaviour is that the surface is not cleared. You can fix this by clearing a buffer in RenderThemeEfl::paintThemePart (RenderThemeEfl.cpp) but obviously something has changed inside EFL framework since version 1.1.0 (checkout from svn r63607 26.09.2011) does not have this bug. Should I prepare a patch that clears the buffer? Best Regards, Michal The above message is intended solely for the named addressee and may contain trade secret, industrial technology or privileged and confidential information otherwise protected under applicable law. Any unauthorized dissemination, distribution, copying or use of the information contained in this communication is strictly prohibited. If you have received this communication in error, please notify sender by email and delete this communication immediately. Powyższa wiadomość przeznaczona jest wyłącznie dla adresata niniejszej wiadomości i może zawierać informacje będące tajemnicą handlową, tajemnicą przedsiębiorstwa oraz informacje o charakterze poufnym chronione obowiązującymi przepisami prawa. Jakiekolwiek nieuprawnione ich rozpowszechnianie, dystrybucja, kopiowanie lub użycie informacji zawartych w powyższej wiadomości jest zabronione. Jeśli otrzymałeś powyższą wiadomość omyłkowo, uprzejmie proszę poinformuj o tym fakcie drogą mailową nadawcę tej wiadomości oraz niezwłocznie usuń powyższą wiadomość ze swojego komputera.

1 0

Notification of code cleanup related with ewk_tiled_model.
by KwangHyuk Kim 13 Jan '12

13 Jan '12

Dear webkit-efl developers, As _ewk_tile_paint as well as _ewk_tile_paint_rgb565 and _ewk_tile_paint_rgb888 are never used in anywhere, I have created bug to cleanup the unused code on ewk_tiled_model. https://bugs.webkit.org/show_bug.cgi?id=76031 Best Regards - KwangHyuk

1 0

Re: [webkit-efl] Layout results of DRT
by Leandro Pereira 12 Jan '12

12 Jan '12

Kangil, On 01/12/2012 06:56 AM, 한강일 wrote: > What are the rule or standard EFL port would follow for rebaselining > result expecially in layout tests. > This should be made on a case by case basis: if a test is failing but should not, see what is causing this. Sometimes there's a bug that must be fixed, sometimes our baseline was wrong in the first place and for some reason the current output is now correct. Since each test is different (although they might share the same kind of output: colored rectangles, "passed"/"failed" texst, etc.), there isn't much one can do. There are rebalining tools that might help here, but I've never used them so I can't comment. A good rule of thumb would be watching other ports when they rebaseline a lot of stuff; this might (but not always!) indicate a change in behaviour. Leandro

1 0

Layout results of DRT
by 한강일 12 Jan '12

12 Jan '12

1 0

Re: [webkit-efl] Announcement of Coding Style Change in WebKit EFL port
by Tomasz Morawski 09 Jan '12

09 Jan '12

Hi, I see your point of view now I will make a huge patch instead of few smaller and open a bug for that. Best Regards, Tomasz > Hello Tomasz, > > In order to change coding style, we permitted to land a huge patch. But, > if this is not simple work, let's file a master bug for this issue. > > - gyuyoung. > > On Mon, Jan 9, 2012 at 5:13 PM, Tomasz Morawski <t.morawski(a)samsung.com > <mailto:t.morawski@samsung.com>> wrote: > > I think it will better when whole community will change our source > according to new/delete operator use not by single person. Maybe it > will be better to create a master bug for this task? Due to that > remove malloc/calloc/realloc calls and replace it them by smart > pointers/new/delete it is not simple task to do this in single > patch. There will be need to make some refactorization etc. > > Hello Tomasz, > > > Will you replace malloc/calloc/realloc with new / delete ? If > you want > to do this, please file a bug. > > > Cheers, > > Gyuyoung > > ------- *Original Message* ------- > > *Sender* : Tomasz Morawski<t.morawski(a)samsung.__com > <mailto:t.morawski@samsung.com>> Software > Engineer/Poland R&D Center-Linux (MSD)/삼성전자 > > *Date* : 2012-01-04 17:55 (GMT+09:00) > > *Title* : Re: [webkit-efl] Announcement of Coding Style Change > in WebKit > > EFL port > > W dniu 03.01.2012 13:27, Raphael Kubo da Costa pisze: > > Tomasz Morawskiwrites: > > > >> W dniu 03.01.2012 12:49, Raphael Kubo da Costa pisze: > >>> Tomasz Morawskiwrites: > >>> > >>>> What about using new/delete operator for memory > allocation/free > >>>> everywhere inside internal implementation? For example: > >>>> > >>>> tile = static_cast(malloc(sizeof(Ewk___Tile))); > >>>> > >>>> I think it could be simpler and better to just write: > >>>> > >>>> tile = new EwkTile; > >>> > >>> That'd be really good. Plus using smart pointers wherever > possible :) > >> So, since there is general agreement for that, could we add > this to > >> EFL coding style? > > > > Sure. Could you add that to the Trac page? > > > It is done. > _________________________________________________ > webkit-efl mailing list > webkit-efl(a)lists.webkit.org <mailto:webkit-efl@lists.webkit.org> > http://lists.webkit.org/__mailman/listinfo.cgi/webkit-__efl > <http://lists.webkit.org/mailman/listinfo.cgi/webkit-efl> > > > _________________________________________________ > webkit-efl mailing list > webkit-efl(a)lists.webkit.org <mailto:webkit-efl@lists.webkit.org> > http://lists.webkit.org/__mailman/listinfo.cgi/webkit-__efl > <http://lists.webkit.org/mailman/listinfo.cgi/webkit-efl> > >

1 0