Hi Daniel,
On Apr 26, 2010, at 16:52, Daniel Ehlers wrote:
> Hi,
>
> first thanks for the hint, we started to look into your bindings and
> start playing around with it. While playing aroudn we found that you
> haven't implemented much of the nearby libraries to webkitgtk, libsoup
> and JavaScriptCore, for the last I found comments in WebFrame.chs.pp.
> For libsoup, we had started to bind basics of this library, so my
> question: Are there any plans to add libsoup to the gtk2hs repo?
I'm currently trying to move the non-core Gtk2Hs libraries into their
own repositories. The current darcs version now contains cabalized
version of glib, pango, cairo and gtk while the other libraries still
need converting (but it's on it's way).
> Maybe I can merge our work to gtk2hs, after some code review ...
>
Yes, it would certainly be better to have one binding.
> Now back to WebKitGTK, during the work on our bindings we found
> that many of the signal callback functions parameters should be
> peeked with maybePeek, for example resourceRequestStarting, which
> has following signature:
>
> resourceRequestStarting :: WebViewClass self => Signal self
> (WebFrame -> WebResource -> NetworkRequest -> NetworkResponse -> IO
> ())
>
There are probably many things like this that need adjusting. I've
added the possibility to have Maybe Object types as arguments to
signal handlers. If it is very common to pass a string that may be
NULL, I could add a new type for that, too.
> The NetworkResponse is not null iff the request is redirected,
> arcording
> to the webkitgtk api docs. I remember there are some more sticking
> points like this.
>
> And to repeat one of my questions last time:
> Is there any reason why gtk_*_get_type functions aren't binded?
>
They are bound. The are called gTypeBlah for any type Blah. In the tar
balls these version are still only available from
Graphics.UI.Gtk.Types. In the current darcs, I would like to avoid re-
exporting the Types module from Gtk and so I have included the
gTypeXXX functions into every module that defines a widget XXX.
I hope this helps,
Axel
> mfg Daniel
>
> On Sat, Apr 10, 2010 at 10:16:47PM +0800, Andy Stewart wrote:
>> Hi Daniel,
>>
>> We have finish webkitgtk binding in gtk2hs.
>>
>> Please check darcs version, under sub-directory "gtk2hs/webkit".
>>
>> Cheer,
>>
>> -- Andy
>>
>> Daniel Ehlers <danielehl...@mindeye.net> writes:
>>
>>> Hi,
>>>
>>> we, Andreas and I, currently developing webkit bindings for haskell
>>> on top of webkitgtk. We plan to have our first release next week,
>>> and currently writing doc and work up the last todos.
>>>
>>> Since we working with the gtk toolkit version of webkit, we are
>>> using much of the glib and gtk functions binded from gtk2hs
>>> into the haskell world. In the webkitgtk api some functions
>>> return gtk GObjects and for succesfully integrate them
>>> we have to query some GType informations. Sadly non
>>> of the gtk_*_get_type function we needed is binded in
>>> gtk2hs.
>>>
>>> Is there any reason why these aren't binded? Maybe
>>> there is a other way in gtk2hs we didn't see.
>>>
>>> Here is a code sniped to show one of the main points
>>> where we need those functions:
>>>
>>> webViewGetIMContext
>>> :: WebView
>>> -> IO IMContext
>>> webViewGetIMContext web_view = do
>>> imct <- {#call gtk_im_context_get_type#}
>>> objectGetPropertyGObject imct "im-context" web_view
>>>
>>>
>>> If there is no reason for them to be not binded, I would
>>> like to provide patches to integrate them into mainstream
>>> gtk2hs and get rid of the direct ffi/gtk+ in our package.
>>>
>>> Please excuse my cruel english.
>>>
>>> regards Daniel
>>> ------------------------------------------------------------------------------
>>> Download Intel® Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev_______________________________________________
>>> Gtk2hs-devel mailing list
>>> Gtk2hs-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/gtk2hs-devel
>>
>>
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> Gtk2hs-devel mailing list
>> Gtk2hs-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/gtk2hs-devel
> ------------------------------------------------------------------------------
> _______________________________________________
> Gtk2hs-devel mailing list
> Gtk2hs-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/gtk2hs-devel
------------------------------------------------------------------------------
_______________________________________________
Gtk2hs-devel mailing list
Gtk2hs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gtk2hs-devel
