ext Alberto Garcia wrote:
On Mon, Sep 07, 2009 at 05:09:25PM +0300, Kate Alhola wrote:
The importance and the goal of Hildon is to provide a set of essential
widgets so
a) app developers don't have to waste their time in writing them
themselves
b) there is UI consistency between all applications
If there are no Qt equivalents for all Hildon widgets, none of
these two problems are solved, no matter how easy it is for
developers to write their own widgets or how compact the code is.
If you think that if there is not 1 to 1 equivalent for everything,
there is nothing.
No, I haven't said that.
What I say is:
* The Fremantle UI style depends heavily on a set of widgets that
have been specifically designed for it.
* These include some very fundamental widgets such as HildonAppMenu,
HildonPickerButton and HildonStackableWindow.
* If you take a look at the N900 you'll see that these widgets are
used ALL OVER THE PLACE.
* Example: there's no application in the N900 using a menu other than
HildonAppMenu.
* If you want to use a menu in your application you must use
HildonAppMenu or a widget designed to mimic its look and feel, else
your application will look different.
* There's nothing necessarily wrong with that (e.g. Canola), but
developers should be aware before starting to write their apps.
I agree that. Antonio just answered, you have HildonAppMenu and
StackableWindow functionality in Qt.
They look exactly same than Hildon equivalents. I said already that you
will soon see also
PickerButton functionality. Qt API is different than GTK+, that's clear
to everyone.
So:
* Do you want to provide Qt libs for developing Fremantle apps? Good
* Are all the widgets that have been designed as a central part of the
Fremantle UI available in Qt? Good
* Aren't they available yet? Fair enough, but then make sure that
developers are aware of this.
Try to understand difference between functionality and widget.
In most of cases you don't need any special widgets.
You have StackableWindows and AppMenus but you don't
need to have separate widgets for this.
Let's get back to roots once again. There is two ways to implement
UI for mobile.
1- You can make new widgets ( Hildon)
2- You can enhance existing widgets to support mobile usecase (Qt)
The model used in Qt makes much easier to port applications because
it minimizes needs to modify your application.
The bottom line:
* Telling people that it's completely reasonable to write Fremantle
apps in Qt without making clear that some fundamental Fremantle
widgets have not been written yet it not a good idea IMHO.
Telling is very good idea, it leaves freedom of choice to developer.
I think that telling about choices and their consequences is allways
good idea. Maemo Qt is at the moment as Beta whereas Hildon for maemo 5 is
about final. That is one argument. The other argument is that Harmatan will
be Qt based, S60 will have Qt, Qt is a cross platform toolkit .Qt has
QGraphicsView,
animation API etc for eye candy, Qt has OpenGL-ES2.0 support, it has
webkit ...
It is also other question that which is more wasting time, writing
couple of dozen lines when you can save couple of thousand lines in
all application by more compact and efficient code with C++ and Qt.
I'm not going to start a C vs C++ debate, but I don't think this is an
argument here since there are already C++ bindings for Hildon:
http://maemomm.garage.maemo.org/docs_unstable/tutorial/html/sec-TouchSelector.html
It's just not C++ issue, Qt is written from scratch as C++ classes and
it just
makes many things like writing derived classes much easier.
I know the C++ bindings, I was the one pushing maemo to support C++ .
Kate
_______________________________________________
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers
