On Thu, 2006-03-02 at 09:09 +0100, Robert Staudinger wrote:
> Hub,
>
> there were discussions in #abiword before you introduced that stuff.
> And people have expressed their reservations. I reiterate that IMO the
> one size fits all header abstraction with switch of behaviour at
> runtime depending on the class of the wrapped widget is a rather
> limited approach. Also at the current rate of one new dialog every few
> months I can hardly see the new framework taking over evolutionarily.
> Regarding rewriting working dialogs just to save a few lines of code,
> hmm, kudos if you can motivate anyone to do that.
Given that I ported the code to Cocoa, at least large parts, I have a
good idea what is going on. The platform glue code for dialogs is just
"set widget content to foo", "enable widget bar", etc.
I know Cocoa programming, I know Qt, I know Gtk, and I have had a little
experience in Win32 in a past life. I can tell you that that they all
have a "set widget text to "foo"", "disable widget bar", etc. And that
can be abstracted. After all we have the same set of data being use in
dialogs (with a very few excpetions in the platform) and the same UI
logic. So why shouldn't we share most of the code?
My 'framework' is not perfect, nor it is complete, but it is there, and
tries to address some problems.
> That said i'd very much favour a i18n solution for AbiWord which
> allows the use of the default strings in glade instead of "%s" (i18n
> accounts for a bulk of boilerplate code). Maybe when the planets align
> ...
That could be taken into account.
My framework start by moving the localisation up in the XP code as we
set the widget label through the abstraction layer that my framework
provides. Do you know how many lines of code I wrote on Cocoa to do
that? Too much.
But all in all, I can't push anybody to use it. I was just trying to
provide an elegant solution to a real problem.
Maybe we should switch all to gtk...
Hub
Received on Thu Mar 16 16:36:45 2006
This archive was generated by hypermail 2.1.8 : Thu Mar 16 2006 - 16:36:46 CET