Hello,
I just subscribed here on Amiga.org only to throw in some sentences into this Thread.
I've been following this post since yesterday and also saw similar posts showing up even on other sites. I think that we should ask ourselves first for what we actually need GTK+ ? As we saw in the first post the person asked for GTK+ but don't had a particular idea what apps use GTK+ only. He named Mozilla and Open Office here as bad examples, since Mozilla uses XUL for Widgets stuff but uses GTK+ for some lowlevel stuff and Open Office is mainly SFC but Ximian took it and start GNOME'ifiying it. They are basicly changing some visible aspects over to GNOME and changed the filehandling to GNOME-VFS - I call it totally breaking a working application. I say this because I know that most of the work these people do are mostly done half. They take good applications, change it over to whatever they think but do it only half. That is while they change some visible aspects over to GNOME the rest stays in the old code it's a mismatch of old and new stuff and probably remains so for the upcoming years.
I do come from the GTK+/GNOME world and spent the past 5 years with them (that is I know the majority of the core developers working on GTK+ and even GNOME (to extend it) and I use to write own stuff using GTK+ and fix around in the GNOME world etc.
We should really ask ourselves if it's worth the trouble with GTK+ and what exact benefits it will give us or if there are native apps we can use. Right now I only have a few in my mind. The GIMP, XChat, GAIM maybe the one or other app as well. Most other apps available using GTK+ are usually also using GNOME libraries these days and porting all that stuff over to Amiga (be it AmigaOS, MorphOS or AROS) is more than a pain in my opinion.
If we get GTK+ ported then what ? A million apps that can easily be compiled ? I don't think so, since many of these programms usually depend on again other libraries and other libraries depend again on others.
Application <-> Library dependency
Library <-> Library dependency
Satisfying all these dependencies will be a hard work to do.
Porting GTK+ itself shouldn't be a big problem (theoretically) and you don't even need X11 for it or any XServer solutions.
GTK+ architecture is divided in the widgets library and the GDK backend. One of these GDK backends is for X11, one for Win32, one for DirectFB and in my understanding it could be easy to write an AmigaOS backend for GDK and the rest should more or less compile.
The current dependency of GTK+ 2.4 is as follows:
PKG-Config (probably optional or removable), Glib, Atk+, Fontconfig, Freetype2, Pango, GTK+
Although a bunch of GTK+ only applications already are written in a way where the Toolkit is just optional. That is you can easily switch between true Windows MFC or GTK+ or QNX Photon etc. And it should be no problem (only work is required) to write an MUI or Reaction frontend for it as well.
So it's questionable whether it makes sense porting the entire GTK+ widgetset including all its dependencies over to Amiga only for a few apps. It would be better to port these apps directly instead.
If someone had in mind to get stuff like Evolution running (only one example) since it is a GTK+ app then I must disappoint you since Evolution is not a GTK+ app it's more or less a GNOME app which requires around a good dozen other libraries to be ported (libraries that depend on other libraries as well). Honestly it would be easier to use Linux in this case.
We should also look at the philosophical aspects here (I know Mr. Frieden already replied to me that there isn't room for this). I for my own think that Open Source and Free Software does give us a good push getting new software to our systems and this imo is a nice thing. But we should clearly differ between core stuff such as compilers, debuggers etc. and GUI stuff. There are no benefits turning the entire Desktop experience from once AmigaOS look and feel over to a GNOME or GTK+ look and feel.
You should also think about that. A lot of people simply dislike or don't agree (anymore) in the way how GTK+ or GNOME apps use to work and they are searching for alternatives with it's own philosophical spirit. Not to mention that there isn't anything special around GTK+ either.
We should clearly concentrate writing either own software solutions or port the good bits from Open Source and the Free Software world over to our systems but then we should do it in a way that makes the app fit entirely into the Desktop look and feel experience. We already have to many toolkits on Amiga and we should seriously concentrate on MUI or Reaction these days. No need for another toolkit.
Sorry if I as only one sound contra GTK+ here but I am a signed person here. It's not just GTK+ itself it's also the entire community of these people that I dealt with the past 5 years. It was clearly enough.
greetings,
oGALAXYo