Re: Docking merged into GTk / Anjuta 1 being ported

[Biswapesh Chattopadhyay <biswapesh_chatterjee tcscal co in>]

> I would like to suggest to try to integrate  the code for the docking
> system in Anjuta 2 into GTk, because I think it should  be part of the
> overall user interface and is as usable as docking in MFC  apps.
> Further I wonder why Anjuta 1 is being ported to  GNOME2, because Anjuta 2
> looks great and just lacks some features that are  easily added. And in my
> opinion the great advantage of Anjuta2 is its  modularity.

1) About docking, an attempt was made to merge it into GTK+ in the long
term by moving it into libegg by Gustavo. I'm not sure what the progress
it, but Havoc/GTK team wanted the following:
	a) Investigate the dock widget API of various toolkits (KDE, .NET (i.e.
Windows.Forms), QT, MFC, Java).
	b) Come up with an API that incorporates the best parts of each and is
flexible enough to remain API/ABI compatible for a long time without
compromising on features.
	c) Post the API and let others comment on it.
	d) Move the code to libegg and let it stabilize there.
	e) Move to GNOMEUI/GTK+ depending on dependencies (For example, if
you're using GConf to store the widget positions for GdlDock, then you
can't move it to GTK+ - you'll have to move it to GNOMEUI)

The above approach definitely makes sense, but it tends to slow things
down. I'm not sure how far Gustavo has progressed but I haven't seen any
API analysis posting on GtkDevel/GnomeDevel - so my guess is not much.

2) About Anjuta1 being ported to GNOME2, it is a short-term solution
till Anjuta2 catches up on the feature-set, etc. Porting from GNOME1 to
GNOME2 is not as big a deal, unless you want a *real* port using all the
latest GNOME2 technologies with ATK, Pango, etc. and removing all use of
deprecated widgets. In fact, most of the Anjuta1 modules are already
ported - only the main src module remains and that too is almost 30%
done AFAIK.

Rgds,
Biswa.