Re: What goes in gnome office ?

[Martin Sevior <msevior mccubbin ph unimelb edu au>]

On Wed, 10 Apr 2002, Rui Miguel Silva Seabra wrote:

> On Wed, 2002-04-10 at 14:13, Mikael Hallendal wrote:
> > ons 2002-04-10 klockan 15.00 skrev Jody Goldberg:
> > > Definite Apps
> > >     Abiword
> > I'm curious, what is the Abi-developers thoughts about this. At least in
> > the past I've gotten the feeling that they are not to keen on
> > integrating more with GNOME (since they want to be
> > super-multi-platform). Or could this be done as a GNOME-only-addition?
> 

That is an interesting statement! We sometimes get the feeling Gnome
doesn't want us because we're not Open Office.

I'm all in favour of using all the great tools Gnome has available on the
gnome platform. We actually do make extensive use of the gnome platform
already- drag'n drop, gnome-vfs, gnome-print, gdk-pixbuf, librsvg, gimp,
gdict, esd, galeon and very soon bonobo.

Subclassing and plugin's mechanism make it easy to add gnome functionality
without having to port to all our supported platfroms. I'm not very
interested in porting gnome apps or libraries to windows though.

Regarding glib, pango and freetype since these are already portable
libraries I don't have an issue with moving these into the xp
layer. However since Abiword is a C++ program, it makes much more sense
for us to to continue to mostly use our own xp C++ implementations of hash
tables, vectors (lists) and string types. Having used both glib and Abi's
own tools, I find the abi tools much more more convient.

All in all I suspect code sharing can occur easily at the library level
(gnome-print, pango, freetype) and at the high level bonobo level eg
GtkMathView for maths, abiword as a embeddable word processor
widget (this works nicely now) Guppi, SVG render(s) etc.

I read the discussion about the general escher drawing layer. I can't help
but think it would be too much overhead for many apps. Not just in the
sense of being big and slow but also in the sense of taking too long to
get right. The next itteration of abiword will add lots of high level
features that do not need anything as sophisticated as an escher drawing
layer. I would hate to see us delayed by diverting resources to build that
or worse being delayed while some high level feature we would never used
is implemented.

Cheers

Martin