On Wed, 23 Oct 2002, Edwin Leuven wrote:

> It just makes me sad to see scarce resources (developer time) being directed
> towards the xforms frontend which in turn delays qt with in turn delays inset
> unification which in turn etc etc

We have three frontends.  I'm sure there are a few people (most
probably users or leaders of free software organisations) who are sad
that we are wasting developer time on XForms and Qt when we should be
getting the Gnome port working. Likewise OS-X users who wish we'd get
our acts together and switch to the best interface in the known
universe.

GUII gives choice to users _and_ developers.  The XForms port won't
die while people use and maintain it -- or want to use it as a
experimental playground.  Besides, the XForms port has such a
developed class structure (maybe a few refinements left) that porting
to FLTK should be a relative breeze -- start by copying the tree and
then fix the compiler errors -- and it should form a good base for
other ports too.  That's why it was always called the reference port.

Now that XForms has source available we should hopefully see distros
picking it up so it shouldn't be as obscure as it has been.  Oh and
one other reason: It was hoped that GUII would attract other
developers to the project as _they_ ported LyX to _their_ favourite
toolkit and hopefully stuck around to maintain that port and work on
other parts of LyX.  This aspect has been successful IMO.  It is wrong
to expect that LyX should be a single toolkit app.

FWIW, if Angus makes the XForms Document code use the controller he's
working in familiar (complete) territory and will not only fill in the
gaps in the existing Document controller but is likely to fix bugs
that a known working interface ported to use that controller will show
up.  Sure this doesn't get the Qt Document code written but it still
benefits Qt because once the controller is known to work your Qt
coding skills can complete the task quickly.  Likewise when he moves
onto MVCing Preferences (or someone else tackles it) that will benefit
Qt.  Maybe it might ease your mind to think of the XForms port as a
unit test rig for the Qt frontend -- if the XForms port continues to
function then the specifications have been met.

In summary, we all win if Angus directs his time and energy into the
stuff he's the expert at.

Allan. (ARRae)

Reply via email to