[GRASS-dev] trunk r47474 breaks OGSF rendering of points on surface

[Hamish]

> > Martin wrote:
> >> Also note that nviz is going to be removed from trunk when
> >> wxNviz will be completed (I hope soon).
Hamish:
> > well, no. there's no reason to remove it in the short term,
> > even as the wx version matures nicely. an optional tcl/tk
> > dependency in the source code costs very little, and is
> > useful to some folks.

perhaps to phrase my words better: I would rather appreciate it
if you didn't remove code which I'm actively using.

by "short term" I mean just for now, when we are near to being
ready for 7.0 to be released of course we can evaluate the
situation and see if we are ready to drop tcl/tk nviz from the
code base. but until then there is no reason why we can't have
more than one way to do something. I'm certainly not asking
anyone to take on the burden of maintaining anything more than
what interests them. until then, this is pretty hard to beat for
a fast typist:
 GRASS> nviz dem color=ortho points=stations

Martin:
> AFAIK there was consensus about droping TCL/TK dependency
> in GRASS 7.
> There is no reason to keep TCL/TK dependency,

mainly what I request is the way we go about doing that, and how
we communicate important changes to the group before they happen.

Thus when we are ready to call wxNVIZ feature complete, only then
that --with-tcltk in ./configure default to "no", but no code be
removed, and importantly do not drop tcl/tk support from the
build system just because the main GUI items don't use it any
more.  Others are very likely to have 3rd party tcl/tk code for
grass they have written over the years and have no interest in
porting over to wxPy. Similarly we have --with-blas and
--with-lapack which AFAIK are mostly unused by us, but cost
little to keep there and probably help some people out a lot.


Asking for a go-slow approach costs others much less than asking
folks to revert the deletion of some source code they need/want,
& it doesn't hold anyone back.


cheers,
Hamish