Manuel Bouyer <bouyer@antioche.lip6.fr> writes:
> On Sun, Apr 09, 2000 at 04:20:28AM +0200, Hubert Feyrer wrote:
> > In retrospect i have to say that the idea of renaming the ghostscript most
> > people use and expect was really suboptimal. A gsnoX wouldn't have
> > required all these changes. ;(
> 
> Yes. And I didn't think about this when I saw the change but now I'll have
> to teach my users to use gsx11 instead of gs ... Much harder than fixing
> softwares.
> 
> Can we eventually change this ? (i.e. gs->gsnoX and gsx11->gs) ?
> We could eventually keep gsx11, and just rename gs to gsnoX, so that it's
> at last possible to put a symlink.

Good point.  In retrospect, keeping the interactive program with the
old name makes more sense.  I see two choices:

	1) gsx11 -> gs, and gs -> gsnoX/gstty (any other good names??)

	2) Make gs a shell script that parses its arguments looking for
	   -sDEVICE=x11* and calls ${PREFIX}/libexec/[gs|gsx11]
	   appropriately.

Which is better?

     -- Johnny C. Lam <lamj@stat.cmu.edu>
        Department of Statistics, Carnegie Mellon University
        http://www.stat.cmu.edu/~lamj/