We were recently talking about a feedback mechanism for
process shared visuals, and someone mentioned that he was
considering working on it.
What's the state of this ? Are there chances that we'll see
this kind of enhancement to the shm memory visual in the
near future ?
I'm asking since I communicated with Raph Levien (current
maintainer of ghostscript), and I decided that I may want
to put my hands on a GGI driver for it. As always, ghostscript
would be executed as a child program, with the DEVICE variable
indicating the driver to use. Assuming that the calling process
is in control of the shm visual (and the pipe), ghostscript would
use the GGI driver (with that feedback mechanism) to draw into
it. Basically the exact same way ghostscript is deployed inside
gv or any other X postscript viewer.
Regards, Stefan
PS: for the curious: Raph is indeed considering to replace the
driver registration in ghostscript with a plugin mechanism,
so installing third party drivers should be possible soon.
PPS: given the ubiquitous use of ghostscript (especially with the
pdf features comming soon), I'm very keen on using it for
one of the first serious applications for berlin. But of course,
it would be before all a serious GGI application, too :)
PPPS: and while working on the GGI embedding feature, I came across
the 'giiJoinInput' function, which I use (as suggested by cube3d)
to return an input structure from the visual.
I strongly suggest adding a function 'giiGetInput' (or something)
which is a bit more intuitive to be used for that purpose.
(Should be a one-liner). Falling back to 'JoinInput' to *get*
the input looks like an aweful hack...
_______________________________________________________
Stefan Seefeld
Departement de Physique
Universite de Montreal
email: [EMAIL PROTECTED]
_______________________________________________________
...ich hab' noch einen Koffer in Berlin...
