On Fri, 19 May 2000, Jon M. Taylor wrote:
> It would be very easy to add these two functions quickly to the MISC
> extension, and lots of targets have accelerated cursor support.  LibWMH,
> LibGWT, LibXMI and Berlin could all use cursor functions....

Just to be utterly redundant, the sprite/bob/cursor/texture feature
management API not being developed is really what is holding GGI back
bigtime.  We need that more than we need 3D IMHO.

However, even after we had that system, which would handle
allocation/negotiation of such objects, in place we will want higher level
API's for each such object.  Andy made one such simple API.

I'd suggest the creation of "official" (i.e. bundled in the tarball)
extension(s) which define and partially implement the basic actions that can
be performed on each major family of these objects.

"sprite" -- move, get size, set magnification, prioritize, show/hide, copy
            to/from buffer, copy to/from other sprite, collision.
"cursor" -- move, get size, set magnification, show/hide, copy to/from buffer, 
            hotspot, plus how to handle ROP pixels.
"bob" --    place, orientation/tesslation, get/set size, full ROP suite.

...textures leave to the 3D folks.

By "partially implement" I mean we do not specify how to obtain the objects in
the first place.  Instead we tell the app developers that for now, until the
feature negotiation stuff is done, they should implement their own code
to create the objects they need on the targets they do their testing on and 
tuck it in a subroutine which will later be replaced by calls to the feature 
management API.  Any code they generate to actually implement the above 
though should go in the extension.

Anyone who hasn't read the conversation about the feature management API,
please see the list archives for anything involving "feature negotiation" or
"sprites."  It would be nice if the above extensions were designed to 
have the feature management stuff neatly slid in underneath them.

--
Brian

P.S. Unless of course we have now the talent and teamwork to actually do the
feature API.  I know I've been pretty useless except to talk about it.



Reply via email to