On Mon, 8 Nov 1999, Bernd Kreimeier wrote:
> "Michael I. Gold" wrote:
> > I am going to make one more appeal for context dependent pointers now.
>
> I second that. If nothing else, the amount of discussion
> might just indicate that we do not fully understand the
> implications of context independence here.
>
> Why not tread carefully? The noop function return as well
> as the context independence seem attempts to create a
> fool-proof mechanism - with all the intricate complexity
> such attempts invariably generate.
I agree. I like the *idea* of context independence, and
it would certainly make life easier for the OpenGL newbies
out there.
However, OpenGL is not ever likely to be bullet-proof to
broken programs...I don't see the benefits of all this
work to make it semi-bullet-proof.
When I actually look at my application code, context
independance really doesn't buy me a lot. I need a
table of which functions are allowed in which contexts
- and it's just as easy to store the address of the
function for each context into the same table. I don't
care how long it takes to build the table because I'm
only ever going to do that when I open a new context
- which is never a 'hard realtime' event.
However, getting this in as an ARB blessed extension is
important, and having something that's reasonably easy to
implement (so it gets done right first time on every
implementation) is even more important.
That's why I voted for context-dependance in the first
place. It's "good enough" - so I don't care either way.
Steve Baker (817)619-2657 (Vox/Vox-Mail)
Raytheon Systems Inc. (817)619-2466 (Fax)
Work: [EMAIL PROTECTED] http://www.hti.com
Home: [EMAIL PROTECTED] http://web2.airmail.net/sjbaker1
