Re: [oglbase-discuss] OK, back to Re: 2nd draft:GL_EXT_get_proc_address spec

[akin]

John Dennis wrote:
| 
| I think my proposal did indeed have the property of static offset for ALL
| functions, there would be no use of a run-time offset. ...

You're right.  I apologize for obsessing about dynamic offsets.  :-)

So the cost of the mechanism you proposed is that the drivers need to
be able to update variable locations in the dispatch table, rather
than fixed locations that are known at compile time, when state
changes; plus any extra development effort needed to change existing
code to support that.

I don't have a good sense of the tradeoff, but I guess we can take the
discussion offline.  As far as the Linux OpenGL Base proposal goes,
the main question is whether we want to guarantee that GetProcAddress
always returns the same address for a given function name, no matter
what the current rendering context is.  I agree with you that we
should make that guarantee.

Allen