[Xcb] Re: CRAY bitfield support in protocol headers: does anyone care?

[Egbert Eich]

Barton C Massey writes:
 > 
 > I think it's pretty hard to plan software around what some
 > really small hypothetical anachronistic community might want
 > or experience in the indefinite future :-).  Does anyone

True. Still, X is something that is not new and that is used
in very many areas. This legacy is a burdeon. I (infrequently) 
talk to people who tell me about use cases that originated in 
the distant past and still matter - cases that I never imagined.

 > know any person who still directly works with X on these
 > architectures, or any place we could look for such a person?

Maybe. These people usually work (live?) in environments that are
disjoint from our world ;-)

 > 
 > > I would think the client side would be more important than
 > > the server side though.
 > 
 > I agree. The normal use of this stuff is (was?) to use a
 > remote display to visualize stuff run on the Cray.  In that
 > case, I suspect that the dozen people who wanted this stuff
 > to work could play some games with the XML protocol
 > descriptions and the XCB dispatch layer to get libXCB-Cray;
 > we could probably leave the Xlib stuff alone, so they would
 > be happy after they did a little work, I think?
 > 

Cool, so if there is someone around who is able to hint these
people directions should they show up it's probably fine.
I never really went into understanding this stuff nor have I
looked deeply into the XML protocol descriptions so I trust
your judgement.
But I agree: leaving the Xlib stuff alone is probably enough 
to keep this community happy ;-)

Cheers,
	Egbert.