[Xcb] XCB naming conventions

[Barton C Massey]

I'm willing to delay the release another week if everyone
agrees a name change is needed, so that we can get it all
straightened out.

Of course, we can open Pandora's box wider by asking about
the structure-based typing (e.g. cookie types); there's been
some talk that this should go also... :-)

	Bart

In message <20060913190453.GA30836 at id.minilop.net> you wrote:
> 
> --===============0216440475==
> Content-Type: multipart/signed; micalg=pgp-sha1;
> 	protocol="application/pgp-signature"; boundary="liOOAslEiF7prFVr"
> Content-Disposition: inline
> 
> 
> --liOOAslEiF7prFVr
> Content-Type: text/plain; charset=us-ascii
> Content-Disposition: inline
> Content-Transfer-Encoding: quoted-printable
> 
> On Wed, Sep 13, 2006 at 02:19:05PM -0400, Jeremy A. Kolb wrote:
> > xcb_window_new
> > xcb_glx_create_context.
> 
> I love this. But I still want to release on Friday, and I kind of hate
> to break everyone's code just before release. Of course, we're not going
> to break it afterwards, so now is the time if it's going to happen.
> 
> > XcbWindow
> > XcbGlxContext
> 
> I don't love this quite as much. I'd be OK with following Cairo's
> example of xcb_window_t or something.
> 
> This should let us eliminate the distinction between *Rep, which is a
> type, and *Reply, which is a function -- using *_reply_t and *_reply
> would be nice. If there are more abbreviations in the auto-generated
> names, we should eliminate those too: for example, *Req would become
> *_request_t.
> 
> Thoughts on CARD8 etc.? It would be good if we were using a different
> name than Xlib uses: the current hack to make that work is painful.
> 
> > We think this would be much easier to read.  What's everyone think?  This=
> =20
> > would require changes to ecore/x11/mesa/xcb-utils etc. But nothing a good=
> =20
> > sed script wouldn't be able to handle.
> 
> If we have a prototype source conversion tool by, say, about 8am UTC
> tomorrow, then I'd seriously consider this proposal.
> 
> I also want feedback from others on how awful this mechanical change to
> the API is for them though.
> 
> It'd help a lot if someone also put together the patch to c-client.xsl
> that generates the desired naming convention, and posted the
> git-format-patch output to the list. Note a complete solution in the XSL
> probably requires making sure it can find word boundaries in xcb-proto
> names. That may require small changes to some xcb-proto declarations.
> 
> --Jamey
> 
> --liOOAslEiF7prFVr
> Content-Type: application/pgp-signature; name="signature.asc"
> Content-Description: Digital signature
> Content-Disposition: inline
> 
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
> 
> iD8DBQFFCFZVp1aplQ4I9mURAua1AJ9WqLiT4rXLdRnw3c/14w4i7zkVNACfVd7Y
> 0K8A2aBEdUXHEco535y+TU0=
> =d9Zm
> -----END PGP SIGNATURE-----
> 
> --liOOAslEiF7prFVr--
> 
> --===============0216440475==
> Content-Type: text/plain; charset="us-ascii"
> MIME-Version: 1.0
> Content-Transfer-Encoding: 7bit
> Content-Disposition: inline
> 
> _______________________________________________
> Xcb mailing list
> Xcb at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/xcb
> --===============0216440475==--