Python byte-compiled files

[tlaronde at kergis.com]

On Sat, Mar 23, 2024 at 11:02:33AM -0700, Alan Coopersmith wrote:
> On 3/23/24 09:24, tlaronde at kergis.com wrote:
> > Furthermore, for example for xcb-proto, the python interface is one
> > interface, but it is not mandatory (there is the primary C interface).
> 
> The python files in xcb-proto are not an alternative to the C interface
> provided in libxcb - they're the mechanism used to generate the C code
> for libxcb from the XML protocol descriptions in xcb-proto.  They're only
> used to build libxcb, not at runtime by applications/clients.

Uh! I did get it wrong... From the description, I gathered that it was
one way to make extensions, but optional (from README.md):

"Python libraries: xcb-proto also contains language-independent Python..."
                             ^^^^

The last sentence may correct this but...:

"Note that this is only necessary on machines where XCB is being built."

Could a native english speaker correct the README.md so that it is
more clear that the python stuff is mandatory for building libxcb? (I
process the modules "topologically", required first, so started by
xcbproto, and indeed there is c_client.py to generate needed files, in libxcb/src/;
I just saw the *.c files from a cursory look...).

> 
> The python alternative API to libxcb was xpyb, but it's long been abandoned:
> https://gitlab.freedesktop.org/xorg/lib/xpyb/
> and points to https://github.com/tych0/xcffib as a replacement now.

Thanks for the pointers. I will put this on the (too huge) stack to have a look because
a python dependency for compilation of C...
-- 
        Thierry Laronde <tlaronde +AT+ kergis +dot+ com>
                     http://www.kergis.com/
                    http://kertex.kergis.com/
Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C