LG events.c changes (was "Re: ")

[Jesse Barnes]

On Wednesday, January 4, 2006 11:53 am, Deron Johnson wrote:
> Paul Anderson told me that the processes for reviewing and
> adopting Xorg architectural changes are not well established
> so in some sense LG is a test case for this process.
>
> So, the specific issue at hand is as follows:
>
> + Should the LG event changes be done via function pointers?
>     Pros: Would group all of the LG event changes into one place.
>           Would minimize ifdefs.
>     Cons: First major structural change to event code in 2 decades.
> 	  Adds several new pointer deferences to the event pipeline.
>
> Please discuss amongst yourselves how you intend to go about
> resolving this issue and what the resolution is, and then let
> me know. Thanks.

Might I suggest instead that you volunteer to maintain the event layer?

It seems to me that X is sorely lacking maintainers at this point, and if 
you were willing to volunteer, that would be one less chunk of code for 
everyone to worry about (and after all, we know that shared 
responsibility is basically no responsibility).  You probably know it 
better than most anyway, and therefore probably have the best judgment 
about how it should be structured and maintained going forward.

As for the function pointer proposal in particular, it sounds like a good 
idea to me, but I'm definitely not an expert in that area so I have no 
idea if superior alternatives exist.

Jesse