[Xcb] [RFC v-1] API with 64bit seq. and more love for requests

[Bart Massey]

I think Jamey's proposal is different from my older one--as far as I
understand it he's proposing to bury the queues in the XCB
backend--but I think in a lot of ways his is better. So, yeah. Let's
do that. :-)  --Bart

On Tue, Feb 26, 2013 at 11:38 PM, Daniel Martin <consume.noise at gmail.com> wrote:
> On 26 February 2013 08:37, Bart Massey <bart at cs.pdx.edu> wrote:
>> As far as "event-driven programming" goes, the
>> never-seriously-implemented plan was always to have a separate layer
>> above XCB, perhaps in a separate thread, that was an event queue and
>> dispatch mechanism.
>
> Do you mean this mail by Jamey (thanks Peter for pointing me there)?:
> http://lists.freedesktop.org/archives/xcb/2010-September/006480.html
>
>> I've got nothing against event queues (OK, maybe,
>> but let's pretend); I just don't think they should be buried in a
>> library that's supposed to be a low-level protocol binding.
>
> With "more event driven" I didn't thought of new queues. I just wanted to
> add an XPending() equivalent, which reports replies too.
>
> But, I must admit, after reading this proposal of Jamey, I like the idea of
> those first-class queues. ;)