Re: [AD] Screen update API again

[Evert Glebbeek <eglebbk@xxxxxxxxxx>]

On Wednesday 25 August 2004 15:24, Angelo Mottola wrote:
> I mean, take your suggested screen update API code. It relies on 
> current functions to work. Will this change, so that we'll have for 
> example no more virtual screens (changing internal init() gfx drivers 
> hook) as they're not needed anymore? If we change it radically, the 
> whole set_gfx_mode code could be put inside the al_create_display 
> code... Then a small compatibility set_gfx_mode function could be made, 
> so that it calls al_create_display.

This is what I had in mind to do in the long run, yes. In fact, this may be 
something that is actually fairly easy to do.
I'll toy around with it some more in the next few days.

> Basically my question is: are we going to emulate the new API using the 
> old API calls, or are we going to emulate the old API using new API 
> calls?

My take, in this case on the graphics API, is to get a working version as 
soon as possible so that it can be tested. This means setting it on top of 
current functions.
Then, when consensus is reached about the API, the situation is inverted 
and the old API is reworked into a compatibility layer.

I think Peter followed the opposite approach with his timer/thread API. 
Does that work on MacOS X, btw?

Evert