Index: [thread] [date] [subject] [author]
  From: WHS <wouters@cistron.nl>
  To  : ggi-develop@eskimo.com
  Date: Sat, 20 Feb 1999 17:10:35 +0100

1. Re: David "Zoid" Kirsch (Quake) on GGI / 2. misc

1. QUAKE:
---------
Evan Martin wrote:

> [Zoid:] GGI was promising, but kinda started and stopped
> development. SDL is too high level. For something like

It still is promising, and the only real option for the future.

> Quake I need direct access to OpenGL (or to the
> framebuffer for software rendering), direct access to the
> sound card's DMA buffer and direct access to the mouse
> input and raw keyboard codes.

Hmm. This is really a whishlist, not a 'need' list. The DMA audio in
particular is not needed as suitable fragsizes mean very low
(unnoticable) delays (doom/heretic/descent with GSI have no noticable
delay, and I WOULD notice if there was as I'm very critical about such
things (e.g. although I've ported Maelstrom to GGI/GSI I don't really
play it for various reasons, the most important one being the pathetic
30 Hz screen refresh which is totally unacceptable for 2D games. Try
flying your ship at full speed in Maelstrom and see if you think it's
smooth..).

However, the windoze version uses DMA and that's why he wants it (I do
have the framework in libgsipcm btw, (the PCM modesetting code is still
in the server), to easily allow direct pcm access. The most important
use I see for this would be though, to make the program using the pcm,
lower level API independant...).

2. MISC:
--------
[ island ]
> > In the mean time, here's a state of the art (1982 ;-))) graphics demo.
> > BBC owners will certainly recognize it (Sengan?). Isn't that variable
> > surf fantastic ;-)))
> 
> Well libggi ain't set up on this box, but from the code I think I knowwhich one
> you mean. Gee... Long ago. Did you have one too?

Still do ;)

[ ggi/gsi_descent 0.7 ]

I inadvertently replaced ggiEventPoll/ggiEventRead by
giiEventPoll/giiEventRead in 0.7 which means a segfault at all times.
Undo and all is well.

Btw, there are no man pages for ggiEventPoll/ggiEventRead. Will they be
removed and will input be made completely separate (no input events from
visuals) from GGI now?

Wouter

Index: [thread] [date] [subject] [author]