Re: PC Vs. Mac

M. Carleer (EST)
Mon, 28 Nov 1994 04:56:51 -0500

> Michael Sattler wrote:

>At 18:40 11/27/94, Stephen Wight wrote:
>>You don't have to support ALL the boards though. Soundblaster would give
>>more than enough of an area at least to start with in compatibility
>>terms. If not that, just the standard Windows MCI driver. If you used
>>Borland C++ it has built in ways of talking to the MCI driver. Most (if
>>not all cards) are compatible with the original SoundBlaster, if there is
>>a place to start THIS is it.
>>[Terribly long signature omitted]
>You are correct. I'm sure that the PCers at CUSM have done their math and
>come to their own conclusions. I have noticed that most games support the
>Soundblaster standard.
>[Another terribly long signature omitted]

First let me say that beeing SoundBlaster compatible is not important for
the development of a sound able CUSeeMe: all what is needed is to be
compatible with the Windows sound interface.
As far as my own experience in writing sound programs for the PC goes, there
are a couple of drawbacks in using the high level MCI commands. The most
important beeing that it only allows recording to a disk file or playback
from a disk file, and not to buffers. This means you are forced to use the
low level interface to the various cards. Another problem is that each card
comes with it's own driver, written by each card's manufacturer. This would
not be a problem if all these drivers were bug free, but this is not the
case. What I have discovered the hard way is that if the playback drivers
usually are good, the recording drivers are far from perfect and you have to
take special precautions to account for the errors in these drivers (like
not returning allocated buffers, or not saying that they have been used,
...). As I already said once on this list, I even know of a card which would
not record correctly using the card's own software, but working well with
the Windows Recorder!!! And this because of a bug in their driver!!!

All this to say that, because PC hardware development is not locked up by
one company, like it is in the MAC world, things get a little bit more
difficult to control; but on the other hand, new ideas can emerge more
rapidly and more easily because of the openness of the standards.

Cheers :-)

Michel Carleer

Laboratoire de Chimie Physique Moleculaire Phone : +32-2-650.24.25
Universite Libre de Bruxelles CPi-160/09 Fax : +32-2-650.42.32
50 Av F.D. Roosevelt e-mail:
B-1050 Bruxelles