Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!elroy.jpl.nasa.gov!ucla-cs!ucivax!milne
From: milne@ics.uci.edu (Alastair Milne)
Newsgroups: comp.multimedia
Subject: Re: Help with PC multimedia packages
Message-ID: <281FA4BA.18592@ics.uci.edu>
Date: 2 May 91 05:33:46 GMT
References: <1991Apr29.200448.14543@isc.rit.edu>
Distribution: na
Organization: UC Irvine Department of ICS
Lines: 83

In <1991Apr29.200448.14543@isc.rit.edu> rhm5684@isc.rit.edu (R.H. Mowery ) writes:

>	My only reason for questioning the IBM system is that it seems 
>that the run-time (at least from what I read) will only work on PS/2
>systems.  If this is true, that means luggin our system with us and also
>limits us from sending out a disk to let the clients view our machines.

    I know, having argued the same thing within our own group, that 
    MicroChannel machines (what people usually mean when they refer
    to the PS/2's) are having limited acceptance, but is it really as bad
    as that?

>Has anyone out there used the IBM AVC software with the Motion video
>Apdater A card, or the 750 motion media card?  IS there anything out
>there that will allow me to do this kind of quality?  We do have true
>blues at work, but of course at home I don't, so no taking work home.

    A PS/2 is not totally out of the reach of the home market,
    though IBM's prices may make it seem as if they'd like that.
    A model 55x will do fine for using M-Motion/A.

>I also cannot
>seem to find any software that can handle live motion video other than
>the IBM motion video.  

    I'm not quite sure what software you mean, but it's true that 
    running the M-Motion video/A card requires loading 3 TSR's (MMEDIA,
    MMOTION, and drivers for the players you want to use).

    However, your application can be in anything you like as long as
    it can allocate the variable blocks M-Control wants, raise M-Control's
    software interrupt, and ideally, be able to create and link in
    a WakeUp routine.  I've been doing it with Turbo Pascal 5.0 and 6.0 .
    Once I'd got some communications problems sorted out, it worked
    very nicely.  This is giving me a Turbo Pascal unit which any 
    application can use to obtain simple, straightforward calls to
    use video.

>They also said the card  only works on the PS/2
>series because of the Multi-Channnel Arch.  Why would this be true?

    IBM are pushing the microchannel bus as hard as they can, even
    though they're making it more expensive to build machines with it
    than with ISA bus.  However, there are actually machines from other
    manufacturers appearing which use MicroChannel.  I only know this,
    though, because BYTE examined a set of them last October.  Apart from
    that review, I've never heard them mentioned.

    MicroChannel is a *really* different beast from the old ISA bus.
    Adapters built for it are a different shape from adapters built
    for ISA.  Its abilities, signals, etc. etc. are all quite different.
    You might just as well try to use a board made for the Mac's NuBus.

>Could it be designed around this -- guess it's just there way of selling
>more machines???  I must say they were very helpful in getting me
>information, but I still would like opinions from the rest of the world.

     I found them reasonably helpful too.  In one case where they were
     less so, I got a handsome apology by e-mail after I mentioned it
     on the net -- in fact, in this group.

     My feeling at the moment is this: take MicroChannel boards, like
     M-Motion; ISA boards, like VideoLogic's DVA-x000; boards for the Mac,
     boards for the Amiga, etc.; and establish some portable,
     system-independent interfaces on which applications can be built,
     which use features common to them all.  As Borland's GRAPH unit
     for Turbo Pascal does across the variety of graphics adapters 
     for PC's.  

     Why do that, at the risk of losing some favourite feature from
     a particular one of them?  Because this area is still very young,
     and it remains to be seen exactly what standards will emerge and
     stabilise.  Until then, you want your applications and their support
     as portable as possible.

     And as I always do, I will again recommend high-level languages 
     for this work, not authoring languages.  You need readability, 
     maintainability, portability, and strong support for software 
     engineering precepts (and moderately efficient code generation 
     is rather nice too).


     Alastair Milne