Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!ucsd!ucbvax!agate!garnet.berkeley.edu!deadman
From: deadman@garnet.berkeley.edu (Ben Haller)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Can the Mac actually do animation?
Message-ID: <1991Apr20.211346.7328@agate.berkeley.edu>
Date: 20 Apr 91 21:13:46 GMT
References: <671126426.4@blkcat.Fidonet.Org> <91101.195933YUNSANJ@YaleVM.YCC.Yale.Edu>
Sender: root@agate.berkeley.edu (Charlie Root)
Organization: Stick Software
Lines: 38

God I'm posting a lot today.

In article <91101.195933YUNSANJ@YaleVM.YCC.Yale.Edu> YUNSANJ@YaleVM.YCC.Yale.Edu writes:
>
> Although reluctant to divulge tthe processes, it is possible to do mediocre
>arcade animation on a Mac. I've been able to get full background scrolling
>with several sprites moving around fine on either a Plus or SE in B/W or a Mac
>II in color. Of course, you're going to need to get down and dirty with
>assembly (yes, I know all you wimps out there are afraid of writing directly
>to the screen in fear of incompatibility). Actually, the incompatibility risks
>aren't very high, generally juts with graphics accelerators (then why do you
>need to write directly?) Offer a choice for users to switch between custom and
>standard routines and everything should work out fine.

  I find this extremely annoying.  The poster says "I know how to do this,
and it's really nifty, and I'm going to make you drool in anticipation of
this neat technology, but I'm not actually going to tell you how to do it."
Very immature.  I can understand not being willing to publish source code
(I don't publish some of my source, either) but restricting the free flow
of knowledge just because you're "reluctant" is a bit much.
  I've done a lot of direct to screen code, and my firmly held opinion is
that there is no way, with brute-force change-every-pixel-value code, to
scroll the full screen in 8-bit mode on a Mac II in less than about 3/60 of
a second, too long for good animation.  As I've mentioned before, there are
tricks, like color table animation, page flipping on some monitors, using
diff algorithms instead of redisplaying the whole screen, etc., but the
brute force method simply doesn't work.
  The poster seems to directly contradict my above assertion.  I challenge
him: either present proof that your assertion is true, or don't waste our
time.  I'd present my proof, except that I think I deleted it a long time
ago; however, anyone can write a routine that just stuffs values into
video memory, and clock it.  Unless I made some stupid mistake when I did
this, I believe my assertions are correct.

-Ben Haller (deadman@garnet.berkeley.edu)
"You take what you get and get what you please
 It's better to die on your feet than live on your knees"
    -Midnight Oil, "Power And The Passion"