From: William Lachance <billl AT synapse DOT net>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Tile based scrolling?
Date: Mon, 20 Jan 1997 13:08:14 -0800
Organization: Synapse Internet <www.synapse.net>
Lines: 34
Message-ID: <32E3DEBE.1A43@synapse.net>
References: <5bjnrh$bu2 AT dailyplanet DOT wam DOT umd DOT edu> <01bc0405$868f0a00$20d6b5cf AT default> <5bus5e$308i AT elmo DOT cadvision DOT com>
Reply-To: billl AT synapse DOT net
NNTP-Posting-Host: violin-03.synapse.net
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp

frenchc wrote:

> Uhm... Excuse me?!? Do you know what the word means? Fast. You won't get
> any faster than allegro. 

Not entirely true. A general purpose system will always be slightly
slower
than something customised for the individual needs of your app. Of
course,
the time investment in writing your own high-speed routines probably 
wouldn't be worth it.. anyway, on to the matter at hand:

> real purpose to doing what he is doing. Just redraw the whole screen once
> again to your double buffer and dump it, don't add rows. I once wrote a
> dirty rectangle tile-based engine and trust me, it was waaay a bad idea.
> I did eventually get it working, but it wasn't worth it. Just dump the
> whole screen, that's what I'm doing now...

Mmmm.. actually, my experience has been slightly different. Without
using
dirty rectangles, the double buffer approach was unacceptably slow at
640x480 on my 486dx2/66. With dirty rectangles, however, it does manage
to get up to around ~40 fps (compared to.. somewhere between 20-30 using
the other method).

On a pentium, it doesn't make much difference (it'll probably run at
refresh either way) but if you want to include slower machines (and
your in-game objects aren't very big, or too numerous) dirty rectangles
are probably the way to go. 

Cheers,

Bill Lachance
billl AT synapse DOT nmet