Message-Id: <3.0.1.32.19990218121801.009135a0@iut.ujf-grenoble.fr>
X-Sender: sletuffe AT iut DOT ujf-grenoble DOT fr
X-Mailer: Windows Eudora Pro Version 3.0.1 (32) [F]
Date: Thu, 18 Feb 1999 12:18:01 +0100
To: djgpp AT delorie DOT com
From: LETUFFE SYLVAIN <sletuffe AT iut DOT ujf-grenoble DOT fr>
Subject: Re: Line frame buffer speed problem my 3D engine
In-Reply-To: <36CAC0B9.6BA3@club-internet.fr>
References: <3 DOT 0 DOT 1 DOT 32 DOT 19990216140256 DOT 0091b100 AT iut DOT ujf-grenoble DOT fr>
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id GAA17243
Reply-To: djgpp AT delorie DOT com

>Okay, I'll send you the LFB-detect stuff right away, it's
>sort of ripped out, I'm sending the include files so include them
>and call init_graphics and then you should be able to
>use the descriptor "video_ds" that points to the lfb.
>
>You might also want to change the mode, it is actually
>640x480x16bit(mode 0x111) and it's in init_lfb_graphics().


>Well, if it doesn't work, sen me a mail.
>
>
>second: the rotation stuff:
>Hm, it's not as easy as that, or, actually, it's rather easy, but
>I don't have any easy code to show you "like that".
>So, I'll try to write down something, and include some of
>my code if you want to.

Many many thanks....
>
>And now a question from me: what are you using(planning to use) for
>algorithm in your engine? resolution etc?

Which algorithme ? (there are so many I am using and I will use !! )

- graphic resolution > all of course ( I'd like to get the most
configuration as possible)
		But I am only workink on 8 bits colors modes for now 
		with 320x200, 320x400, 640x480.... 1600x1200
- Map resolution > Well I am doing it (for now) in long int format and I
shall go later in float 
		But I am not sure of that yet...( speed.... )
- Main algortihms > For drawing I use the "bresenam line algorithm"
	it is very usefull for lines 2D, lines 3D, gouraud, phong, maping...

- The wole stuff > all the map is tored in memory (and it doen't change )
		All formulas are computed from one position, one direction.
And the result is stored in a second memory buffer which I dump to the screen

The "you" position is defined by x,y,z and the "your_direction" is define by 

theta and phi ( angles used in "coordonnées sphériques" )
Rotation becomes very simple ( add a value to both theta and phi )
while a translation result in an add to x,y and z.

Well this looks quite good, but the "formulas" I 've told about are quit
expensive in term of machine computing. And I don't really know if that'll
do !!!