Re: [AL] Per pixel effects, Speed?

[Thomas F. <tfj...@ho...>]

On December 29, 2001 08:45 am, you wrote:
> i dont think i have the time to learn enough about it, what i need is some
> blender / blitter code that uses MMX and SSE and i can then rip that code
> and use it in my warez, and learn from it..... it seems to be the best way
> for me to learn.
> my C/C++ optimization skills are weak, but  i have tried to learn from
> compiling snippets and reading the asm code... but now i need to learn
> about this MMX and SSE stuff....
> is there any basic info about it.. tutorials etc.. URL's  that you know of
> ?

Get Bob's Fblend (http://sf.net/projects/fblend). It will do what you need.

MMX and SSE are processor specific extentions to accelerate certian actions.
MMX is pentium and above and SSE is pentium 3 and above. (IIRC)

-- 
Thomas Fjellstrom
tfj...@sh...
http://strangesoft.net
>From <all...@ca...> Sat Dec 29 12:59:59 2001
Received: from dahlsten.mdh.se [130.243.79.38] by canvaslink.com with ESMTP
  (SMTPD32-6.00) id A49D4FA402DE; Sat, 29 Dec 2001 12:59:57 -0500
Received: from student.uu.se (lobbyn-6-isdn-ppp.mdh.se [130.243.78.62])
	by dahlsten.mdh.se (8.11.6/8.11.6) with ESMTP id fBTHtoT08593
	for <al...@ca...>; Sat, 29 Dec 2001 18:55:50 +0100 (MET)
Message-ID: <3C2...@st...>
Date: Sat, 29 Dec 2001 19:16:35 +0100
From: Sven Sandberg <svs...@st...>
X-Mailer: Mozilla 4.06 [en] (Win98; I)
MIME-Version: 1.0
To: al...@ca...
References: <3C2...@st...> <002b01c1900e$d38f2480$010...@ms...>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by AMaViS-perl11-milter (http://amavis.org/)
Subject: Re: [AL] ANNOUNCE: Spritelib 1.0.0
Precedence: bulk
Sender: all...@ca...
Reply-To: al...@ca...
X-UIDL: 300937880
Status: O
Content-Length: 651
Lines: 15

Elias Pschernig wrote:
> I shortly looked at it, and there's a minor problem with mingw:
> The example won't compile because you forgot END_OF_MAIN(), and
> there's a warning because you use "return NULL;" inside of main().

Oops, thanks for mentioning. These are fixed now.

> If I didn't just overlook it, there should maybe also be a function to
> restore all the current video sprites. (Useful inside a
> switch_callback, after complete video memory has been lost.)

Good idea, and also easy to implement inside the library. This now
exists in ASU 2.0.0 (see announcement).

-- 
Sven Sandberg   svs...@st...   home.student.uu.se/svsa1977
>From <all...@ca...> Sat Dec 29 13:24:43 2001
Received: from dahlsten.mdh.se [130.243.79.38] by canvaslink.com with ESMTP
  (SMTPD32-6.00) id A4B73FD303D4; Sat, 29 Dec 2001 13:00:23 -0500
Received: from student.uu.se (lobbyn-6-isdn-ppp.mdh.se [130.243.78.62])
	by dahlsten.mdh.se (8.11.6/8.11.6) with ESMTP id fBTHuGT08635
	for <al...@ca...>; Sat, 29 Dec 2001 18:56:16 +0100 (MET)
Message-ID: <3C2...@st...>
Date: Sat, 29 Dec 2001 19:17:01 +0100
From: Sven Sandberg <svs...@st...>
X-Mailer: Mozilla 4.06 [en] (Win98; I)
MIME-Version: 1.0
To: Allegro normal <al...@ca...>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by AMaViS-perl11-milter (http://amavis.org/)
Subject: [AL] REANNOUNCE: ASU 2.0.0
Precedence: bulk
Sender: all...@ca...
Reply-To: al...@ca...
X-UIDL: 300937880
Status: O
Content-Length: 554
Lines: 13

This is the reannouncement of ASU 2.0.0, which is the new name of
Spritelib which I announced yesterday. The new address is:

  http://home.student.uu.se/svsa1977/programs/asu/index.html

See previous announcement for a description of what it is.

Since Spritelib already existed (thanks to Grzegorz Adam Hankiewicz for
pointing this out), I've changed the name to ASU (Allegro Sprite
Unification), and I've also fixed the bugs and added the function
suggested by Elias Pschernig.

-- 
Sven Sandberg   svs...@st...   home.student.uu.se/svsa1977
>From <all...@ca...> Sat Dec 29 13:46:35 2001
Received: from smtp2.mts.net [205.200.16.113] by canvaslink.com with ESMTP
  (SMTPD32-6.00) id AF8946F90380; Sat, 29 Dec 2001 13:46:33 -0500
Received: from hppav (wnpgas04-p289.mts.net [205.200.59.106])
	by smtp2.mts.net (8.11.4/8.11.3) with SMTP id fBTIgS608807
	for <al...@ca...>; Sat, 29 Dec 2001 12:42:28 -0600 (CST)
Message-ID: <002101c190a8$e4495080$6a3bc8cd@hppav>
From: "Chris Barry" <cr...@mt...>
To: <al...@ca...>
References: <003501c19004$f765d700$4f1286d4@p2cel400> <5.0...@oi...>
Date: Sat, 29 Dec 2001 12:39:18 -0800
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2615.200
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
Subject: Re: [AL] artsoftware , scrolling backgrounds
Precedence: bulk
Sender: all...@ca...
Reply-To: al...@ca...
X-UIDL: 300937880
Status: O
Content-Length: 1215
Lines: 29

> >It's easily done, but it'll eat your memory like a trough of hogs. A
bitmap
> >that's one screen high and 20 screens wide at 640x480x16 resolution will
> >gobble almost 20 megs of RAM (I think). And that's just for the
background,
> >let alone the rest of the game. Use a tilemap. Or vastly redesign how
your
> >game works (limit or eliminate the scrolling).
>
> do cartoonist have the same problem ?

Do cartoonists have RAM problems? No, if they need more room to store their
backgrounds, they get a bigger drawer :) They don't need to "tilemap"
background for RAM's sake, though they do reuse them (as others said) to
save time. This is an apples and oranges comparison. They don't really
relate. If you want games with one-image backgrounds, try Final Fantasy
VII - IX, Resident Evil and Abe's Exodus. The FF games scroll minimally, and
the other two don't scroll at ALL. Yet Super Mario Bros. for the NES scrolls
for more that 20 screens on some levels. All the magic of tilemaps. And of
course, tilemaps hold more information than just image data, especially for
collision detection.

Trust me. You want tilemaps.


Chris Barry

"After all, when one is serving bait, presentation is everything" -
Megatron