15 replies to this topic

[Mihail121]

So...

I'm glad to finally write something about the software rendering articles after so much time! As you know Hans and i are exposed to great school/uni stress (mental pressure that is..) in the last couple of months but i decided yesterday to continue the work on the articles because i hate unfinished business and besides, it wont be fair to disappoint all the ladies screaming "Misho - your my software rendering god!!!" :notworthy: So here i present my development plans, so you guys can post some comments!

First, i'm very unhappy with our engine systems used in the different tutorial parts and besides, OOP languages (except for Java and D) are making me sick so i bought a book on C 15 days ago with the hope to remove the C++ additions from my head and to leave the pure procedural C :cool2: So i'm currently reading the book and i'm making progress fast.

After finishing that book i start the work on code base for the next tutorial, which covers in details the topic dynamic lightning! As already stated everything will be coded in C and ASM to gurantee code simplicity (which i greatly tolerate...). My main goal this time will be to create an extensible system ready for further additions so that we don't waste time to reinvent the wheel for each tutorial!

Further tutorials will include: clipping (other than scanline) + culling + z-buffering, texturing, lightmapping

That's it! I'm open for comments and critics!

P.S.

Please don't start another OOP vs. NON-OOP flamewar - it's just my personal choice, that's all :)

[Ed Mack]

I'm looking forward to it. I initially taught myself proper C, and love going back to good old malloc every now and then :)

[bladder]

yeah, im looking forward to this too. it's great that youve decided to finish them up - though C? really? well .... ok, it's your articles...

But yeah, having an extensible framework is always a good idea when writing multiple programs base dont eh same stuff.

[coelurus]

Assembler? Woah, that's rather ambitious, but is it really necessary? I don't mind C at all (in fact I think it's a great idea to promote that language :D ), just try to concentrate on the software rendering-bit more than the optimizing bit.

Plus if you don't feel like being all alone on this, you know where to find me :happy:

[Mihail121]

All alone? The only thing i can do alone is to brush my teeth... and few other actions that i won't comment here ;) I would be very happy with some help indeed!

And about the assembler... well it's not really needed for demonstrational puproses but i just want to show how damn smart i am ;) No, just kidding, i just care about the speed but i guess for the software rendering tutorials we should really concentrate on the software rendering only! That's why i'm thinking of using some external libs for the system like SDL for an example!

[anubis]

assembler in a tutorial sounds a little crazy to me unless the article is called : "optimizing software rendering through assembler magic". for learning purposes the code should be kept as easy to understand as possible

[Mihail121]

Agree! It's decided then - as simple and descriptive as possible!

[Ed Mack]

I thought asm would be good - kids like me missed out on that part of games education (Although I'm gradually learning ASM and next year it's on our course so I'll properly learn it)

But, ok. No ASM is good for newbies (and non-masochists)

[coelurus]

Hey hey, not everything in life is b&w. Personally, I felt that starting on assembler already in the series would be a wee bit too early, but it'd be a very good idea to slip in some asm later on (and call it similarly to Anubis' proposal). Asm is a golden tool used very seldom these days, but the tutorial should only show that it's a wonderful tool when writing fast software rasterizers, not a necessity. Small things make all the difference.

And asm is good for newbies, it shows them how computers really work and they'll be able to figure out what things in high-level languages are bad.

So, make plans for an asm-optimized tut! :)

[anubis]

it's no question that some carefully sprinkled assembler code is great but you have to be careful that it doesn't get in the way of the article's intended topic. in the end you want to tell people about software rendering and not how a computer works. imo that's the real reason to introduce asm code to this series in a seperated article that deals with asm only without introducing any further rendering concepts

[Mihail121]

Hmm... i guess Hans and i can decide whether to include asm-fragments after the main work on the tutorial is done and we have an overall look. But i guess that in the end we'll go without, 'cause, as stated, we want to concentrate entirely on the software rendering. Following that goal will play even more important role in the future, where the tutorials will get quite complicated.

P.S.
Ahh heck, we can just code the whole damn thing in some BRAINFUCK-clone!

[anubis]

Quote

P.S.
Ahh heck, we can just code the whole damn thing in some BRAINFUCK-clone!

may i suggest :

http://en.wikipedia....amming_language

or :

http://en.wikipedia.org/wiki/Malbolge

[Mihail121]

anubis said:

http://en.wikipedia.org/wiki/Malbolge

Dear Lord Jesus.... :blink:

[Mihail121]

So, a quick status update. The work on the article is already in progress and it appears this will be the biggest article we've ever wrote! It will be atleast as big as the all the previous together or even bigger :)

Hans and i are mainly discussing the following topics:

- physiks of lighting
- how do we "see" the world around us
- transforming the real world lighting model to a greatly simplified computer version (including math, implementation details)
- ideas for further research

However the problem that troubles us is that we probably won't be able to release a source (demo) with the tutorial. So we have two options:

1. release the tutorial without a source and release the source later
2. release the tutorial with a source, which will take a little bit longer

So we ask you, our readers, for an advice!

Thank you!

[Ed Mack]

Cool! wo0t ;)

[coelurus]

I take that as 'it'll be a little interesting even though the source might come a bit later'? :D If we could update the tutorial with a demo just a little bit later, I think that would be the best solution.