pouët.net

I guess thats compensated by the c64 sceners/hangarounds that I've argued with who think that all pc demos are boring 3d shows that look like first person shooters, and dont require any skill to make at all since its all done by the hardware and directx.

Transform was pretty cool!

Otherwise, I'm not really feeling the C64 vibe. I have a lot of respect for C64 sceners but I think I overestimated the machine.

It is a shame coders don't push current hardware the way they push the C64 or we'd see some crazy shit!

Don't forget about such classic like SAMAR - great prods like Digital World, Opium, or Air Power.
Also, try 1.67 Years by Obsessed Maniacs.

[url]http://pouet.net/prod.php?which=1459[/lurl]

some more to try:

http://pouet.net/prod.php?which=1459
http://pouet.net/prod.php?which=1413
http://pouet.net/prod.php?which=1678
http://pouet.net/prod.php?which=11666
http://pouet.net/prod.php?which=1464
http://pouet.net/prod.php?which=1586
http://pouet.net/prod.php?which=1462
http://pouet.net/prod.php?which=1428
http://pouet.net/prod.php?which=18410
http://pouet.net/prod.php?which=14978

its easy to overestimate the c64 as it is being pushed so hard, hitech effects gets to look natural, but actually the cpu is so slow that in games it is a big problem to copy around the charscreen&color ram at 50fps.

You cannot push the modern hardware to the limit the same way you can push a fixed machine like the C64 or Amiga or any other of those old home computers. This has been explained to you and numerous other people a countless times and still you're clueless (then again, I wouldn't expect anything else). Programming on the C64 doesn't have anything to do with programming on the PC, except when considered from a very theoretical standpoint, so comparing them is like comparing apples to chocolate.

and it's not like pushing the limit is an obvious thing.

I don't think in coder logic. To you it's apples and chocolate, to me it's computers and more powerful computers.

if you dont think in coder logic, then the processing power shouldnt be relevant.

You can't upgrade C64 by buying more RAM, HDD, faster CPU etc. - but on PC you can do that. Got the point, Buttler?

Of course PowerCPU and similar doesn't count as a "upgrade" - those are peripherials. <- right?

Yes, but I don't understand why you can't push CPUs and GPUs like the C64 or Amiga etc like Preacher said.

feel free to try :)

Feel free to push the limits of your piano.

Buttler: because you have to tackle with compatibility issues and drivers and ... therefore it's not possible, and not recommanded either, to go too far ( except in specific categories such as 1k and the likes where it's basically "anything goes" ) with YOUR machine.

Having a fixed hardware voids the problems I just mention, therefore YOU can push as far as you want. It will work on any machine and at the same speed.

That's because, as you admit, you don't know anything about hardware or coding. Please, stop digging a hole for yourself!

There is nothing that stops coders from writing something that pushes their setup to the limit. What stops them is that they are to lazy to learn the ins and outs of specific hardware which is something you have to do on a system that died in 1987 and have stayed the same since.

And face it, everything above 640k ram on an IBM compatible is fucking peripherial.

When someone tries to write something for a specific setup (let it be 32-bit colour in DOS, a graphics card with hardware T&L, or OpenGL cubemap extension, or any model of programmable GPU and so forth), everyone else will instantly flame that one person because it does not run on anything but that setup.

The real reason we don't do that anymore on PCs is that nowadays we have these things called "operating systems" and "application programming interfaces", which come between the programmer and the hardware, and either heavily discourage bit-pushing and hardware banging or then just flat-out make it impossible (and rightfully so, that sort of programming belongs to the Eighties). Some people just don't realize that the same development that gives them USB peripherals that plug'n'play (finally!) and fancy GUIs on multicore machines makes it impossible and unnecessary to do the same things we used to do in the DOS era when you could tweak the VGA registers between scanlines and whatnot.

It would be rather hard to max out code on PC the same way you do on a C64 or other oldschool platform. The PC hardware just moves too fast, there are too many different combinationsand there simply isn't time or motivation to really go to the extreem limits.
If say a Pentium 166 with a Voodoo 1 had been decided as the standard that everyone should use for all eternity on the PC, I'm sure we would see some demos today that took it to the limit, as coders would have had ten years of time tinkering and learning all the details and tricks. But when new hardware comes out every few months, that kind of optimization just isn't realistic.

I think it's great that we have these two ways for coders to do their stuff - either hardcore coding on fixed, well-known old hardware or experimenting with bleeding edge new technology. Each has its charm.

Word.

Thread closed.

It makes more sense now. Funny how a passing comment can cause such a stir.

So, do many PC coders use assembler to code in or does that cause problems too?

There's no point coding in assembler on a PC since it's fairly difficult to beat modern compilers for speed and optimization. The only benefit is in extreme size optimization where the order of instructions might help packing.

I always wondered why assembler wasn't discussed much in the PC scene. Must be a right bitch to learn, especially for different hardware configs.

Does anyone code in Amos these days? :D

It doesn't cause problems any more than any other programming language (they're all converted to assembly anyway), but it's pointless in most situations. Nowadays, for the most part, assembly language is used in operating system programming (there are things that need to be done in the kernel that few compilers have support for), in some form of optimization with special instruction sets in newer processors (mmx, 3dnow!, sse, sse and all those), or in very special circumstances (size optimization, code profiling, perhaps in buffer overflow attacks and the like).

And you musn't forget that modern processors tend to shuffle, change and reorganize your code around to optimize its execution anyway. Because of that, and the fact that the code is more and more likely to run on multicore processors which have their own issues and peculiarities, writing hand-optimized assembly is both pointless and impossible. I still remember all execution times, cache timings and pipelining rules on the original pentium on which I learned assembly.. nowadays, none of that matters one bit.

Algorithmic optimization (doing the computations and whatever theoretically as efficiently as possible) is much more important than low-level things like assembly, and higher level languages are a lot better suited for that.