New years challenge!

[abachler]

This doesn't make any sense

X = (X * K) MOD P

X = 5, K = 2, P = 3

Like what? Is that not a mathematical equation in the traditional sense?

Rather a sequence X sub N = (X sub N-1 * K) MOD P?

Its a clue, not the solution. But Ill elaborate a little more. There are conditions to the variables in that equation-

P must be prime

1 < K < P

0 < X < P

thus P must be greater than 2

[abachler]

OK, since im not getting very many entries, Ill remove the requirement for it to be in assembly, since I think this is the biggest barrier for most people.

[manav]

Not like a barrier although, but I was curious why in asm, and that too on a C++ forum?

[whiteflags]

As interesting as it is, I don't think people are taking this request seriously. That could also be the reason you aren't getting replies. I'm sure some of the grizzly people here still remember assembly, so that isn't the problem.

[dwks]

Not like a barrier although, but I was curious why in asm, and that too on a C++ forum?

C++ and C . . . and C#, in a way.

[abachler]

I don't think people are taking this request seriously.

Well, my contests tend to be fairly difficult, not the usual ' write a program using only macros and only use the letter S 3 times'. I could dumb them down, but then what would be the point. A contest is supposed to be challenging.

[oogabooga]

Well, my contests tend to be fairly difficult, not the usual ' write a program using only macros and only use the letter S 3 times'. I could dumb them down, but then what would be the point. A contest is supposed to be challenging.

Do you have a solution?
If so, could you answer the following questions about it?
1. Can your method play any rhythm, or just one note per beat?
2. Can your method play any diatonic tune (perhaps within a bounded note-range)?

[Cactus_Hugger]

My solution can play any rhythm. It can't however, do chords, or rests. (Although rests could easily be added if I am allowed a call to Sleep(), or perhaps one could play a tone outside of human hearing.)

I thought abachler's challenge was a good one, even for cboard. Leave your programming comfort zones, and give it a try.
Personally, I failed at getting nasm to assemble anything Win32 related. I finally gave up on that, and really bit the bullet, and sat down with gcc's assembler. (I greatly prefer Intel-like syntax over AT&T... oh well.) After much crashes, and me finally discovering that a semicolon doesn't actually comment a line (but doesn't give an errors...), I got one working. Although I'm still perplexed by some of the assembler that gcc generates, such as:

	movl	$0, %eax
	addl	$15, %eax
	addl	$15, %eax
	shrl	$4, %eax
	sall	$4, %eax

Why not just precompute whatever that results in, and stuff it in eax...?

My solution is attached. Should compile on Win32/x86.

[alreadyinuse]

Personally, I failed at getting nasm to assemble anything Win32 related. I finally gave up on that, and really bit the bullet, and sat down with gcc's assembler. (I greatly prefer Intel-like syntax over AT&T... oh well.) After much crashes, and me finally discovering that a semicolon doesn't actually comment a line (but doesn't give an errors...), I got one working.....

lol :-D

have you tryed FASM ? It has a syntax very similar with nasm, a wonderful macros system (well, this is my opinion), some great library and include files....

[oogabooga]

Unfortunately yours is not a solution to the original problem.

You can't use a table of values. You have to calculate the next
note and duration through some algorithm.

That's what makes it difficult.

As far as the assembly contortions go, what exactly was the C code
that generated it? I don't see that in my "gcc -S" of simple program.

[Cactus_Hugger]

#include <stdio.h>

int main()
{
	return 0;
}

Generates the assembly that I'm curious about, and posted above.
EDIT: Ok, turning up optimizations to -O2 generates more sane assembly, and precomputes it. I just wondering why gcc is doing eax = ((0 + 15 + 15) >> 4) << 4 in the first place... (other than a call to _alloc and __main, eax is never used. Even then, I'm not sure if _alloc/__main use eax... don't know the calling convention.)

Totally missed the table of values thing, so I guess not. Makes me wonder exactly how we're supposed to play a song, unless there is one specific Christmas tune that follow a mathematical equation...? Such as the one hinted earlier, but that gives values (way) too low to be audible frequencies, and durations too short to be more than a click - I'm obviously 'not getting it'. If it's pitches, relative to some starting pitch... then I need log() to make the conversion to frequency, again, breaking a rule.

[dwks]

You cant use a table of values. You have to calculate the next note and duration through some algorithm.

My algorithm is fgetc()! -- or whatever you use in assembly.

[abachler]

hehe, too bad it doesnt seem to work on my system :0

[GanglyLamb]

So when is this contest ending? (not that i want to participate im just curious as to how this is solved - alos looking forward to see abachlers solution )

[Cactus_Hugger]

hehe, too bad it doesnt seem to work on my system :0

My example? It should compile with gcc ... what errors do you get? A simple "gcc -o music.exe music.s" is all that's needed. (Of course, you obviously need Windows.)
Now... if you're using MSVC... well... no idea what its assembler is like.

dwks: My understanding was that all you could use was Beep(), so fgetc() would be against the rules.

I am also interested in abachler's solution.