hacker.org

Posted: **Fri Dec 05, 2008 10:47 am**

my code:

0<10>0^0<<0^82*?:1-3?0<<0<1+0>5cdp

Posted: **Fri Dec 05, 2008 1:52 pm**

you have to make it much shorter: there's one with 21 instruction limit

Posted: **Sat Dec 13, 2008 3:27 am**

Bah. I'm a Java programmer, I don't do terse. So here's one that neither (ab)uses the call stack nor writes to memory:

0<991++1^1^<0^2v:1+4?1v2vd1-0^8?084*1+-gdp

Posted: **Sun Dec 14, 2008 1:30 am**

BTW, lukas, you code has a bug: it terminates when it encounters a 0, so if that's part of the actual data, and the largest number comes afterwards, the result is wrong. But apparently that case does not occur in the tests that determine whether you pass.

And after sitting over this for 3 hours, I can't see for the life of me how to get it down below 29 instructions, let alone 21, seeing how even the simplest operations take 5 or more instructions.

I only managed to get down to 28 by using 0 to terminate and assuming only positive numbers; unfortunately that's not a valid assmption.

Posted: **Sun Dec 14, 2008 5:46 am**

brazzy wrote:BTW, lukas, you code has a bug: it terminates when it encounters a 0, so if that's part of the actual data, and the largest number comes afterwards, the result is wrong. But apparently that case does not occur in the tests that determine whether you pass.

And after sitting over this for 3 hours, I can't see for the life of me how to get it down below 29 instructions, let alone 21, seeing how even the simplest operations take 5 or more instructions.

I only managed to get down to 28 by using 0 to terminate and assuming only positive numbers; unfortunately that's not a valid assmption.

I would say that 3 hours is not that long. It took me a while , may be 3 weeks may be more to find the correct solution for 27. Then I put away HVM challenges and switch to others. Then I came back and got those mice down to 22. The mice problems are good in sense that it's pure coding and no IDE or assembler would help you.

Posted: **Sun Dec 14, 2008 8:55 am**

Well, it didn't take me too long to discover the 21-solution. So I don't think that many hours or even days are necessary. Of course, if you're stuck, it may be reasonably to stand back and come again later. I also found quinine much harder (and actually needed some days to solve it), because there are many more possibilities to try out.

@brazzy: It's nice and important if you know how to write valid programs. But this is the wrong place for the "correct" way of programming... this is hacker.org: everything that passes the test is a correct solution.

Posted: **Mon Jul 06, 2009 1:23 pm**

My Code:

Code: Select all

preserve 100

	var pointer
	var tmp

mov tmp, [pointer]
start:
	push tmp
	push [pointer]
	jg next		
	mov tmp, [pointer]
next:
	inc pointer
	push 20
	push pointer
	jg start
	pi tmp

Assembled to:

Code: Select all

55*4*<<99*92*+2+>99*92*+2+<55*4*<<:1-98+?55*4*<<99*92*+2+>55*4*<1+55*4*>54*55*4*<:1-792+*01-*?99*92*+2+<p

Yes, its kinda big, but im lucky my assembler finally works so far

Posted: **Sun Aug 30, 2009 7:22 am**

My code:

Code: Select all

0<45*>10^<0^45*<:1:3?d4g45*>1+0^45*:01-:158*-?45*<p

Posted: **Sun Dec 20, 2009 6:52 pm**

Well I'm quite ashamed of mine.

Code: Select all

0<00^991++:75*1-?1+0^2v2v<1^1^:1-01-*7?1vd1v3cd1v3cdp

Too long, call stack flooder, etc.
The "21 instructions" challenge is gonna take me some time...

Posted: **Thu Oct 21, 2010 10:26 pm**

Another straightforward version:

Code: Select all

0<999++>12g1+0^54*6g54*1+g:62**g       999++<p!0^<999++<:77*4+*g                                                     0^<999++>055*5*  -g

NOP slides! xD

Posted: **Sat Dec 25, 2010 10:56 pm**

My code:

Code: Select all

[9*9]=0;
[9*8]=[[9*9]];
while (4*5>[9*9]) {
        if ([[9*9]] > [9*8]) {
                [9*8] = [[9*9]];
        }
        [9*9] = [9*9]+1;
}
printi [9*8];

compiled to:

Code: Select all

99*00^2v>d98*00^2v>d45*99*<:1:92+      ?01        g10:991+*5+  ?99*<<98*<:1:92+      ?01        g10:92*6+    ?98*99*<<0^2v>d0        g99*99*<1+0^2v>d0996+*4+-g98*<p!

Let's hope there is only one with an instruction limit....

Posted: **Thu Jun 23, 2011 2:36 am**

look, here comes my quick'n-dirty solution for this one...

it uses a !subroutine! for comparing the s0/s1 values in place ..

44*g!1^1^:1:2?1vd$! 0<1<5c 2<5c 3<5c 4<5c 5<5c 6<5c 7<5c 8<5c 9<5c 91+<5c 92+<5c 93+<5c 94+<5c 95+<5c 96+<5c 97+<5c 98+<5c 99+<5c 99+1+<5c p

lots of bits wasted, but why not, i got enough of them here

... its all for speed and read-ability of course

!! ...

happy coding / greets - max.

Posted: **Fri Aug 19, 2011 1:50 am**

10<29*1+2^:1+74*?1^<1^:1+4?d0^<1v1+1v094*6+-gp

i wrote it out in pseudo-code and compiled by hand

Posted: **Thu Sep 22, 2011 3:15 pm**

01+0^<0<:1:01-:5?0^<0>0^45*1-:3?1c0<p

I thought 37 instructions weren't that bad until I saw that mouse challenge

Posted: **Mon Jan 02, 2012 9:44 pm**

gosh, did this, but I won't tell you my #s of instructions... for King Mouse I will have to strip it down significantly...

hacker.org

King Rat

King Rat