hacker.org

I have a problem with the Quine-Challenge:
I solved it, but I can't submit the answer because it is too long:
The server answers:

Request-URI Too Large

The requested URL's length exceeds the capacity limit for this server.

So what can I do? Is there an alternate way to submit very long answers?

Try sending it as post data?

btw, this is kind of neat: http://steike.com/code/useless/zip-file-quine/

I tried it using POST method, but my answer is not accepted. It says the answer was incorrect.
Are there any limitations concerning code-length or cycle-count?

btw, the empty hvm-program should be also a correct answer.

it's true the empty program is an answer, but not a very interesting one so we don't allow it =)

the max cycle count of 10000 applies to the quine challenge, but it should tell you if that's the problem. if you want, email me your code (adam at adum.com) and i'll see what isn't working. there is probably a limit on code size, but i'm not sure what offhand.

gfoot wrote:btw, this is kind of neat: http://steike.com/code/useless/zip-file-quine/

Very impressive.

how is that done?

The zip "deflate" algorithm has instructions for "output the following N bytes" and "repeat the N bytes starting from the Mth byte backwards in the existing output". So naming these operations N(n) and P(n,m), show output in comments and no comments on lines that are just data:
Note that in P(3,2) it outputs *three* items starting from the second-last printed, so the third item has not been printed yet - the third item ends up being the *first* one printed by the current statement.

It is quite clever. I wrote this from memory of reading a description of the method, so I hope it's correct - otherwise, try searching for "zip quine" and see if you can find the description yourself.

Should also be noted that "HVM run ERROR: invalid instruction (PC=0, STACK_SIZE=0)" is also quining program that isn't accepted

I think I'm really stuck at quine. I've solved quite a few hvm-problems, and I've also thought quite a bit about this one. But I just can't imagine how the solution would look like. I think that I can't encode the program text explicitly, because every time I add a character to encode another character, this additional character must be encoded by another additional character and so on... The only approach I can imagine is some kind of procedural description of the characters (i.e., the characters get computed implicitly). But this procedural description must be contained in the program itself. So this seems to by quite abstruse.

The best solution would be the initialization of the memory-cells with appropriate values, but this seems to be only possible in the local test environment. Oh, and I also searched for a weak spot in the hvm itself to jump to the code-string. But (as expected) this also seems to be impossible.

So has anybody some hint to get me on the right track?

There are fairly standard techniques for writing quines in different kinds of languages - you should be able to get some good pointers from a Google search. There's still plenty of work involved in getting one of the common forms to work in hvm.

I know how to write a quine program in hvm, but I cannot find a quine challenge, although
I've solved more than 120 challenges (including some hvm challenges). Where is the quine
challenge hidden?

It comes just after Spiral Bits.

tog wrote:It comes just after Spiral Bits.

Ok, thanks. I was too lazy to try that.

You could get practicing on the hvm quine anyway, so you've got it ready when the time comes. Bear in mind the 10000 cycle limit, that bit me a few times.