Hot'n Tropic Climate

User avatar
dangermouse
Posts: 89
Joined: Sun Jun 05, 2011 8:14 pm
Location: deep space computing AG
Contact:

Hot'n Tropic Climate

Post by dangermouse »

Does the challenge title imply some algorithm to be used to decode the message? Or rather the author of the algorithm?
User avatar
TheBigBoss
Posts: 29
Joined: Thu Jun 07, 2012 12:07 pm
Location: Germany

Hot'n Tropic Climate

Post by TheBigBoss »

The substring "n Tropi" from the title sounds like "entropy".
So it's more likely a data compression rather than a crypto encoding.
One of my first guesses was a continued fraction, from which I got 24 integers below 128.
But these integers don't make sense at all, so I think it's the wrong way.
And what's about the number 25 from the challenge?
AMindForeverVoyaging
Forum Admin
Posts: 496
Joined: Sat May 28, 2011 9:14 am
Location: Germany

Post by AMindForeverVoyaging »

Seeing how teebee is back on the forum:
Could we please have a hint for this challenge? Only 4 out of 56 people have solved it, which means that it is pretty damn hard.
User avatar
teebee
Posts: 89
Joined: Mon Nov 10, 2008 3:21 pm
Location: Germany

Post by teebee »

The hint is already given by TheBigBoss ...
AMindForeverVoyaging
Forum Admin
Posts: 496
Joined: Sat May 28, 2011 9:14 am
Location: Germany

Post by AMindForeverVoyaging »

teebee wrote:The hint is already given by TheBigBoss ...


Well, he never solved the challenge. Nor did anybody else in the past five years.
User avatar
TheBigBoss
Posts: 29
Joined: Thu Jun 07, 2012 12:07 pm
Location: Germany

Post by TheBigBoss »

I know how the text is encoed. The big problem is to find the right alphabet and the frequencies. All my brute-force-solvers took too much time to find the right answer.
User avatar
teebee
Posts: 89
Joined: Mon Nov 10, 2008 3:21 pm
Location: Germany

Post by teebee »

Oh, I see. I have reverted the changes from revision https://en.wikipedia.org/w/index.php?ti ... =659346887.
eulerscheZahl
Posts: 58
Joined: Thu Nov 29, 2012 7:45 pm
Location: Germany

Post by eulerscheZahl »

What did you do against rounding errors?
I can clearly see two familiar words at the beginning.

I blew the numbers up to calculate with sage (BigInteger library supporting fractions), but the end of the text looks quite random to me (ends with 'gaso').

By the way: the continued fraction is [0, 1, 6, 26, 2, 124, 1, 1, 4, 3, 3, 13, 1, 42, 1, 1, 8, 1, 55, 1, 16, 1, 3, 2, 2, 1, 5, 1, 45, 2, 1, 1, 1, 2].
I'm telling this, because this doesn't help at all :)
User avatar
TheBigBoss
Posts: 29
Joined: Thu Jun 07, 2012 12:07 pm
Location: Germany

Post by TheBigBoss »

When I say, I know the encoding, I do not talk about continued fractions.
eulerscheZahl
Posts: 58
Joined: Thu Nov 29, 2012 7:45 pm
Location: Germany

Post by eulerscheZahl »

Me neither.
As I got the first two words, I should be on the right track.

Against the hot climate, you could use an air conditioner - maybe that's the reason for the challenge name.
User avatar
TheBigBoss
Posts: 29
Joined: Thu Jun 07, 2012 12:07 pm
Location: Germany

Post by TheBigBoss »

I fed my algorithm with the familiar words too and came quite near to the number given in the challenge. But this approach is not very promising, because the Greedy-algorithm approach will not necessarily converge to the plain text in each and every case.
The NTL library does a good job for me. I have no trouble with rounding errors.
User avatar
teebee
Posts: 89
Joined: Mon Nov 10, 2008 3:21 pm
Location: Germany

Post by teebee »

eulerscheZahl wrote:[...] (ends with 'gaso').
Use 2.360 instead of 2.361 ...
eulerscheZahl
Posts: 58
Joined: Thu Nov 29, 2012 7:45 pm
Location: Germany

Post by eulerscheZahl »

2.360 - that was really helpful.
Now both sage and pari/gp find the correct solution.
TheBigBoss wrote:I fed my algorithm with the familiar words too and came quite near to the number given in the challenge.
You know, that you can reverse the function, right?
AMindForeverVoyaging
Forum Admin
Posts: 496
Joined: Sat May 28, 2011 9:14 am
Location: Germany

Post by AMindForeverVoyaging »

teebee wrote:Use 2.360 instead of 2.361 ...
eulerscheZahl wrote:2.360 - that was really helpful.
Now both sage and pari/gp find the correct solution.
The whole challenge falls apart when one value is off by one thousandth? Now that seems a bit silly.

Especially when neither 2.360 nor 2.361 can be proven to be the absolute correct value. It really depends on which source you use; and there is no way of telling which source is the right one to use.
eulerscheZahl
Posts: 58
Joined: Thu Nov 29, 2012 7:45 pm
Location: Germany

Post by eulerscheZahl »

Compare the value for 'w' in wikipedia 2009 and wikipedia now on the on hand and wikipedia 2015 on the other hand.
The value was changed back to 2.360 just two days ago from a German IP (teebee, is that you :D?).
Maybe that's the reason, that noone solved it for a long time.
Post Reply