unforum

[Ziola]

I haz transcription (sorry it took me so long...I had to find it)

[Rogi Ocnorb]

Kudos, Ziola!

The entire collection uses 40 out of 64 codons (39, not including the "CU" at the end.)
So, it's not likely that it's any kind of real Codon string.

In any 1000 codons collection, there are usually only a couple of missing codons, max.

This collection was of 1907 codons, total.
The counts of the values are:

[Ziola]

I've been extremely honest about my lack of puzzle aptitude. Beyond the simple things (binary, hex, base and rot) I'm more of a happy little cheerleader

Do you think we need to find the letters that represent the strings? I mean, if you follow this link and scroll down to the inverse table, you can see a three letter and a single letter that corresponds to each codon. For example, AAC can also be represented by Asn/N.

I hate puzzles

[catherwood]

[sixsidedsquare]

Here is the codons replaced with their single letter short name for corresponding amino acids (hopefully no mistakes):

[la!uep]

Hi, long time lurker here.

There's definitely a pattern there... what I did was to take all of the letters that sixsidedsquare decoded, and placed them in a text-wrapping window box (after removing the hard returns and using a mono-spaced font), and then played with the window width to change the wrapping line-length. I can literally see the patterns in the letters. For example, here's what happens when wrapping it at 56 characters per line:

[sixsidedsquare]

[la!uep]

[la!uep]

So I thought I'd try my hand at tweaking six's script, and here's what I ended up with:

http://stashbox.org/467310/letters-to-colours.html

This version allows for real-time window resizing and rearranging of the patterns. I also gave it my take on a nicer color palette (works with up to 64 different symbols). Just remove all returns and spaces and wrap the colors at will.

Thanks for the javascript idea, six. I was at first thinking of some sort of .bmp file manipulation thing, but this is much more convenient.

[sixsidedsquare]

Ha, I've actually got an irc script to flip text like that.

But back to the puzzle, ascii binary was my first thought with the 8's and having a bit more time this afternoon to play with it I've got it. Here's the process of thought I went through for anyone interested.

First I tried using Rogi's version with a unique character for each triplet looking at the sets of 8. With binary ascii the first character is always a 0 for all normal characters, so I tried replacing whatever characters those were with 0's for the whole text. Then for all letters a-Z the second character is a 1, so I replaced those, but the problem was there were already a number of double zeros at the start of sets of 8. I tried working though reasoning like this for a bit but nothing good was really coming out of it.

However I was getting to thinking that it would probably be better to go back to the original triplets and see if the ones that came first in sets of triplets (which were repeating throughout) had anything in common. So I lined up groups of 8 triplets to look. I've included the first few groups of 8 here for those of you who are playing along at home:

[Ziola]

Hmmmm...I honestly have no idea. But I will poke the puzzle maker to peek over here and see if there's more. Kinda looks like it, what with the last line and all.

[la!uep]

Good job six! Now, I'm looking at the discarded info, and I agree when you say there might be something there. I'm seeing more patterns in octets, and ASCII is spitting somewhat non-gibberish stuff, but with a few odd animals (spoilered just in case):

[sixsidedsquare]

[la!uep]

[sixsidedsquare]

Interesting, and somewhat hilarious find there. I think I can elaborate a bit more on the what, but the why still escapes me.

If we pull it back to the original characters and look a bit closer at that, what were seeing is that the spaces in one translation are becoming character returns in the other. You can see this from the original character string that corresponds to every space in the original decode (don't think it misses any):