Code Cracking

This one should be pretty simple - chosen a nice long text to make it easy:

Ceilcleo tehdnmlt irI ktewlteh ala eayIo nn ldtaaOehtnfnamt cneCeilcleo tehdnmlt irT m mtpyteceoleiso2aHenyade3rnrimgdOsdsfho awFl e irooseclelwwthdnI otehwlswh ealh et yTohlt iahu lhpnnrga ea dT rwhsrseooWpoprhdne to iren ocleF fitsaolesh yri2iwMdi aynuenbu drrgetaPsnaissoHhodrhdnu we irsn acleImteh ytu2iw sbtsaTwrolenoeyfi d a2faDeisepoctnRtomclee wyhdnsn2irF owlwosne aro layIohcmnn ea dtt laT ihqeeut

admin: ooo I like it!
I'll save this one for tomorrow, (travel day)

nzlockie: Wow, a whole week and nothing at all. Seesh. I guess even the simplest codes can be super tough sometimes. I swear you'll kick yourself once you see it.

admin: I'll admit this is giving me an issue. I've spent a solid couple of hours looking at this without any major break throughs.
The frequency patterns seem to indicate transposition rather than substitution.
I'm observing several groups of common letters which obviously have my attention.
The numbers are throwing me a little but I'm ignoring them for now. I don't trust the spaces but at this stage I'm using the capitals to separate the text. There's nothing to support this, it's just a guess.

my best guess is that this is some kind of route cipher.

I have one question, is there a key required for this message?

nzlockie: Would you say your last code had a key?

admin: No, you just had to write the code out letter by letter across three columns. I wouldn't consider that a key. So I'm guessing yours doesn't either?

nzlockie: In that case, no, it doesn't have any sort of key.

OMG! Two full weeks and no crack. Wow, I really genuinely thought this would be simple.

OK, that's it, I'm releasing a clue. This one might be a bit blatant but never mind. You see those numerals in the code there? Numerals stand for consecutive spaces. If the code would have two spaces in a row, for example, those two spaces are replaced by the numeral "2" at the end of encryption. This allows the code to be transmitted as HTML text (since html ignores multiple space characters as whitespace by default - this was necessary). The first step in decoding the code is to replace numerals with consecutive spaces.

And a second hint - other than that, this code contains no substitutions.

admin: Sorry! I've been really busy this week and haven't looked at it at all!
Thanks for the clue, we'll see if that makes a difference!

Not that I'm going to have much chance to work on it this weekend either...

It's been 3 weeks now. If no crack in another week I'm literally just going to give away the answer.

Here's a bit more of a hint though: kind of like your last code had a sort-of "key" of 3, this code has exactly the same "key".

OK, since this code is one month old, here is the solution.

THE CODE NOBODY COULD CRACK

Note: I'm going to use - for a space in codes (because sometimes spaces will be doubled up and you can't show that nicely in html) and _ for an "unknown" character that we still have to decode

TO ENCODE
Seperate cleartext by sentence and remove all punctuation other than spaces, except for spaces following full stops. Replace any numerals with the numbers in word form, so that the entire text of each sentence has only letters and spaces. If there are any words with capitals except the first word of the sentence, those words need to be put in lower case.

Write out every third letter of the cleartext into the code until the end of the clear string is reached.

Clear: I-am-really-awesome
Code so far: Imelase

Then starting at the second letter of the cleartext, repeat:

Imelase--aywo

Then finally repeat starting at the third letter of the cleartext:

I am really awesome
Imelase--aywoarl-em

Replace multiple spaces with a single numeral indicating the number of spaces to get the final code for the sentence:

Imelase2aywoarl-em

Finally combine all the sentences together into a single string.

TO ENCODE (Faster way)
Write the text out into three rows, and copy it off line by line, then do the numeral replacing for multiple spaces:

Imelase
--aywo
arl-em

Imelase--aywoarl-em
Imelase2aywoarl-em

TO DECODE
Do one sentence at a time. You can tell where each sentence begins and ends by the position of the capital letters.

Begin by replacing numerals with multiple spaces.

Imelase2aywoarl-em
Imelase--aywoarl-em

Now write out each letter, leaving two gaps for the next two letters:

Code: Imelase--aywoarl-em
Clear: I__m__e__l__a__s__e

Repeat for the second and third letters:

I-_m-_ea_ly_aw_so_e
I-am-really-awesome

TO DECODE (Faster way)

If you can count and do math fast, count the number of letters in each sentence and divide by three. Then copy that number of letters out from the code across three rows, and read the code down. If you have a remainder of one-third, round down and add one extra letter to the first row. If it's two-thirds, instead add one extra letter to the first and second row.

Imelase aywoarl em = 19 characters
19/3 = 6 + 1/3

Imelase
--aywo
arl-em
(look at this with a monospace font to see how fast it is)

admin: Hey sorry, I didn't get to look at this code as closely as I was hoping to. As it turns out, it's unlikely I would have gotten this though. This is like a Rail Fence cipher but across three lines instead of the traditional two. Ironically, I did actually experiment with that idea at one stage and I remember testing three lines but the code didn't resolve itself in the first few lines so I moved on. I'd noticed that it spelled "child" with three lines - which I thought was a bit of a coincidence.

With your explanation I can get it to read now. The single spaces tripped me up at first.

Nice one!

nzlockie: No worries. I thought the line "Deisepoctn" would be a bit of a giveaway because there's literally only one sensible anagram of that word, and then you apply the same technique to "T ihqeeut" to understand how the spaces fit in.

admin: Got some time a decrypted the rest of it. I feel a bit silly now. I totally saw a bunch of clues that this was the kind of encryption you'd used but I just didn't focus on them. Fair play, that was a good code. Interesting poem choice too!

I have an interesting challenge for you, if you're up for it:

This is an algorithm I'm currently working on IRL. I haven't broken it yet, but I've narrowed it down to this.
002891 = 41334
003187 = 41423
004374 = 44215
011592 = 53315
074997 = 42544
075878 = 53143
200015 = 56621
281015 = 52224

I need to work out how the second number is calculated from the first. I don't know how many of the digits in the first number are used, but I know that it's not MORE than this. The first number is base10 but I strongly suspect the second number is a different number base.
The second number can never contain a zero and I have no record of it ever containing higher than a 6 - although it would make sense with one of my theories if it WAS able to contain a 7.
The second number must always be 5 digits long.

These are the only pairs I have to work with, although there is one last pair that I haven't looked at yet which I will use to confirm any algorithm I can find that converts all of these ones.

Feel free to have a crack!

nzlockie: Lars hasn't cracked it! D:

Pinkie: Neither have I! And I've been working on it for almost a year!

nzlockie: Yeah, I've pretty much forgotten about this. I just found my notes on it and they're all over the place.

I am pretty sure that the first 4 numbers are different from the second four. Just occurred to me to ask - do the first four numbers form part of a set in any way? Because there's a lot of striking similarities between them. Like if you add up every second digit in all the numbers, for example, you get:

11,9
11,8
11,7
11,8
13,23
12,23
3,5
4,13

The second four numbers are all over the place, but the first four are pretty sequential. There's a few other things about those numbers that struck me but it got me nowhere in the end.

nzlockie: That must be super hard!

admin: Sorry, no set. The first digits are the unique serial numbers from each vehicle. There are a mix of models represented here and they're spread throughout the list.

One of my issues is that I can't confirm how many of the digits from the vin are used to calculate the second number, although I know it's no more than six.

I would be very surprised if the conversion didn't involve a bit function of some description.

I have to say that at this stage I'm running short on inspiration!

Pinkie: It's pretty hard but it's a good lesson in patience. I worked on one such algorithm that took about 4 years to crack! It was a good feeling to crack it though!