# Whatever it is, it should be in a museum

I have a friend who collects early computer technology. In fact, he's got a bunch of bizarre peripherals that probably should be in a museum instead. Some of this stuff is so esoteric that he's had to create his own hardware just to communicate with them. (You should see the setup he's got for running his probably-soviet punch card reader... sooo janky).

Anyways, his cable management is atrocious, so when I was over at his place the other day, he was trying to figure out what some of the wires were. Instead of tracing them to their ends like a normal person (which, admittedly, would've been a task), he was reading the raw data as it goes across the wires using a strange device that would've been more at home in a questionable spy movie.

This sorta worked okay: he figured out the one for his speakers, oscilloscope, and keyboard fairly quickly. However, one set of wires was giving him a lot of trouble. These wires were transmitting the following messages in parallel:

Wire 1:    VEEHVUVVEVEVUEHEGEUVHEUEHHUEGEUEEVEHVVEEHEVEVVVHUEUEUEEEEHE
Wire 2:    44HIGH4GG4G4H4PGGHHGR4HGHRHHGHHG44HRG4HHPGGH4G4HHHH4H4GH4PG
Wire 3:    44GSUU4UE4E4U4LNGHUUL4UEHLHHGHUE44VLU4GHLEUG4U4HUVH4U4EG4LN
Wire 4:    44HIGI4G1414I4JHGHIGH4G1HHHHGHI144HHG4HHJ1GH4G4HIHH4G41H4JH
Wire 5:    4EEHVH4VE4E4HEHEVEHVHEGEEHUEVEHEE4HHV4EEHEVE4V4EHHUEGEEEEHE

Wire 1:    VGEVEVUUEVEHEHEUVVHUEUEHEEUEUHHVHVGEUEUVVVUEGEUVHEEHEEGEEHE
Wire 2:    GG44GGHHHG4HGRHH4GRHHH4PH4H4HHP4H4GHHHHGG4HHGHHGRHGPHHG4HHH
Wire 3:    UG44EUHHHU4HELHH4ULUHU4LG4H4HHL4H4GVUHUUU4HHGHUULVNLVVG4HHV
Wire 4:    GG441GHHHG4H1HHH4GHGHI4JH4H4HHJ4H4GHHHIGG4HHGHIGHHHJHHG4IHH
Wire 5:    VVE4EVUUEVEEEHEU4VHGEHEHEEUEUEH4E4VHUEHVV4UEVEHVHHEHHHVEDEH


(The messages have been cut into two parts for readability only). After transmitting the above, there is a short pause, then the messages repeat.

Can you determine the meaning of this data, and to which device the wires are likely connected?

Yes, historical and technological accuracy has gone out the window with this one. Pretend this is a parallel universe or something. :)

• The title makes me want this to be related to Indiana Jones. – Ian MacDonald Dec 7 '16 at 21:50
• Wires transmit voltages, not letters. Are we supposed to assume these are, say, ASCII characters sent one bit at a time? Or should we suspend disbelief and suppose that somehow the wires really are transmitting actual letters and digits? – Gareth McCaughan Dec 7 '16 at 23:19
• @GarethMcCaughan see the footnote :) I've actually been toying with the idea of adding a hint which explains a bit more how this is working, but that may give away too much too soon. For all intents and purposes, you can assume that the sender, receiver, and reader all speak the same protocol. It may be that each symbol is a specific voltage, or each symbol is indeed ASCII or some similar variant. All are viable, though some may be better suited to the task at hand than others. – Phlarx Dec 8 '16 at 0:17
• I'd wait longer before dropping hints, I think. I did see the footnote but thought my question was still worth asking -- but I'm not surprised to see you not answering it just yet. – Gareth McCaughan Dec 8 '16 at 1:04
• The frequencies are quite different on the different wires. 4 appears 26 times on each of wires 2,3,4, 13 times on wire 5, and never on wire 1. V appears 26 times on wire 1, 19 on wire 5, 7 on wire 3, and never on wires 2 and 4. Etc. (Those 26s and 13s are suggestive, but my current guess is that they are coincidence.) – Gareth McCaughan Dec 8 '16 at 3:23

For convenience, here's a transposed representation of the data on the 5 wires:

V4444 E444E EHGHE HISIH VGUGV UHUIH V4444 VGUGV EGE1E V4444 EGE1E V4444 UHUIH E444E HPLJH EGNHE GGGGV EHHHE UHUIH VGUGV HRLHH E444E UHUGG EGE1E HHHHE HRLHH UHHHU EHHHE GGGGV EHHHE UHUIH EGE1E E444E V4444 EHVHH HRLHH VGUGV V4444 EHGHE EHHHE HPLJH EGE1E VGUGV EHGHE V4444 VGUGV V4444 HHHHE UHUIH EHVHH UHHHU E444E UHUGG E444E EGE1E EHGHE E444E HPLJH EGNHE VGUGV GGGGV E444E V4444 EGE1E VGUGV UHHHU UHHHU EHHHE VGUGV E444E HHHHE EGE1E HRLHH EHHHE UHHHU V4444 VGUGV HRLHH UHUGG EHHHE UHUIH E444E HPLJH EHGHE E444E UHHHU E444E UHHHU HHHHE HPLJH V4444 HHHHE V4444 GGGGV EHVHH UHUHU EHHHE UHUIH VGUGV VGUGV V4444 UHHHU EHHHE GGGGV EHHHE UHUIH VGUGV HRLHH EHVHH EGNHE HPLJH EHVHH EHVHH GGGGV E444E EHHID HHHHE EHVHH

Notice that some of these quintets repeat. In fact, of the 118 quintets used, there are only 18 distinct values. (Notably, this is less than the number of letters in the alphabet.) For convenience, I'll assign these to the letters a-r in alphabetical order:

• E444E = a
• EGE1E = b
• EGNHE = c
• EHGHE = d
• EHHHE = e
• EHHID = f
• EHVHH = g
• GGGGV = h
• HHHHE = i
• HISIH = j
• HPLJH = k
• HRLHH = l
• UHHHU = m
• UHUGG = n
• UHUHU = o
• UHUIH = p
• V4444 = q
• VGUGV = r

This reduces the message to:

I shall attempt to solve this as a standard cryptogram. So I'll start by listing the ciphertext letters in order of frequency.

a: 13, q: 13, r: 13, e: 10, b: 8, m: 8, p: 8, g: 7, h: 6, i: 6, k: 6, l: 6, d: 5, c: 3, n: 3, f: 1, j: 1, o: 1

But the lack of spaces makes this hard to solve. Maybe one of the symbols is a space. But which one?

If we assume that the plaintext is “trimmed” with no spaces at the beginning or end, that rules out q and g. If we further assume there are no double spaces, that rules out g, m, and r. So maybe a is a space, since it's the most frequent of what's left.

q djrpqrbqbqp kcheprl nbilmehepb qglrqdekbrdqrqipgm n bd kcrh qbrmmer iblemqrlnep kd m mikqiqhgoeprrqmeheprlgckggh fig

But that leaves us with a 27-letter “word”, which seems unlikely. Hmm...

Edit: Oh! I've got an idea! If you take each wire symbol and treat it

as a base-32 digit (0123456789ABCDEFGHIJKLMNOPQRSTUV) represented in binary, then the pattern of 0's and 1's makes a 5×5 pixel glyph for a character.

Knowing this makes the “cryptogram” far less cryptic. It's solution is: