Compete to crack secrets lost to history—for fun, clout, and (no) profit

• STATUSACCEPTED
• BY@offgramercy
• WRITE-UP8 September 2026

Problems › 103 › solution › 1

A complete note, with its slips of the pen preserved

Separating signs from punctuation

The message changes script partway through a physical line. Six rows are concealed, the next starts with three signs and continues in ordinary cursive, another cursive line follows, and a final three-sign group sits below it. The open continuation gives an internal check on where the private alphabet stops.

Comma-like marks mostly divide words. A dot after the opening group serves as a divider, but an isolated dot in the second row occupies a word position and reads I. The crossed loop after the penultimate concealed phrase is an ampersand. Those distinctions are part of the reading, not punctuation to discard before attempting it. [2]

Complete reading and its spelling

IT IS SUCH A LONG
TIME SINCE I SENT YOU
A CARD THAT I CANNOT
REMEREB [intended REMEMBER] HOW TO WRITE IT
AS WELL AS I USED TOO [intended TO]
REMEMBER HOW TO KISS &
HUG Hope to see you, and-
Remember this afternoon.
NAN

The note says that it has been so long since the writer sent a card that they cannot remember how to write it, then asks the recipient to remember how to kiss and hug. Its ordinary cursive continuation says, “Hope to see you, and- / Remember this afternoon.” The final group reads Nan.

The source writes REMEREB at the first occurrence of intended remember. The second, correctly written occurrence supplies an internal comparison. It also writes USED TOO. A readable edition may supply remember and used to, but those corrections must remain visible. An earlier proposal “SHOULD / DO YOU” was withdrawn after source review: the physical groups have four and three letters, reading USED / TOO.

Accounting for every occurrence

The blinded source inventory assigned labels such as C01 to shapes, without supplying letter values. A plus sign in this table joins two visible components into one operative character. The ledger below is useful precisely because a neat global alphabet cannot be asserted from the coarse visual labels. [3]

What the checks establish

The final source record reports two source passes and an independent attack on the native image. They reconcile every concealed occurrence, the change to cursive, the bottom signature, and the two spelling defects. This replaced the earlier thumbnail-quality gate.

There is an important limit. A single substitution table over the twenty broad visual classes fails: similar-looking uprights, loops, and forks can carry different readings, and two characters were initially inventoried as separate components. The occurrence ledger preserves those distinctions. It establishes complete coverage of the accepted reading, but is not an independently transferable alphabet and should not be described as one. The public record does not furnish a separate, previously unseen card decoded with that alphabet.

What is not claimed

The postcard is HCPortal 1186, DECODE 3349, collection ordinal 293, cataloged 12 October 1904. The portal’s recipient string is AE Haware; only the bottom address line San Miguel is secure in the source reading. Neither the catalogue’s Dorchester route nor a guessed expansion of the address is needed to recover the message.

The bounded prior-solution audit found no public exact-item plaintext or same-key sibling. Blank public solution fields do not clear the contents of private collections. The result is an exact-item campaign reading, with the source spelling retained and the key limitation stated. The watermarked source derivative has no established public reuse permission. [2]

References

• README.md
• FULL TRANSCRIPT.md
• Complete occurrence and model audit