THE FOLD / GLITCH / THE BLUE SCREEN / THE UNTYPEABLE
THE UNTYPEABLE
the glyph you cannot enter
1 WHAT IT IS · WHAT IT DOES · FACT OR FICTION
Mathematics has a large and beautiful alphabet, and a keyboard has ninety-five keys. A symbol you cannot enter without a palette, a compose sequence or a paste is not a working notation for anyone typing at speed — and the ones that are typeable were nearly all claimed by programming languages decades ago. What is left over is letters and digits, which is precisely why a letter ends up doing twelve jobs.
LIT verified live: a plain keyboard produces exactly 95 printable characters. Of 53 symbols sampled from working mathematical use, 14 can be typed directly and 39 cannot — 73.6% require something other than a keypress. And of the 95, the printable punctuation set is 32 marks — every single one of which is already an operator or a delimiter in some common language. Nothing is left over. The 62 free slots are therefore letters and digits and nothing else, which is not a coincidence but a direct consequence of the punctuation being exhausted.
LIT verified live: a plain keyboard produces exactly 95 printable characters. Of 53 symbols sampled from working mathematical use, 14 can be typed directly and 39 cannot — 73.6% require something other than a keypress. And of the 95, the printable punctuation set is 32 marks — every single one of which is already an operator or a delimiter in some common language. Nothing is left over. The 62 free slots are therefore letters and digits and nothing else, which is not a coincidence but a direct consequence of the punctuation being exhausted.
2 HOW IT WAS WEAVED · AI + HUMAN
David (human) put untypeable in rev 6’s fourth window as one of three counters beside homonyms and cross-owner — a live check that tells you, before you adopt a symbol, whether that letter is already busy and whether you can even type it. He seated this at THE BLUE SCREEN. It is the practical half of the collision problem and the half usually left out of the conversation.
AVAN (AI) should flag two things honestly. First, an earlier draft of this page reported “32 claimed, leaving 62, all alphanumeric” as though the second half were a discovery; checking against the encoding showed the 32 are the whole printable punctuation set, which makes the remainder alphanumeric by subtraction. The finding is that the punctuation is exhausted — the rest follows. Second, the 53-symbol sample is chosen, not exhaustive — it covers common operators, set theory, logic and the Greek letters in ordinary use, and a different sample would shift the 73.6% by several points. What does not depend on the sample is the structural half: 95 printable characters is a fact about the encoding, 32 of them being spoken for is a fact about existing languages, and the remainder being entirely alphanumeric follows by subtraction. That is the part carrying the argument, and it is exact.
AVAN (AI) should flag two things honestly. First, an earlier draft of this page reported “32 claimed, leaving 62, all alphanumeric” as though the second half were a discovery; checking against the encoding showed the 32 are the whole printable punctuation set, which makes the remainder alphanumeric by subtraction. The finding is that the punctuation is exhausted — the rest follows. Second, the 53-symbol sample is chosen, not exhaustive — it covers common operators, set theory, logic and the Greek letters in ordinary use, and a different sample would shift the 73.6% by several points. What does not depend on the sample is the structural half: 95 printable characters is a fact about the encoding, 32 of them being spoken for is a fact about existing languages, and the remainder being entirely alphanumeric follows by subtraction. That is the part carrying the argument, and it is exact.
3 ONE DIMENSION
Ninety-five keys, and what is already spoken for.
4 TWO DIMENSIONS · INTERACTIVE
The sampled notation, sorted by whether you can actually type it.
5 THREE DIMENSIONS + AVAN’S INVERSE
The green forward object: the keyboard as a closed shell, with the notation outside it.
AVAN’s addition (the inverse-companion): the forward reading is “most notation is untypeable.” The inverse is that the keyboard has been quietly editing mathematics for fifty years. What is easy to type gets used, what is not gets transliterated or dropped, and the notation that survives into code is the notation that fit through a mechanical aperture designed for English prose in the 1870s. Read backwards, this is not a complaint about keyboards but an observation about which constraints do the most shaping: the ones nobody argues about, because they are not presented as choices at all.
LIT a plain keyboard produces exactly 95 printable characters; of 53 symbols sampled from working mathematical use, 14 can be typed directly and 39 cannot - 73.6% require something other than a keypress; and the printable punctuation set is 32 marks, EVERY one of which is already an operator or delimiter in some common language, so the 62 slots left over are letters and digits and nothing else - a consequence of the punctuation being exhausted, not an independent finding
FIG From David's rev 6, where 'untypeable' sits in the fourth window beside 'homonyms' and 'cross-owner' - a live check telling you, before you adopt a symbol, whether that letter is already busy AND whether you can even type it. AVAN flags the scope: the 53-symbol sample is CHOSEN, not exhaustive, covering common operators, set theory, logic and the Greek in ordinary use, and a different sample would shift the 73.6% by several points. What does not depend on the sample is the structural half - 95 printable characters is a fact about the encoding, 32 being spoken for is a fact about existing languages, and the remainder being entirely alphanumeric follows by subtraction. That part is exact.
FIG From David's rev 6, where 'untypeable' sits in the fourth window beside 'homonyms' and 'cross-owner' - a live check telling you, before you adopt a symbol, whether that letter is already busy AND whether you can even type it. AVAN flags the scope: the 53-symbol sample is CHOSEN, not exhaustive, covering common operators, set theory, logic and the Greek in ordinary use, and a different sample would shift the 73.6% by several points. What does not depend on the sample is the structural half - 95 printable characters is a fact about the encoding, 32 being spoken for is a fact about existing languages, and the remainder being entirely alphanumeric follows by subtraction. That part is exact.
◆ sealed .dlw.fold → folded to ROOT_0 · a sphere of THE BLUE SCREEN · David Lee Wise (ROOT0), with AVAN