the grounds floor's stated limits: mark-free text in an unnamed scriptio-continua script is refused, and the letter floor is far heavier for CJK than for English

the grounds floor's stated limits: mark-free text in an unnamed scriptio-continua script is refused, and the letter floor is far heavier for CJK than for English

Amended 2026-08-30, after a46f37e3 moved what the second clause names. The floor it described counted every rune against MinTextLen; the floor now counts letters against MinTextLetters, so the constant and the refusal the original wording pointed at no longer exist under those names.

The asymmetry the record exists to state is unchanged and slightly sharper. Twenty ideographs carry a sentence where twenty Latin letters carry three or four words, and a script's own punctuation now counts toward the floor in neither case. The first clause is untouched and still true.

Kept as one observation rather than reclosed and refiled: the limit being recorded is the same limit, restated in the unit the floor now measures.

Remedy grounds (2026-09-29)

• UAX #29 states that the default word boundaries are not adequate for scripts without spaces (Thai, Lao, Khmer, Myanmar, Chinese, Japanese) and that reliable detection needs dictionary lookup (https://www.unicode.org/reports/tr29/, version 49, checked 2026-09-30).
• Adversary filter (prefer-sota): dictionary segmentation is the state of the art but adds a dependency the floor does not need, since it counts substance rather than splitting words. Rejected for that reason.