From: Arcane Jill > Of course, back in the days of the ZX80 (a device which, by the way, had > its own custom, non-ASCII character set) and its offshoots, there was > indeed a SPACE LETTER - a character which looked like a space, but > acted like a letter, so "Louis XVI" could be made to count as a single > word. It would never line-break. Of course, on the ZX80, all characters > were fixed-width, but if we imagine a proportional font with the same > properties, it is clear that SPACE LETTER would no more stretch or > shrink than would the letter 'K'. > > Whether or not this would be useful to Unicode is another matter entirely, > but it sounds good to me, so I'll raise it for discussion. > Phillippe's idea does have precedent.
Beside the funny allusion to the old ZX80 toy (not sure if ZX80 had this space letter you describe, because I remember that it had in fact only a very small 8-bit character set with a monospaced font available in only one poor bitmap resolution... I also remember the fact that this toy did not have a display interface (so when you needed to perform some computing, you had to suspend the display refresh, and the screen was garbled...) When I was a child, I also played on that toy and it must have been the first "computer" I could play with, and I was not interested in his poor BASIC language, only in its Z80 assembly code... This was the good old time of PEEKs and POKEs... something that new generations will never experiment with, as they are so used to benefit of out-of-the-box omnipotent libraries, without understanding what is really performed in the background. ;-) My real first computer has been a TRS-80, same processor, but at least a serious engine for true applications, and not a toy... But it's true that a space letter was discussed in the past and proposed as a base letter to hold diacritics in the middle of a word and without creating a word break or line break opportunity (at least not more than break opportunities which may be necessary to render a too long word that can't even fit in a single line). May be it could be a candidate for hyphenation/syllable-breaks, but this is highly language-dependant, and not part of Unicode job to define such linguistic hyphenation feature.
