Warning: rabbit hole
https://polymaths.social/@rl_dane/statuses/01K1XR4D4GCKDD7PQMSRAMV5EA
You'll recognize this one as well, no doubt. There's probably an old Usenet FAQ somewhere with it, although I haven't checked.
But it got me to thinking of when the idea of a-z collating *entirely after* A-Z became a norm; only to falter again when ASCII systems met other countries. When the era was.
I checked with a couple of my old printed dictionaries from the 20th century, and they had case-insensitive collation in English. One even put 'a' ahead of 'A'.
It's definitely thus an #ASCII notion, but did it pre-date #Unix? Did it pre-date regular expression syntax? Or shell globbing?
My initial educated guess would be that the golden period of A-Za-z was the middle 1960s with ASCII and some Unix predecessor to the early 1990s when all of the books/posts/whatnot on the new Standard C were spreading the knowledge of locales.
#UnixHistory #retrocomputing #locales #collation #RegularExpressions #fnmatch