Hi all,
Randall kept urging me to make PicoLisp more independent of the case
sensitivity of underlying file systems. He is now a happy Mac user, and
the Mac OS file system is case insensitive (shudder).
As far as I see, there are two places where PicoLisp depends on the case
of file names:
1. In the localization directory "loc/" are files like "DE", "JP" or
"RU" (for the country) and "de", "jp" and "ru" (for the language).
2. Database blobs are stored in paths like "./5/-/F/T.b.pdf", which are
composed from the names of external symbols (here: {5-FTb}).
Point (1) is now defused, because the language files are renamed to
"DE.l", "JP.l" or "RU.l" with today's testing release. Actually, the
".l" extension is more correct, as these files do indeed contain
executable Lisp code.
Point (2) will not be addressed in 32bit PicoLisp, as this goes to
deeply into the printed representation of external symbols, and it would
break to many existing systems. 64bit PicoLisp, however, will handle it
differently.
Cheers,
- Alex
--
UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe
