Date: Mon, 16 Jan 95 10:40:44 MST From: terry@cs.weber.edu (Terry Lambert) To: ache@astral.msk.su (Andrew A. Chernov, Black Mage) Cc: phk@ref.tfs.com, dgy@seagull.rtd.com, freebsd-hackers@freefall.cdrom.com, roberto@blaise.ibp.fr Subject: Re: CVS stuff Message-ID: <9501161740.AA23380@cs.weber.edu> In-Reply-To: <wTo5E6lGjY@astral.msk.su> from "Andrew A. Chernov, Black Mage" at Jan 15, 95 11:48:50 am
next in thread | previous in thread | raw e-mail | index | archive | help
> 0xff is valid russian letter and I already do basic things > in source tree to handle it correctly. If you know any code > confused, please tell me. Any code that uses a signed character as an lvalue for a getch/getc, and then checks for -1. Presumably, the response will be "that code is broken". I make the same claim for any code using '\0' as a string terminator, since it assumes a particular encoding, and I make the same claim for any code written not using wchar_t instead of char for strings. I also make it for copyinstr/copyoutstr (which Linux doesn't have, BTW), since it assumes NULL termination by single character of a string. The problem is the distinction between process and storage encoding, and the use of storage encoding in crossing the user/kernel boundry and in internal processing. I have yet to see alternate message catalogs for, for instance, ash, csh, and bash... even though I think unshared catalogs are a mistake in any case, and thus the XPG3/XPG4 code is probably the wrong way to go in trying to do internationalization, unless you like per program catalog bloat. Most code is, by definition, poorly equipped for localization. It's a question of where you draw your line in the sand. Terry Lambert terry@cs.weber.edu --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9501161740.AA23380>