Date: Sat, 26 Mar 2005 22:32:18 +0100 From: Anthony Atkielski <atkielski.anthony@wanadoo.fr> To: freebsd-questions@freebsd.org Subject: Re: mot de passe root Message-ID: <1176697413.20050326223218@wanadoo.fr> In-Reply-To: <86d5tmkst7.fsf@amidala.datadok.no> References: <968029093.20050326192627@wanadoo.fr> <86d5tmkst7.fsf@amidala.datadok.no>
next in thread | previous in thread | raw e-mail | index | archive | help
Peter N. M. Hansteen writes: > I fail to see how switching from one set of message strings files in a > correctly written application would destabilize it. As I've explained, changing string lengths can be a source of trouble; string copies that worked before are suddenly overflowing buffers. Inserting variable data into messages is also difficult with localized software, as the order and format of the variables must often change, and sometimes a lot of extra code is required to accommodate this. Special characters can cause code to fail if it is not prepared to handle 8-bit data--setting the high-order bit is especially likely to make trouble. There are lots of potential problems. And that's just with software that facilitates localization, such as Windows executables. In other environments, it gets a lot worse. > Oh, you're talking about Windows. Yes, there's been a lot of > localization related trouble there. But then we're relatively safe > from the secret brainfarts of Microsoft developers here. As I've already pointed out, localization is actually cleaner and safer in Windows programs than in most other types of software. Windows allows you to isolate strings and similar data in resource files built separately from the executable code. In many other environments, you have to either hard code all language data directly into the program, or you have to write your own code to allow broad localization. -- Anthony
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1176697413.20050326223218>