![]() |
> I downloaded and installed the file : hylafax-v4.0pl2-3rh5.i386.rpm. > > As I understand this release should handle international characters, but = The source release does not as far as I know; you will have to trace and ask the compiler of the .rpm as to what patches he applied. > when I try to fax a textfile containing norwegian characters I get the = > message: Can not determine file type. This is because the rule for recognizing text files does an ascii test, which looks for printable characters in the current locale. Other parts of Hylafax expect the default, C, locale, which uses ASCII as its character set. (In particular the bit that writes out postscript does floating point number conversions and breaks if it is run in a locale which uses "," as the decimal separator, as Postscript uses US/UK conventions exclusively.) You will need to change the code which handles typerules, use a weaker test for text files in the typerules, or try to arrange that just the relevant part of the locale is overriden. Please see the archive for discussions of confusion in the man page for typerules. A better solution might be to find those parts of the code that rely on the C locale and force that locale locally. However, many other programs, e.g. all SCO Unix a.out programs, misbehave when you change the locale globally.