Localization issues in glibc-2.1.95

Damjan arhiv@freemail.org.mk
Fri Oct 13 06:16:00 GMT 2000


Hello, I'm testing the new features of the localization in glibc-2.2,
and updating the macedonian localization, and I encountered some
problems.

1. I have made a slight change to localedata/locales/iso14651_t1 so that
it sorts macedonian cyrillic right. Here is a diff (sorry if it's not
in a good format):

	242d241
	< collating-symbol <CYR-TSHE>
	247a247
	> collating-symbol <CYR-TSHE>
	498d497
	< <CYR-TSHE>
	503a503
	> <CYR-TSHE>

2. When I execute this: 
	# LC_ALL=mk_MK.UTF-8 locale -c LC_PAPER
I get:
	LC_PAPER
	297
	297

although in the locale source I have: 
	LC_PAPER
	height   297
	width    210
	END LC_PAPER

is this the right behavior?

3. Also this command "LC_ALL=mk_MK.UTF-8 locale -c LC_TIME"
first shows undecoded (..<0xd1><0x81>..) the names of the months and
days and then dumps core with a "Segmentation fault". If I try with a
iso-8859-5 locale it shows the names OK, but still dumps core.

4. Can "country_isbn" in LC_ADDRESS be a string? (All I could find about
the ISBN prefix of my country was 9985+86, what's the meaning of that)


I have glibc-2.1.95, installed on a separate partition.

-- 
damjan





More information about the Libc-alpha mailing list