In message "Re: MES as an ISO standard?",
'kuhn@cs.purdue.edu' writes:
> good typography (proportional fonts, kerning, ligatures, etc.) and
> therefore will also not care about combining characters, bidi,
> and representation forms.
I think BIDI is a bit more fundamental than "good typography" :-)!
> Yes, leaving lots of confusion. Am I allowed to use combing characters
> in conforming Java variable names? Will they link to precomposed
> characters? The specification says "Don't know".
That's one issue that always surprises me a bit. Java is a programming
language and, as such, I don't care if I can only write variable names
in 7-bit ASCII. After all, all the keywords are English. The only place
in the source where one really needs Unicode support is in strings and
chars (hard-coded). The language in which I write my program is one
thing, the languages/scripts it can deal with is what I worry about.
> VERY lonely and really miss the company of others offering software with
> UTF-8 support.
Speaking of which, I just went thru a book on JDK 1.1. How does one
write UTF-8 in Java? All I can see are the DataInput- and OutputStreams.
If I understand right, they can read/write strings in UTF-8, but assume
a two-byte length before each. Not quite plain-text files.
Did I miss something? In my original hacking under JDK 1.0.2, I wrote
my own UTF-8 input and output streams because I couldn't find anything.
I was expecting the situation to improve somewhat with JDK 1.1.
Pierre
P.S. Markus, send me a sample of one of your UTF-8 files (in German?).
We'll see if I can read it...
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:35 EDT