SC22/WG20 N962
From: Keld Jørn Simonsen [keld@dkuug.dk]
Sent: Sunday, June 09, 2002 2:59 PM
To: sc22wg20@dkuug.dk
Subject: (SC22WG20.3949) NP on i18n functionality TR

Hi WG20-ers,

SC22 requested in resolution 01-26 that "SC22/WG20 develop an NP for
a Type 2 TR, or a Type 3 TR, which addresses the functionality of
15435 and addresses current practice in existing programming languages."

I have been looking into drafting a NP for this, but are a little at
odds, mostly because I did not go to the plenary and thus have
not got the background info.

So here are some thoughts:
SC22 talks about programming languages, are they then only talking
about SC22 standardized languages, or can we include eg Java?

Also they are not mentioning libraries, and I would actually
like to also look at ICU.

Operating systems are not mentioned,
although I think both POSIX and Microsoft windows have interesting
i18n features that would be useful in a PL. 

They mention 15435, so we should look at that functionality too.
But should it be a kind of new advice on i18n functionality to PLs?
Or should it be an experimental standard (Sc22 hints at this
by asking for a TR type 2).

I know that a number of SC22 WGs would like to have advice from
WG20 on i18n functionality, I have heard such from prominent
people at least in C, C++ and COBOL. I would like that this
TR (or maybe another like 10176) would be able to give guidance.

Should it be type 2 or 3? What was the thoughts with this from SC22?
TR type 2 is an experimental standard.

We are surely going to discuss this next week, and I send this
email so that people not attending can give their thoughts too.

Kind regards
Keld