Re: New /etc/unicode POSIX system database

From: Kenneth Whistler (
Date: Mon Nov 30 1998 - 21:26:22 EST


> What would be extremely important is to define a file format for a
> standard /etc/unicode database ... ... and should
> provide the latest update version whenever the standard is extended.

What the Unicode Consortium will do is provide the character data
it always has, plus some yet-to-be-determined additional increment
of character data, in machine-readable and easy-to-parse text files
in defined formats. And we hope that the next round will include
better version-control mechanisms than we have had in the past.

It will then be up implementers to decide what to do with those
data files. If the Unix community decides it wants to use some
portion of that data in a text database in /etc/unicode (as Markus
suggests) or /usr/share(/lib?) (as John Fieber suggests), that is
up to them. The Unicode Consortium would neither define the exact
format of that text database nor its location. If, as Dean Abrams
suggests, it is better to provide that data to applications through
an API implemented in a binary (either as a system-level API or
as an add-on library, plug-in, component, or whatever), then
vendors are free to use the character data to implement such an
API and get people to use it.


This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:43 EDT