Re: Variant locales?

Date: Mon Apr 22 2002 - 16:24:56 EDT

On 04/22/2002 02:18:52 PM John Hudson wrote:

>This raises a related topic. The codes to tag combinations of script and
>(the poorly named) language system in OpenType together represent a set of
>specific typographic conventions that may or may not be identified along
>language lines and which are best presented to users independently of
>locales (however locales end up being defined)...

There is a section in my paper that discusses typographic conventions. I
think Paul got a copy of an early draft, but I didn't get any comments from
him. Even so, I see what I have proposed as just that -- a proposal -- in
the hope that others (like yourself) who are able to think through both the
real-world linguistic-related issues as well as practical IT implementation
issues can also work together to refine and revise as needed.

>More recently, in discussions
>with Paul Nelson at Microsoft, I have come to the conclusion that there
>be instances in which it might be desirable to tag specific sets of
>typographic conventions independently of 'writing system' per se. This, of
>course, depends on how one is defining the term writing system. Peter, how
>do you define it?

I have defined writing system as the implementation of a set of characters
and writing behaviours for representing a particular language. So, whenever
two languages are involved, this implies two distinct writing systems. That
may not match the way some people view the world; e.g. some might react to
a suggestion that (say) English text and Indonesian text involve two
different writing systems. But the model doesn't have to exactly match
every detail of reality, and specifically doesn't have to be perfectly
relational; it just has to be implementable, practical and adequate for
realistic IT purposes.

Now, for what you're talking about, I think I've mentioned to you before my
idea that the OpenType "language" tags should really be conceived of as
denoting groups of writing systems that share common typographic
conventions. So...

>It might be possible to define the term to include
>distinctions between different typographic conventions (is, for example,
>English written with the long-s a different writing system from modern
>English?), but it is also possible to conceive of typographic conventions
>as a separate layer above writing systems.

That is pretty much the conclusion I come to in the paper.

If you'd like to get a peek at the paper, I can send it to you, or you can
get it from the ISO/TC 37/SC 2/WG 1 doc repository
( It's long, but it
attempts to deal with a lot of issues. Typographic issues are the specific
focus of section 6.2.

- Peter

Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <>

This archive was generated by hypermail 2.1.2 : Mon Apr 22 2002 - 17:19:23 EDT