Re: Latin ligatures and Unicode

From: John Jenkins (
Date: Tue Dec 28 1999 - 12:14:36 EST

on 12/28/99 5:07 AM, Michael Everson at wrote:

> Ar 11:27 -0800 1999-12-27, scríobh John Jenkins:
>> For me, the AAT and OpenType mechanisms adequately answer this point, as
>> they allow full control over arbitrary (or automatic) ligature generation or
>> overriding.
> But they don't because you can't predict for many instances what "default"
> "turning ligation on" could be. There _are_ no default ligatures for many
> scripts, times, documents, etc. And I know that the designer can add in a
> zillion levels of defaultness, but this is a clumsy kludge that burdens
> designer and user alike.

But AAT and OT don't restrict you to the defaults. The user has complete
control over what actually happens.

> Merle Tenney showed me how the AAT mechanisms work in a font workshop in
> London last year. I was impressed. And I have been thinking about it all
> this time. And impressive as it is, it doesn't satisfy my needs.

I still haven't seen any specific instances where it fails. Inasmuch as AAT
allows *exactly* the kind of behavior you've been asking for (other than
presence in plain text), I don't see why it fails as a model.

It turns on certain ligatures by default.

The user can turn these defaults off in specific instances and turn on other
ligatures in specific instances.

John H. Jenkins

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:57 EDT