Re: Etruscan

From: Mark Leisher (mleisher@crl.nmsu.edu)
Date: Mon Nov 03 1997 - 15:52:00 EST


>> I think you don't want to use overriding to reverse things. What
>> you want is a mechanism at a higher level (e.g. styles). Setting
>> a text in left-to-right direction that runs over several lines
>> with overrides is not absolutely impossible, but extremely
>> tedious; every character has to be enclosed in an LRO-PDF pair.

Did you intend to say every line? I'm curious to know if there is a
example of ordinary text (pathological examples are too easy to come up
with) that requires wrapping every character instead of every group of
characters with the same directionality.

>> So for the directionality, it will be so that the inherent
>> directionality can be used in plain text, the other one probably
>> shouln't be used in plain text, but only formatted. Bustrophedon
>> can only be achieved with styles, not with BIDI controls.
>>

    John> Or with hard line breaks.

    John> Of course, I don't hear anybody asserting that Unicode *ought*
    John> to support boustrophedon in plain text, anyway.

Not necessary to support boustrophedon in plain text. With language
markup, it turns out to be quite simple.

Start with the user selected directionality for Etruscan, determine the
line wrap point, add the appropriate directional markup at the beginning
and end of the line segment, change to the opposite direction, and do it
all over again until the whole line is rendered.

Line separators are simply a signal to change to the opposite
directionality, and multiple line and paragraph separators cause a reset
to the original user-specified directionality.

>> - The fact that BIDI implementations will have difficulties to be
>> updated with new character properties when new characters are
>> added (this would in essence be a strong suggestion that
>> everything except U+0580-U+7FF is LTR) would get consideration
>>

    John> Er, I don't think we can do that. I don't think implementers
    John> should count on all RTL scripts being encoded in that one
    John> range. Certainly we anticipate various dead RTL scripts being
    John> put on Plane 1 (Meroitic, South Arabian).

Let's not forget the not-quite-dead RTL scripts like Dihevi.

    John> People should be getting the directionality properties from
    John> the properties file, nowhere else.

Exactamundo.
------------------------------------------------------------------------
mleisher@crl.nmsu.edu
Mark Leisher "A designer knows he has achieved perfection
Computing Research Lab not when there is nothing left to add, but
New Mexico State University when there is nothing left to take away."
Box 30001, Dept. 3CRL -- Antoine de Saint-Exupéry
Las Cruces, NM 88003



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:37 EDT