Re: Inconsistency in ISO 8859-1/8 tables

From: Martin J. Duerst (
Date: Mon Jul 14 1997 - 08:22:58 EDT

On Fri, 11 Jul 1997, Werner Lemberg wrote:

> On Fri, 11 Jul 1997, Unicode Discussion wrote:
> > >I thought Hebrew was "bidirectional".
> >
> > Technically, "bidi" has been used to refer to the ability to process both
> > right-to-left (e.g., Hebrew and Arabic) and left-to-right languages. I am
> > unaware of any living language that is written bidirectionally in the sense
> > of boustrophedon.
> To a certain extent, Chinese (as used in Taiwan) needs this. There are
> newspapers which have headers written from left to right and right to left
> on the same page; the main text is nevertheless written vertically.

Please be careful. Indeed Chinese/Japanese texts, in particular titles
and names, are sometimes writter RTL, for example on the right side of
a truck so that they read from front to back. This can be seen as a
one-char-per-column vertical style.

But such things, as well as the question of whether these languages
are written horizontally or vertically, is purely a matter of style
(that's why it is not supported in HTML (neither RFC 2070 nor HTML
4.0) but has to go into style mechanisms such as CSS. Mixed Hebrew-
English (or similar combinations) text, however, needs some structural
information to be rendered correctly at all. This is what is meant
by BIDI, and this is why BIDI support is in HTML (RFC 2070 and HTML

Regards, Martin.

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:35 EDT