W3C home > Mailing lists > Public > www-international@w3.org > October to December 1998

RE: WebFont?

From: Chris Pratley <chrispr@MICROSOFT.com>
Date: Wed, 16 Dec 1998 02:16:18 -0800
Message-ID: <5F68209F7E4BD111A5F500805FFE35B90DF5C80D@RED-MSG-54>
To: "'Martin J. Duerst'" <duerst@w3.org>, Jerzy Karczmarczuk <karczma@info.unicaen.fr>
Cc: www-international@w3.org
Of course, another option for Arabic browsing is Internet Explorer 4.0cs
(Complex Scripts version), as well as IE5 beta 2 (any language). IE supports
the BiDi HTML4 extensions for proper BiDi rendering, multiple Arabic
encodings, UTF-8, etc. IE can render Arabic on Windows95, Windows98, and NT4
or greater, and provides this support on any language version of these
operating systems. IE uses standard TrueType fonts for Arabic display, so
you can pick from many widely available fonts.

-----Original Message-----
From: Martin J. Duerst [mailto:duerst@w3.org]
Sent: December 16, 1998 1:40 AM
To: Jerzy Karczmarczuk
Cc: www-international@w3.org; www-international@w3.org
Subject: Re: WebFont?

At 08:24 98/12/16 +0000, Jerzy Karczmarczuk wrote:

> There are several issues here. As far as the dynamic fonts
> (Bitstream style, .pfr packing; IExplorer uses also different dfonts)
> are
> concerned, there is nothing in Netscape to support or not to support
> an Arabic *font*, provided that:
> A. the appropriate glyphs are accessible through LINK FONTDEF, and the
>    document defines its name correctly.
> B. their coding corresponds to one of established codings; it may be
>    user-defined, or Unicode (or any other "faked"...)
> C. The document author prepares himself the layout: right align, every
>    line coded from right to left, etc. Horrible, but possible.

Definitely a very bad idea, not up to the standards.


>    (In
>    fact, I don't really know - if Netscape declares in the "preferences"
>    the possibility to accept Unicode, it should provide for some
>    control characters as well; I tried this once and it didn't work...)

Unicode includes an enormous amount of characters. The HTML 4.0
spec explicitly says that you are not required to render all of them;
it was just not possible to make such requirements that nobody was
ready to accept.

For Arabic and Hebrew, you need bidirectional controls. They
exist in Unicode as "control characters", but mixing control
characters and markup is not a good idea. Therefore, HTML 4.0,
inherited from RFC 2070, contains appropriate markup for
bidirectional text. Here again, a browser is not required to
support it, but if it doesn't, it shouldn't show Arabic/Hebrew
at all to avoid showing a complete mess.

For details, please see
http://www.w3.org/TR/REC-html40/charset.html#h-5.4 and
http://www.w3.org/TR/REC-html40/struct/dirlang.html#h-8.2.


There are browsers that do it, for example Tango from Alis.
Others are working on it.


Regards,   Martin.





#-#-#  Martin J. Du"rst, World Wide Web Consortium
#-#-#  mailto:duerst@w3.org   http://www.w3.org
Received on Wednesday, 16 December 1998 05:16:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 19:16:53 GMT