Re: Trademark Entity and Entity Problems in General

E. Stephen Mack (estephen@emf.net)
Sat, 09 Aug 1997 22:57:59 -0700


Message-Id: <3.0.3.32.19970809225759.00e497a0@emf.net>
Date: Sat, 09 Aug 1997 22:57:59 -0700
To: www-html@w3.org
From: "E. Stephen Mack" <estephen@emf.net>
In-Reply-To: <3.0.1.32.19970810071426.006a2424@mail.netvision.net.il>
Subject: Re: Trademark Entity and Entity Problems in General

Jonathan Rosenne <rosenne@NetVision.net.il> wrote:
>According to the standards, numerical character references are in the
>document character set, not in the encoding used for transmission. Thus
>they do not depend on the charset.

Absolutely -- but I was talking about implementations, not the
standard.  As the test page (a draft in progress) on i18n by Alan
Flavell shows [1], the popular browsers don't get this right.
Specifically, with Navigator 3.01 and IE 3.01, the number entitities
are interpreted according to the current character encoding, not
the document character set.  (A true train wreck.)

An HTML file containing an entity such as &#250; will show
different things in different browsers, depending on what
character encoding is selected by the user or the author
(in violation of the specification).

So change your "do not depend" to "should not depend" -- the reality
is somewhat different from the specification, sadly.

[1] http://ppewww.ph.gla.ac.uk/~flavell/iso8859/internat.html
    go to the tables half-way down
-- 
E. Stephen Mack <estephen@emf.net>    http://www.emf.net/~estephen/