W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > May 1997

Re: Comments on Part 1: Encoding declaration

From: James Clark <jjc@jclark.com>
Date: Thu, 29 May 1997 14:01:36 +0700
Message-Id: <>
To: w3c-sgml-wg@w3.org
At 11:57 29/05/97 +0900, Murata Makoto wrote:
>Comments on Part 1: Encoding declaration
>The current draft requires that all text entities have
>encoding declarations unless they are encoded in UTF-8.
>(1) Internal entities
>Does this requiement apply to internal entities?  Is it
>possible to apply different encodings to internal entities?

How could it?  An encoding converts a sequence of bytes into a sequence of
characters.  The replacement text of an external entity is a sequence of
bytes, so it needs an encoding to convert it into a sequence of characters.
But the replacement text of an internal entity has been parsed and so is
already a sequence of characters.

>(2) Duplication of encoding declarations.
>Suppose that we have a document comprising one document
>entity and one hundred external text entities.  I believe
>that these external text entities should not be required
>to duplicate the same encoding declaration.
>The encoding specified for the root entity or an external
>text entity should be inherited by directly-referenced
>external entities, unless they have encoding declarations or
>they begin with a Byte Order Mark.

SP does this.  If you don't specify an encoding it gets inherited.  I think
of this as being similar to relative URLs.

Received on Thursday, 29 May 1997 03:18:18 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:26 UTC