W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > November 1996

Re: ERB discussions and decisions: entities

From: Michael Sperberg-McQueen <U35395@UICVM.UIC.EDU>
Date: Fri, 15 Nov 96 12:48:42 CST
Message-Id: <199611151854.NAA02784@www10.w3.org>
To: W3C SGML Working Group <w3c-sgml-wg@w3.org>
On Thu, 14 Nov 1996 20:54:23 -0500 <lee@sq.com> said:
>Note that &quot; is a single quote, which does not occur in ASCII.
>(many implementations include a single quote at position 39 instead
>of an apostrophe, but this is in error)

The documentation I checked said it's a typewriter double quote,
ASCII x21 or whatever.  Unfortunately I can't remember what doc
that was; what's your source?

ISO 10646 and other ISO character sets uniformly use 'quotation mark'
-- which is what the comment in the entity set says -- for the "
symbol, and 'apostrophe' for the ' symbol, and 'grave accent'
(and possibly something else indicating its use as an open single
quote) for the ` character. That seems to me reason enough to
believe the &quot; should be treated as ", and &apos; as '.

>I am very pleased with the decision to reduce the list of predefined
>entities to a manageable number.  I'm not actually sure why any beyong
>&lt; and &amp; are needed, although &gt; is good for symmetry.

providing quot and apos makes it possible for attribute values to
contain both characters; it also allows the user to choose which
delimter to use, if they prefer to use the same one all the time.
Strictly speaking, one would have sufficed.

>The current situation is so much of an improvement that it seems almost
>churlish to ask this, but why is &apos; included when it is not
>recognised in plain text?

It is recognized in literals and may need to be escaped there.

Michael
Received on Friday, 15 November 1996 13:54:18 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 10:03:43 EDT