W3C home > Mailing lists > Public > www-math@w3.org > November 2009

Re: isotech error with lowast?

From: David Carlisle <davidc@nag.co.uk>
Date: Thu, 12 Nov 2009 17:45:37 GMT
Message-Id: <200911121745.nACHjbDu000368@edinburgh.nag.co.uk>
To: bruce@inera.com
Cc: www-math@w3.org, btusdin@mulberrytech.com, dalapeyre@mulberrytech.com

> Shouldn't this be defined as:
> 
> <!ENTITY lowast           "&#x0204E;" ><!--low asterisk -->

One might expect that, but like some others (most notably asymp) the
definitions are somewhat skewed by html compatibility. Many of the html4
entity definitions are somewhat strange but the html4 symbol.ent file
linked from 

http://www.w3.org/TR/html4/sgml/dtd.html

defines lowast to be 

<!ENTITY lowast   CDATA "&#8727;" -- asterisk operator, U+2217 ISOtech -->

The html entity sets are baked into code in multiple browsers of every
desktop and mobile phone on the planet and nothing that is put in a dtd
entity file will change that as most html systems don't read these
definitions from any kind of declarative file, thus is seems futile to
publish an html dtd with definitions different from the ones actually
implemented, and it would be odd to try to make xhtml incompatible to
html with respect to entity definitions.

If the mathml set were incompatible with xhtml then the meaning of lowast
throughout an xhtml+mathml document would depend on the technical
details of how the dtds were combined, but whichever definition "won" it
would mean the same thing throughout the document, you can't make the
expansion of the entity sensitive to which element the entity is in.

The main aim of taking the entity set definitions out of mathml into
their own spec at

http://www.w3.org/TR/xml-entity-names/

is to get a common set of definitions that can be used accross
languages, but where existing usage in different communities is
incompatible, getting a common set means something has to change and the
above considerations mean that essentially if there was a conflict the
html definition was taken.

I hope that explains why things are as they are

David

________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.

This e-mail has been scanned for all viruses by Star. The service is
powered by MessageLabs. 
________________________________________________________________________
Received on Thursday, 12 November 2009 17:46:11 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 20 February 2010 06:13:06 GMT