W3C home > Mailing lists > Public > www-math@w3.org > August 2003

Re: Accented variables

From: Robert Miner <RobertM@dessci.com>
Date: Thu, 7 Aug 2003 10:03:49 -0500
Message-Id: <200308071503.h77F3nx03163@wisdom.geomtech.com>
To: jpederse@wiley.com
CC: S.Pepping@elsevier.nl, www-math@w3.org


Hi.

> Thanks. I was interested to see so many combining marks on this list. Does
> this mean that, despite their deprecation, MathPlayer allows for them in
> MathML markup, or is it do with dealing with them outside of MathML?

Some of each.  The situation is this: MathPlayer reads the IE (HTML)
DOM to figure out what MathML it should render.  Consequently, we are
seeing a pre-processed input stream, but the rules of the
pre-processing are unclear to us.  For example, IE replaces entity
names it recognizes with characters, but the list of entities it knows
is some private MS list -- more than strict HTML, less than MathML.
The IE HTML parser clearly isn't a true XML parser, so I'm sure it
doesn't do the combining accent normalization in full generality.  But
just exactly when/if it does is not clear to me.  The only way I know
to find this stuff out is to conduct painstaking experiments.

So, as far as MathPlayer is concerned, we put in rules to handle
reasonable characters whether or not we were sure we would ever see
them in IE's DOM.  Also, we do use the MathPlayer rendering engine in
other contexts, and it is perfectly plausible there characters might
turn up in *some* context.

Sorry there isn't a better answer.

--Robert

------------------------------------------------------------------
Dr. Robert Miner                                RobertM@dessci.com
MathML 2.0 Specification Co-editor                    651-223-2883
Design Science, Inc.   "How Science Communicates"   www.dessci.com
------------------------------------------------------------------
Received on Thursday, 7 August 2003 11:04:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 20 February 2010 06:12:55 GMT