W3C home > Mailing lists > Public > www-math@w3.org > May 2004

Re: http://www.w3.org/TR/mathml-units

From: Stan Devitt <jsdevitt@stratumtek.com>
Date: Sat, 22 May 2004 13:42:54 -0400
Message-ID: <40AF911E.7000405@stratumtek.com>
To: Robert Miner <RobertM@dessci.com>
Cc: steve.swanson@mackichan.com, www-math@w3.org

Steve and Robert,

Sorry for replying so late on this issue. I actually had several things 
in mind when we wrote this example as class="MathML-Unit" but primarily 
it boils down to simple practices when styling any document. 

It is for much the same reason I prefer to look at the markup of the 
source documents instead of the markup of  the same documents after they 
have been transformed to the detailed format required by a particular 
device whether that be HTML, FOP,  eps,  or some specific printer.

1.   Names force the style to be  treated the style uniformly throughout 
the document.
2.   Named styles are easier to read in the source than the same 
document cluttered with style details.
3.   Named styles are easier to change.
4.   Names convey additional semantics.

There is certainly no intent to limit the markup to CSS aware 
applications.  There are numerous techniques available for transforming 
such source document to a format  that can be rendered by a particular 
application, and I would always author a document for such a device by 
using named styles and transforming them using one of the many 
transformation tools available.

Having said that,  I second all of Robert's remarks below.


Robert Miner wrote:

>Hi Steve,
>>Just a few gentle remarks.
>>The presentation examples in the note use
>>  class='MathML-Unit'
>>But there is no explanation of this value.
>>The example in section 8 is a little funny.  Why not mark it up as written
>>(i.e. as a bevelled fraction with fences)?
>I don't think anyone every responded to you on this.  
>I think the idea of the MathML-Unit value was that it should have the
>same effect as setting mathvariant="normal".  Of course in a CSS-aware
>MathML implementation, then you could accomplish the
>mathvariant="normal" part by actually setting up a CSS rule using the
>MathML-Unit class.  But for cases where you don't have a CSS-aware
>MathML implementation, you probably want to put in the mathvariant
>explicitly.  Even in those cases though, including the
>class="MathML-Unit' at least marks the construct as a unit.
>In any event, now that a Math Interest Group has been rechartered, we
>can republish this Note, with updates and errata.  So if you have
>other thoughts on the matter, please post them.  And I agree about
>section 8...
>Dr. Robert Miner                                RobertM@dessci.com
>W3C Math Interest Group Co-Chair                      651-223-2883
>Design Science, Inc.   "How Science Communicates"   www.dessci.com
Received on Saturday, 22 May 2004 13:42:29 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:27:35 UTC