Re: Comments on the DTD and the spec.

Hi Simon,

> The last paragraph of section 3.3.4.1 might also have specified that the
> rowalign attribute on the mstyle element should correspond to that attribute
> of mtable, and that the columnalign and groupalign attributes on mstyle
> should correspond the those attributes of mtable, mtr and mlabeledtr. At
> least, that is how I would interpret it in the absence of such a
> specification.

I don't understand why you suggest treating columnalign and groupalign
differently from rowalign.  All three can be set on mtable, mtr and
mlabeledtr.  All three have default values.  

Since an mtable can only have mtr's and mlabeledtr's as children,
there isn't any way to use mstyle to explicitly set these attributes
on the rows, but not the table.  That is, the mstyle has to either
enclose the whole table, or it can be in the contents of a cell, but
it can't be in between.  Consequently, I don't think there is really
any ambiguity about whether it applies to the mtable or the rows in
practice.  

However, you make a good point about the language in the last
paragraph of 3.3.4.1, and I would prefer to say that these three
attributes uniformly apply to mtable and not mtr, mlabeldtr.  But I
don't want to do that until I understand why you thought they should
be handled differently.

--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 Friday, 20 June 2003 11:07:31 UTC