Ted Harding <[EMAIL PROTECTED]>:
> As you can see,
> the lineup does its job, despite the intervening text. (And
> it carries over past page-breaks too).
>
> I don't see how this is compatible with your description
> of troff "treating multiple adjacent .EQ/.EN spans as a
> single display"? Maybe it does work like that if the EQ/EN
> blocks are adjacent, as in your example; but there must be
> some other mechanism there when they are not (as in mine).
You're right. I explained the situation poorly. The respect in which
eqn treats these as adjacent displays is just that it remembers mark
and lineup locations after an EQ/EN span has been processed.
Intervening text doesn't matter.
I guess I should have said that all EQ/EN spans are treated as a single
display for mark/lineup purposes, because that's how it works internally.
MathML doesn't have this behavior, When you have multiple MathML
markup spans in a document, later ones know nothing about earlier ones;
<alignmark> information is forgotten at the closing </math> tag.
To do the nearest equivalent in MathML, all the displays that want to
share alignment state have to be wrapped in a top-level table element.
--
<a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>
