# Re: Subscripts in Content MathML

From: David Carlisle <davidc@nag.co.uk>
Date: Thu, 6 Oct 2005 14:43:57 +0100
Message-Id: <200510061343.OAA31802@penguin.nag.co.uk>



> The problem with <ci>a_1</ci> is that it would be treated by our
> algebra engine as an object with name "a_1", and of course this is not
> strictly true.

In this case, I think my original suggestion of
<selector><ci>a</ci><cn>1</cn></selector>

This has the default presentation you want and more or less the right
semantics. the main reason that I've needed to sometimes use
<ci>x_i</ci> (or its openmath equivalent) together with a special
purpose stylesheet is if you need to use these indexed variables as
bound variables, when it's not always equivalent, or desirable, to
bind the whole vector at once. The need to use a special purpose
stylesheet is less of an issue in openmath as openmath has no default
presentation form so you always need _some_ stylesheet.

>  As I said in my original e-mail, the DTD permits the following combination:
>
>   <ci><msub><ci>a</ci><ci>i</ci></msub></ci>    (*)

Yes that's valid but I think the selector form more strongly encodes
what you mean here. But that's just a matter of taste, I think.

> The only problem is to design a content-to-presentation XSLT sheet
> which will transform the content in the above expression while leaving
> the presentation alone.

That should be the default behaviour of any content mathml renderer.

on elipses I think it's best to have an encoding of

> sin[x] = x - x^3/3! + x^5/5! + ... + x^n/n!

as

sin[x] = sum{i=0}^{n} {(-1)^i x^{2i+1)/(2i+1)!}

and leave it to the presentation stylesheet to display using ... and
inline + rather than \sum symbol if that is the presentation choice.

David

________________________________________________________________________
This e-mail has been scanned for all viruses by Star. The