CellML Discussion List

Text archives Help


[cellml-discussion] Content MathML editing language: Binary=>n-ary syntax


Chronological Thread 
  • From: p.nielsen at auckland.ac.nz (Poul Nielsen)
  • Subject: [cellml-discussion] Content MathML editing language: Binary=>n-ary syntax
  • Date: Tue, 21 Nov 2006 00:39:07 +1300


On 2006 Nov 20, at 18:48, David Nickerson wrote:

> I'm not sure that developing a new syntax and spending time worrying
> about where to preserve and throw away extraneous brackets fits
> into any
> of these targets. And in all cases, I would think that most users
> would
> expect that when they type in or edit an equation that they can
> save the
> equation and come back to it later and have it look the same as when
> they saved it.
While I can accept that some users will find it useful to enter
equations in a format similar to fortran/C/matlab, I don't think that
providing such a facility should take up any more resources than
necessary. My feeling is that time might be better spent using one
(or more) of the established language syntaxes. This (these) could
then be transformed using a simple XSLT (or similar) into Content
MathML. In this way we avoid stuffing around inventing yet another
syntax - something I really, really don't think is worth the effort.
It also will provide some of the tools required for automatic
translation between common languages and Content MathML. Issues about
automatic reordering of equations are way down the hierarchy of
importance. In this area, I feel that the translator should reflect
directly the input of the user. There may well be sound numerical
reasons why an author might wish to impose a specific order of
evaluation (e.g. to avoid numerical problems due to cancellation), as
well as user interface issues (least surprise to the user).

Best wishes
Poul




Archive powered by MHonArc 2.6.18.

Top of page