W3C home > Mailing lists > Public > www-math@w3.org > March 2012

Re: Using content-MathML for computation and analysis in Science and Engineering

From: Peter Murray-Rust <pm286@cam.ac.uk>
Date: Tue, 20 Mar 2012 09:05:02 +0000
Message-ID: <CAD2k14N36ht2=5joWkQAvSS1yiyB8=f=SBPAcd=sc4osX=NBtg@mail.gmail.com>
To: "www-math@w3.org" <www-math@w3.org>
I am now trying to summarise my position and hope the list can guide me.

Goal:
To create a completely self-suffcient, easily installled chemical
computation engine that can be distributed under an OSI compliant licence
(but not GPL).

Note: Some of the functionality could be provided by web-services and users
take their chance on having 24/7 access. (We do similar things for
chemistry). The engine will probably be able to service "high-school
physics" and significant areas of engineering and materials science. The
emphasis is on numeric evaluation, not theorem-proving or semantic
deduction; there will be a small amount of symbol manipulation, e.g.
differentiating analytical expressions

Note: I am prepared to implement a MathML engine for the subset of maths
involved (and have made progress on this). Prototyping and correct
execution is currently more important than performance. The engine should
allow for easy extensibility with functions and vocabulary not native to
MathML2.0. The solution architecture should scale to be usable under
distributed computing, Hadoop, etc.

Question: As I don't want to re-invent the wheel, are there existing
solutions? They should generally be based on APIs or web services, not GUIs.

Assuming not, I shall go ahead anyway.

1. I am still considering whether I should use lambda. The pros are that it
is clean and I suspect relatively easy to implement. the cons are that it
will scare off most users (especially chemists).

2. The code is a series of  statements of intent rather than (say)
statements of fact (e.g. fact(3) = 6) or hypotheses to be proved.

Question: Is there an agreed semantics/vocabulary for this? If not, is
there a commonly used one? Otherwise I need to invent my own (e.g. DavidC
suggested "declare")

If we can agree on this, then I think it will be possible to create a
programming idiom which bridges "high-school" mathematical notation and
executable code.

Many thanks for your help

-- 
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
Received on Tuesday, 20 March 2012 09:05:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 20 March 2012 09:05:35 GMT