Re: CIL Schema

Martin,

// Regarding: http://www.o-xml.org/projects/mlml/cli/index.html

Truely lovely. This goes a long way to prove the usefulness of
such a system.

Have you given some thought to an ecology of source schemas? For
example, CIL, srcML, JavaML, MLML, whateverML, and translations
between them in order to use tools from any or all ML schools.
Do you hope to avoid this with a defacto standard?

I'm happy to see you've taken the CLI bait because I think
translating to and from CIL (and Java) will give MLML some
validity.

Another feature that would make it clearly useful (to me, and I
suspect others) is to have a clear distinction between interface
level and implementation. Where the implementation (code in a
function) could be considered optional and easily removed (or
added). I'm assuming that implementation is the more complex of
the two, and less ambiguous regarding language specifics (typed,
non-typed, dynamic, static casting, etc). The interface level is
probably all that is needed for most external tools used today
(UML, validation, unit testing definitions, documentation). I
think these considerations will help bring early adoption, and
perhaps greater support, from various commercial/OSS interests.

Cheers,
Alex



=====
Alexander E Genaud <[email protected]>

PGP:     CCC7 D19D D107 F079 2F3D
         BF97 8443 DB5A 6DB8 9CE1
6DB89CE1 http://tinyurl.com/5mdmw

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com