Lily,
I agree. See my other email. The thing I pointed out there is that we need
better arithmetic support in IRML to achieve this easily (regular expression
becomes cumbersom for this, and sometimes even impossible when it gets very
complicated).
I tend to agree with Hilarie that arithmetic support and function
evaluation in rules can soon get too complicated.
We need to understand whether something like that is really required,
whether there are use cases (Derek mentioned one) and whether there
are alternatives to integrating this into the rule language - Andre
gave some hints for alternatives, and Hilarie outlined the idea for
special subsystems. Both approaches could solve the problem without
"full-fledged" arithmetic support in the rule langage...
-Markus