Challenge my assumptions!

Jim Little
Wed, 14 Jul 1999 17:09:04 -0600

Sorry for the long delay in replying to this.  I published 15
new/updated Prism documents last week and then didn't think about Prism
at all until today.  My biggest fear is burnout.  :)

Anyway, iepos, I've thought about the idea of using a binary tree, or
simply any tree, for the meta-metamodel and I've decided to stick with
what I've got.  My reasons are:

1) The two meta-metamodels are equivalent in power
2) Your approach is a subset of my approach.  I.e., if someone wanted to
define a meta-metamodel in terms of binary trees, he could.  But the
converse is not true.
3) I've got a substantial code investment in place, and your approach
isn't unequivocally better.

However, I'm still open to discussion... can you come up with an example
metamodel where a tree-based approach is undeniably better than a
list-based approach?  I tried, but I couldn't think of any.

Regardless of the final decision, I appreciate your comments!  Thanks.