RE: design issue: query type checking via DTD

I think Rick put it best when he characterized this support for assisting in
query construction as "nice-to-have". Arguments about what the XML look like
aside, the client side doesn't absolutely need this feature int the DASL
protocol. A "correct" query can always be constructed if the client expends
the effort to ensure it; protocol support is not critical to make this
happen. Even with the syntax proposed by Alan, incorrect queries constructed
it's just that one class of incorrectness is ruled out.  

We should keep the grammar extremely-simple. More complex-grammars are fine
when they offer clients something they *must* have. Again, I believe this is
a "nice-to-have".

Thanks,
Saveen

Received on Monday, 8 June 1998 12:53:55 UTC