[sc34wg3] CTM / TMCL Issue: Wildcards - Do we need them?

[Robert Barta]

On Sun, May 25, 2008 at 06:24:23PM +0200, Lars Heuer wrote:
> Now, I took a look at the TMCL draft dtd. 2007-12-08 and I wonder if
> we need the "?" and "?foo" notation at all; it seems that the []
> notation is enough.
> 
> A random example from TMCL:
> 
>   def AssociationTypeConstraint()
>       ?atc1 isa associationtype-constraint
>       AddConstraintToSchema(?atc1)
>   end
> 
> 
> which can replaced by:
> 
>    def AssociationTypeConstraint()
>        AddConstraintToSchema([isa associationtype-constraint])
>    end

[] can replace a named wildcard _only_ if the structure of the
information is restructured into a tree form.

I am not sure whether it is _always_ possible to do that with every
kind of expression. But I am pretty sure that - even if is - it is
inconvenient for a user to perform this mental exercise.

Given the many ... edges ... CTM has saving on the wildcards does not
really make a difference. From an implementation viewpoint both are
no-brainers.

My succus: Keep it as-is, maybe explain one with the other.

\rho