[sc34wg3] Comments on latest TMRM draft

[Robert Barta]

On Thu, Jul 14, 2005 at 09:20:19PM +0200, Jan Algermissen wrote:
> 
> On Jul 14, 2005, at 8:52 PM, Lars Marius Garshol wrote:
> 
> >To say that "values are unconstrained" seems like hand-waving. At the
> >very least some minimal conditions have to be spelled out.
> >
> 
> FWIW, I think the TMRM could only benefit from explicitly saying that  
> values are typed. (They are anyhow, because if they do not have a  
> type, no one can do anything with them. So why not say it? You cannot  
> even compare two values if they have no type.)

[ Speaking again for Tau+ only. ]

When introducing property values, the only thing required _there_ is
that values are from a set. This implicitely makes them 'comparable'
in the sense "this is the same or not" (otherwise the whole set thing
does not make sense).

So, no data type there. But...

> Aside: does the TMRM assume that values can be compared?

Tau+ uses elements from a set V. So these must be comparable using '='
by definition.

> The value types (aka data types) are of course unconstrained.

...But...Yes, Tau+ introduces data types when it comes to the
'disclosure'.

> They can be simple or complex. For the TMRM value types (and values)
> are opaque of course. (Is that what is meant with 'unconstrained')?

I think the term 'opaque' is very appropriate here.

\rho