Re: TMQL, next round

On Wed, Mar 07, 2007 at 03:28:58PM +0100, Lars Heuer wrote:
> I think the date/time datatypes are a 'nice to have' but I don't mind
> if they don't make it not into the standard.
> Sidenote: SPARQL does not support any date/time datatype natively.

I think it is too early to say something about SPARQL.

> Ok. The question is, if we want all functions in the default
> namespace or if we provide an own namespace for it

Or - if we assume that 'tmql' is the default namespace - whether
functions should live there or have their own namespace. And ....

> and define a minimum set of functions which all impl. have to
> support.

Exactly. Which implicitely means "whatever we choose will be wrong for
someone" :-)

\rho