Re: TMQL, next round

On Fri, Mar 09, 2007 at 01:08:54AM +0100, Lars Marius Garshol wrote:
> What we really need for all built-in types, I think, is:
>  - a literal syntax,
>  - comparators.
> 
> Possibly also atomification and de-atomification.

Just, for the record, not sure whether this is practical or not:

It would also be possible to get away without *ANY* primitive data
type being adopted into TMQL. Accordingly, standards-conformant
queries could not use 3.14 but would have "3.14"^what:ever, but we can
allow implementations to inject their own data type with specifying
what you mention above:

   - literal syntax (text serialization/deserialization)
   - binary serialization/deserialization
   - a set of infix, prefix operators

That would be very modular, but clearly also very minimalistic.
Again, not sure how pragmatic it is.

\rho