[sc34wg3] TMQL: Annex A "Predefined Environment"

[Robert Barta]

On Sat, Jul 19, 2008 at 03:30:19PM +0200, Lars Heuer wrote:
> Comments against TMQL draft dtd. 2008-07-15.
> <http://www.isotopicmaps.org/tmql/tmql.html>
> 
> In general: The syntax is not valid acc. to the latest CTM draft,
> especially ";" and "." are missing and "iko" was renamed to "ako".
> If you need a reviewer who is a bit familiar with CTM, ask me.

I definitely need help here. Maybe after the Leipzig meetings, we can
spend a few minutes on fixing this?

> I find it a bit suboptimal that the normative environment is a big
> topic map,

I do not think that the fact, that it is a topic map is an issue.  In
fact, having data types defined ontologically and connected to
functions is nice, IMHO.

But more to your point .....

> ......... I'd wish that at least some functions are defined more
> formally.

... regardless what format is used, what is an appropriate formalism
for defining the functions? What I do not want is to have another
formalism just for that, so what remains is

  - reference some functionality elsewhere (XPath?, suggestions?)

  - describe in prose, add some examples

All can still be held in a topic map. Eating your own dog food is very
educational.

> In my opinion at least the functions "has-datatype" and "has-variant"
> are misleading. The "has-"-prefix indicates (for me) that a boolean
> value is returned which is not the case.

Ok, that would not be an association I had. Haven't used BOOLEANS in
decades.

> I'd prefer "datatype" and "variant".

Fine with me. What about

    'variants'

? Because ...

> ....................................Further, the description of the
> "has-variant" gives the impression, that just one variant from each
> name is returned. Is that the case?

... while the description insinuates that (will be fixed), you definitely
get all the cra^H^H^variants.

\rho