[sc34wg3] TMCL Requirements

Mary Nishikawa sc34wg3@isotopicmaps.org
Wed, 30 Apr 2003 10:45:50 +0900

Hi Partick,

Glad that you are reviewing this now. This needs to be elaborating on, I agree.

Non requirements are for saying, "Of course, we would never specify that!"
This is stuff that can seem obvious, but sometimes needs to be put in 
writing nevertheless.

>I found the Non-Requirement for TMCL of:
>        4.1.1 When these constraints are applied
>rather odd. I am assuming that I am mis-reading the intent of the text 
>since I would think a selector would choose when a constraint is applied. 
>Or does it mean that the selector chooses the where the constraint will be 
>applied and something else applies the constraint?

Behind the choice of which selectors  or restrictions to use or when to 
use then, is a user.

The language will not tell users of the TMCL when they should  or should 
not constrain one TM contruct or another to suit a particular purpose. We 
are not a data typing/ data constraining police. For example we will not 
specify that the users are required  to use the xml schema built in 
datatype "date"for all instances of birthdates used in their topic maps.

Goverment bodies/corporations may want to impose these kinds of rules, but 
this is not in our scope.

This is what I meant when I wrote it. If adding this helps, I will do so. Thanks.

I think we will be spending most of our time looking at the requirements 
that can be agreed to, and others that we need to raise issues for.  We 
are all leaving for London soon, so please bring your list. I probably 
will not be able to respond after today because of trip preparations.