TMCL - draft 2010-03-25

Hi all,

I have to admit that I didn't follow TMCL in detail, but I've some
comments against the current draft:

Introduction
------------
  [...]
  TMCL is defined as a Topic Maps vocabulary consisting of a number of
  topic, association, occurrence, and role types
  [...]

The usage of plural (role types) vs. singular (topic, association
etc.) is disturbing. Maybe the distribution is caused by my limited
English knowledge, though.


Generic criticism: Why do the editors use "contains" if they are
referring a (singleton) property and not "is"? The common
understanding of "contains" is a collection. Even if one item
represents a collection, it does not match the common, naive
(TMDM-conform) understanding of "contains", imo.


3.1 General
-----------
  [...]
  where the phrase is used, or using a qname (an identifying token of
  the form foo:bar)
  [...]

Shouldn't "qname" be written as "QName"?


3.2 tmdm:subject
----------------
  [...]
  The topic tmdm:subject [...] represents the concept of a subject as
  defined in [ISO/IEC 13250-2]. It is the type of which all topics,
  statements, and association roles are instances, and the common
  supertype of all types.
  [...]


Is it defined in TMDM? TMDM is rather unspecific, imo. The common
statement of TMDM seems to be that topics are subjects. And it does
not define that associations are subjects by default. Associations are
(acc. to TMDM) relationships between subjects.


3.6 Being a Subtype
-------------------
  [...]
  A topic type t1 is said to be an subtype of the topic type t2 if
  following the association type tmdm:supertype-subtype from t1 to
  roles of type tmdm:supertype produces either t2 or a subtype of t2.
  [...]

What does "following [...] to" mean? even if I define

  tmdm:supertype-subtype(tmdm:supertype: foo, wtf: bar)

the definition seems to be hold true (acc. to TMDM). It *does not*
define a "supertype-subtype relationship" acc. to TMDM 7.3, but TMCL
refers just tmdm:subertype-subtype and tmdm:supertype and neither
tmdm:supertype-subtype nor tmdm:supertype is constrained.


3.7 Being an Instance
---------------------
Same problem as in 3.6.


3.8 Matching a Regular Expression
---------------------------------

  [...]
  A string s matches a regular expression r if the string is a member
  of the set of strings L(r) denoted by r as defined in appendix F
  [...]

This is sloppy defined. Appendix "F" defines not "r". It defines "R",
further the text is simply a verbatim copy of the XSD text which does
not explain much.


3.9 The Value of an Occurrence
------------------------------
   [...]
   The value of a topic t's occurrence of type ot is referred to as v,
   and produced by finding the occurrence item o in t's [occurrences]
   property whose [type] property contains ot or some subtype of ot.
   If no such occurrence exists there is no given value. It is an
   error if there is more than one such occurrence.
   [...]

This section explains nothing. What's the 'value'? I'd assume the TMDM
Occurrence [value] property. But the section does not say it.

Why does not TMCL define "The Value of a Name" and "The value of a
Variant"?


... to be continued ... maybe ...

Best regards,
Lars
-- 
Semagia 
<http://www.semagia.com/>

<https://twitter.com/larsheuer/> Twitter
<http://www.topicmaps.de/mailinglist/> German Topic Maps mailing list
<http://tinytim.sourceforge.net/> Open Source Topic Maps engine
<http://mappa.semagia.com/> Mappa - Python Topic Maps engine