[xtm-wg] Initial feedback on draft

[Sam Hunting <sam_hunting@yahoo.com>]

Steve/Graham/AG:

It reads well! Good job.

There are substantive (I hope) comments intermixed with copy edits.

The substantive issues are marked with "**" and remarks are put
in square brackets []. Many of these are "consensus" issues.

Note that many of my substantive comments are directed to what I 
regard as nice-sounding but extremely bold, even wild, claims 
about ontologies, semantics, and the like.

I strongly believe that there is no consensus for these claims
in the AG, and so they should be deleted from the spec before
publication.

I'm afraid this is only as 2.3; more later.

S.


status of this document
-----------------------

a. "... is in an Authoring Group Review phase ..."

If so, then should not the title still read "TopicMaps.Org AG 
Review Specification", not "Draft Specification"? 

b. "From that date until February 3rd, 2001" sounds like review
is cut off on the 3rd. This does not reflect dates agreed on
in Paris:

"The editor is instructed to produce an AG review XTM 1.0 
specification based on the Feb 3rd 2001 draft specification.  The 
deadline for this document is Feb 10th.  Ratification by the 
participating members will close on Feb 17th."

The text should reflect the minutes.

Origins
-------

a. "chaired at the date of delivery of this specification"

The minutes read:

"Steve Pepper delegates chair duties to Eric Freese"

Should/need the text reflect this?

Terminology
----------

a. addressable information resource 

An information resource whose identity is computable. (That is, a 
computer system can retrieve the resource and make deterministic 
comparisons between it, and some other resource, to establish 
their identity or difference.) An example of an addressable 
information resource is the online version of the document you 
are now reading.

Replace:

  of the document you are now reading.
  
with

  this document.

b. reification

  Delete
  
    within the terms of the topic map paradigm
    
  Replace
  
    within a topic map.

[Don't know what it means to "discourse
within a paradigm." KISS]

  
c. topic map node

  seems pointless to have nodes, but no graph. Delete?
  
2. Concepts
-----------

a. Concur with Bernard on unusefulness of "real[-| ]world" here

b. Concur with Bernard on *resources*.

c. "The assigment of such characteristics is said to be valid
within a certain scope, or context."

Delete:

  , or context
  
"... at the time of its creation."

Append:

  Topic map authors may provide additional information to facilitate
merging.

d. "gentle introduction"

  Delete throughout (titles and TOC):
  
    A gentle
    
  replace
  
    [A|a]n

2.1 A Gentle Introduction...
--------------------------

a. Well, if we MUST use reifies, let's work it in!

para starting "Because not all subjects"

Delete

  every topic acts as a surrogate for some subject

Replace

  every topic acts as a surrogate for (or *reifies*) some subject
  
b. para starting "It is often useful in"

Delete

  guide the compiler
  
Replace

  guide the author or indexer
  
c. para starting "It is often useful in"

Delete

  representation of relations among subjects
  
Replace

  representation of relationships among subjects
  
[no confusion with relation as in relational DB]

d. para starting "because associations express"

Disagree with Bernard regarding "real-world" here. 

e. para starting "There is no"

Note that if the DTD is changed to allow no topics
as members (IMHO wrongly) the final sentence will
have to be reworked.

f. para starting "Becuase topics"

Delete

  topics and their relations
  
Replace
  
  topics and their relationships
  
g. para starting "We use the word topic map to denote"

delete 

  such objects
  
replace

  such things
  
2.2 Overview of topic map constructs
------------------------------------


a. para starting "this section"

Delete

  "more logical"
  
Replace

  "logical rather than alphabetical"
  
b. para starting "In order to discourse"

Replace:

"In order to discourse on a subject within
a topic map, that subject must be reified
through the creation of a topic. Subjects
are thus the organizing principles of topics."

[Don't know what it means to "discourse
within a paradigm." KISS]

c. para starting "In a consistent topic map"

"each subject will ideally"

This does not conform to the definition in
terminology, which lacks "ideally". Since
the ideal is only achievable with the intervention
of humans or at least intelligent agents (who are
after all the ultimate judges) this is a big
issue. Suggest adding "ideal" to terminology.

"preferably" in same para embodies same issue.

d. para starting "Most subjects"

Delete

  their identity is
  
Replace

  their identities are
  
e. para starting "However anything"

Delete

  However anything
  
Replace

  However, anything
  
Delete

  considered as an HTML document in the file system.
  
Replace

  considered as an HTML document.
  
[internet not equal "the" file system]


2.2.1.2 Reification
-------------------

a. para 1: duplicates terminology, see comment there.

b. para starting "the notion of reification." 

Delete in toto.

[There's nothing concrete here. I don't know what to 
do with "at the very heart", "say anything", '"real"' (does it 
mean processable?), or "come as close as a machine can".]

c. paragraph starting "since anything"

Delete

  themselves, ... assertions.
  
Replace

  themselves.
  
[Without some formal idea of what an "assertion" is,
and whether topic maps do anything more, for example,
than "assert links", which they do via xlink, there's
nothing concrete here.]

2.2.1.3 Subject identity

a. para starting "subject identity" 

** Delete

   -- in effect, to interchange their semantics -- 

[This bold statement can neither be proved nor
disproved. It is, therefore, content-free and can
be deleted without loss.]

2.2.1.4

a. para starting "A subject indicator"

** Delete
  
   and interchanging semantics
   
[This bold statement can neither be proved nor
disproved. It is, therefore, content-free and can
be deleted without loss.]

b. para starting "since subject identity"

** delete

  through the use of standardized ontologies
   expressed as published subject indicators
   
  replace
  
    through published subject indicators"
    
[original statment implies that the ontology
method is the only way to express a PSI. Surely
this is to be proved rather than asserted? Let's
KISS.]

c. para starting "since once and the same"

** Delete
  
    Situations like this ... ontologies.
    
Replace

  For example, given two topics, where one has "Buster Keaton" as 
  a basename topic characteristic, the other has "The Great Stone 
  Face" as a basename topic characteristic, and neither has a 
  subject indicator resource, a subject mattter expert on 
  Hollywood stars might conclude that both topics have the same 
  subject -- the silent comedian named "Buster Keaton" -- and 
  intervene in XTM processing to cause these two topics to be 
  merged.
  
[There is no reason to focus on mediating ontologies
here, or to recommend a particular method. Further, that
the human is the ultimate judge of subjects to be merged
is omitted from the current draft.

2.2.1.5 Topic characteristic

a. para starting "anything that may be"

Delete

  Anything ... Characteristics
  
Replace

  Topic characteristics
  
b. para starting "The assignment ..."

  Delete
  
    scope, or context.
    
  Replace
  
    scope.
    
[What could the extra word for scope add except confusion?]

2.2.16 Scope

a. para starting "the concept of scope"

  Delete
  
    The concept of scope is used to express the extent
    
  Replace
  
    Scope specifies the extent
    
  
  [Implies concept itself somehow actuates the computer!]
  
   Delete
  
    In other words, it
    
  Replace
  
    It
    
  Delete
  
    , whatever the processing context.
    
  Replace
  
    .   
 
2.2.2 Name

a. para starting "Each name may exist"

  Delete
  
    It always
    
  Replace
  
    A name always
    
2.2.2.1 Base name

a. para starting "A base name"

  QUERY: Do we want to specify a length for the string?

  The minimum basename string length that is required to be 
  supported by all conforming XTM applications is 255 characters. 
  This XTM 1.0 Specification does not impose a maximum basename 
  string length limit.

2.2.3 Occurrence

a. para starting "the latter"

** Replace:

  The latter [resourceData] provides (among other things) a useful
  way of assigning metadata to topics
  
With:

  The latter provides a syntactically convenient means of labelling.
  It should never be used for any other purpose, for example,
  metadata, as that would defeat reliable interchange of topic
  maps.
  
[There is no consensus on this statement in the draft at all.]

Putting a whole metadata structure into resource data completely
subverts the entire paradigm.

ResourceData is there only as a syntactic convenience for doing
a label. That's why its only in variant name and occurrence!]

2.2.4 Association

a. para starting "there is no"

  Define "directionality." Is it like hyperlink traversal?
  
2.2.5 Topic Map

a. para starting "a topic map"

** [Still seems strange to have "nodes" without "graph". Where
  else could nodes exist except in a graph?? Even Joe Markup
  knows that -- unless he knows that nodes exist in a DOM tree,
  which is completely deceptive.]
  
b. para starting "1. a serialised"

** delete

    or some other syntax
    

  [There's no consensus that XTM is going to express any other 
  syntax than XML. See the charter:

  "Develop, publish, maintain, and promote a specification, known 
  as XML Topic Maps (XTM), for the expression of topic maps in 
  XML."
  
  Nor, in the absence of a formal processing model, do I see
  a way forward to doing so.]
  
INSERT 2.2.5.2 and thenceforward renumber

2.2.5.2 Topic Map Graph

a. A topic map graph consists of nodes and arcs. The nodes are the
endpoints of the arcs.

** [To me, it is senseless to have nodes that have no plane of
existence, and are
   not connected together in any way. Therefore, I have tried to
provide a vanilla,
   non-controversial, non-application specific definition of a topic
map graph.
   
2.2.5.2 Consistent topic map

a. para starting "A consistent"

  Delete
  
    suppression as defined
    
  Replace
  
    suppression, as defined
    
  [You want Annex F to apply to the whole para, don't you? Not just
"further
  opportunities]
  
2.2.5.3 Topic Map Document

a. para starting "a topic map document"

Delete

  a syntax governed by this or some other specification
  
Replace

  the XML syntax defined by this specification.
  
[There is no consensus on the wording as written.]

2.2.5.4 XTM document

a. para starting "An XTM document"

Delete

  the syntax
  
Replace

  the XML syntax
  
[There is no consensus on the wording as written.]  
  
2.3 Published subjects

a. para starting "A published subject"

  "online via a URI"
  
**QUERY: URN?
  
  QUERY: Does accessible online rule out the use of a file system?
  Say by an intranet or by a developer?
  
 

=====
<!-- "To imagine a language is to imagine a form of life."
     - Ludwig Wittgenstein, Philosophical Investigations -->

__________________________________________________
Get personalized email addresses from Yahoo! Mail - only $35 
a year!  http://personal.mail.yahoo.com/

------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/337252/_/981139434/
---------------------------------------------------------------------_->

To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com