[sc34wg3] SC34/WG3 Draft Minutes - 8 December 2002
Patrick Durusau
sc34wg3@isotopicmaps.org
Mon, 09 Dec 2002 18:44:14 -0500
This is a multi-part message in MIME format.
--------------090907040708010407030306
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Greetings!
Apologies for the message with no attachment yesterday! Yes, we really
did work yesterday and as evidence I offer my failure to attach the
minutes due to fatigue. ;-)
Today's minutes will follow shortly.
Patrick
--
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
pdurusau@emory.edu
--------------090907040708010407030306
Content-Type: text/html; charset=us-ascii;
name="sc34_8Dec2002.html"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
filename="sc34_8Dec2002.html"
<html>
<!--THIS FILE IS GENERATED FROM AN XML MASTER.
DO NOT EDIT-->
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<title>Draft Minutes: SC34/WG3 Meeting, Baltimore, 8 December 2002</title>
<link rel="stylesheet" type="text/css" href="/stylesheets/tei-oucs.css">
<meta name="DC.Title" content="Draft Minutes: SC34/WG3 Meeting, Baltimore, 8 December 2002">
</head>
<body><a name="TOP"></a><table class="header" width="100%">
<tr>
<td align="left">
<h1 class="maintitle">Draft Minutes: SC34/WG3 Meeting, Baltimore, 8 December 2002</h1>
</td>
</tr>
</table>
<hr>
<h2>Contents</h2>
<ul class="toc">
<li class="toc">1. <a class="toc" href="#body.1_div.1">SC34/WG3 Morning Session</a><ul class="toc">
<li class="toc">1.1. <a class="toc" href="#sc34_8Dec2002-div-d0e86">term-scope-def</a></li>
<li class="toc">1.2. <a class="toc" href="#sc34_8Dec2002-div-d0e214">prop-scope-structure</a></li>
<li class="toc">1.3. <a class="toc" href="#sc34_8Dec2002-div-d0e236">err-constraint-violations</a></li>
<li class="toc">1.4. <a class="toc" href="#sc34_8Dec2002-div-d0e264">op-modifications</a></li>
<li class="toc">1.5. <a class="toc" href="#sc34_8Dec2002-div-d0e274">reification-effects</a></li>
<li class="toc">1.6. <a class="toc" href="#sc34_8Dec2002-div-d0e308">term-equal</a></li>
</ul>
</li>
<li class="toc">2. <a class="toc" href="#body.1_div.2">Afternoon Session</a><ul class="toc">
<li class="toc">2.1. <a class="toc" href="#sc34_8Dec2002-div-d0e379">Short Tour of Reference Model</a></li>
<li class="toc">2.2. <a class="toc" href="#sc34_8Dec2002-div-d0e397">term-subject-address-def</a></li>
<li class="toc">2.3. <a class="toc" href="#sc34_8Dec2002-div-d0e422">locator-reference</a></li>
<li class="toc">2.4. <a class="toc" href="#sc34_8Dec2002-div-d0e460">infinite-subject-spaces</a></li>
<li class="toc">2.5. <a class="toc" href="#sc34_8Dec2002-div-d0e509">strings-as-subjects</a></li>
<li class="toc">2.6. <a class="toc" href="#sc34_8Dec2002-div-d0e547">Future plans for SAM</a></li>
</ul>
</li>
</ul>
<div class="teidiv">
<h2><a name="body.1_div.1"></a>1. SC34/WG3 Morning Session
</h2>
<p><b>Attendees</b></p>
<ul>
<li><a name="d0e42"></a>Derek Millar
</li>
<li><a name="d0e45"></a>Mary Nishikawa
</li>
<li><a name="d0e48"></a>Patrick Durusau
</li>
<li><a name="d0e51"></a>Motomu Naito
</li>
<li><a name="d0e54"></a>Lars Marius Garshol
</li>
<li><a name="d0e57"></a>Ann Wrightson
</li>
<li><a name="d0e60"></a>Elliot Kimber
</li>
<li><a name="d0e63"></a>Martin Bryan
</li>
<li><a name="d0e66"></a>Bernard Valant
</li>
<li><a name="d0e69"></a>Steve Newcomb
</li>
<li><a name="d0e72"></a>Holger Rath
</li>
<li><a name="d0e75"></a>Sung_Hyuk Kim
</li>
<li><a name="d0e78"></a>Soon_Bum Lim
</li>
<li><a name="d0e81"></a>Michel
</li>
</ul>
<p>
</p>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e86"></a>1.1. term-scope-def
</h3>
<p><a name="d0e91"></a>The definition of scope is different from that of XTM 1.0 and ISO 13250:2000, in that it explicitly says topic characteristics
assignments are valid for each of the subjects in its scope individually. Is that acceptable?
</p>
<p><a name="d0e94"></a>Lars: All subjects view, only valid where all themes apply, vs. valid where any views apply, vs. say nothing, vs. structured
scope (Ann)
</p>
<p><a name="d0e97"></a>Newcomb: scope is just one of an unbounded number of assertions about assertions, see Steve's email, Scope, again. annoying
to have to define a class instance type to qualify a relationship. scope is an escape valve without defining an asertion type.
</p>
<p><a name="d0e100"></a>Ann: in XML can now only care about part of what is being said (without DTDs). but how to do strict policies where needed?
</p>
<p><a name="d0e103"></a>Newcomb: if scope gets very large and hard to manage, create specialized assertions that take the place of that scope. anything
that might want to say in a scope could have a special assertion type
</p>
<p><a name="d0e106"></a>Lars: does sort of beg the question about why have we encumbered the structure with scope? with the query language will need
to say that queries are executed within a particular scope.
</p>
<p><a name="d0e109"></a>Newcomb: would mean more complexity at query time, should use more specialized assertions if that is a problem.
</p>
<p><a name="d0e112"></a>Bernard: so namespace aspect is no longer a problem?
</p>
<p><a name="d0e115"></a>Newcomb: can make topics instanceOf a class that is to not display (or to display)
</p>
<p><a name="d0e118"></a>Ann: is this difference in topic maps as indexing level concept vs. topic maps as under the hood development tool?
</p>
<p><a name="d0e121"></a>Bernard: Newcomb wants to preserve the user's view.
</p>
<p><a name="d0e124"></a>Lars: declare topic that is not to be displayed and make everything an instanceOf itself.
</p>
<p><a name="d0e127"></a>Martin: Roles, where did they get off to?
</p>
<p><a name="d0e130"></a>Lars: must use published subject
</p>
<p><a name="d0e133"></a>Martin: can distinguish in ISO 13250 by role and scope.
<pre>
<Standard id="RDF">
<Name<Resource Data Framework</Name>
<Name lang="fr">....</Name>
<Used-In project="...">....</Used-In>
Name become a role, Used-In would be the scope
<occurs occrl="Used-In"
</pre>
</p>
<p><a name="d0e139"></a>Newcomb: would be pernicious to tighten up scope.
</p>
<p><a name="d0e142"></a>Lars: tightening here makes the query language easier
</p>
<p><a name="d0e145"></a>Newcomb: three choices, don't say, scopes mean extent of validity, or extent of validity with PSI.
</p>
<p><a name="d0e148"></a>Elliott: applicability
</p>
<p><a name="d0e151"></a>Generally should say extent of applicability and not extent of validity.
</p>
<p><a name="d0e154"></a>Lars: 1. Any Subjects, (XTM, 13250) 2. All Subjects (inconsistent with merging rules) , 3. Leave it open
#1 is dead, choice affects the interchange of topic maps. current merging rules imply "all subjects"
</p>
<p><a name="d0e157"></a>The union of scope is in ISO 13250 and Steve says it is an error.
</p>
<p><a name="d0e160"></a>Martin: certain that merging rules imply "all subjects"?
</p>
<p><a name="d0e163"></a>Newcomb: when merging topicmaps made by small "a" applications, may end up with a scope that cannot be interpreted. does not
want to lose integrity of scopes that have been expressed.
</p>
<p><a name="d0e166"></a>Ann: can't resolve this without resolving the facet war. Some of the examples should be done with facets.
</p>
<p><a name="d0e169"></a>Lars: still need to resolve legitimate use cases
</p>
<p><a name="d0e172"></a>Holger: three levels, lowest level is scope set itself, identity of scope level in RM (not true for SAM, only property of
characteristic), do we care about that in the SAM; next level, merging, scope governed merging, related to general question
of how to interpret scope, leave up to application, in context of SAM, this is a certain application of scope for merging;
can do ALL and use 3 for interpretation
</p>
<p><a name="d0e175"></a>Lars: 1. Merging, 2. Interpretation, 3. Interchange (2 is what does it mean? 3 is what happens when play by differnet rules
for scope) 4. TMCL/TMQL
</p>
<p><a name="d0e178"></a>Bernard: same problems occur in other places, problem occurs when have more than one subject composing the scope
</p>
<p><a name="d0e181"></a>Elliott: if a characteristic has two themes in its scope (Lars: scope does not affect merging of topics, only characteristics.)
</p>
<p><a name="d0e184"></a>Newcomb: do assertions wind up with multiple scopes?
</p>
<p><a name="d0e187"></a>Michel: give RM view of this issue? Easier to understand in these cases.
</p>
<p><a name="d0e190"></a>Newcomb: should we do this now?
</p>
<p><a name="d0e193"></a>Lars: some people have discussed it deeply but few people in this room have gotten the benefits of that discussion. Have a
document on this that people should read before the next meeting.
</p>
<p><a name="d0e196"></a>Ann: to what extent is it necessary to read all the mailing list to prepare for meetings, should try to improve processes.
</p>
<p><a name="d0e199"></a>Lars: could not finish SAM and did not get list of issues out for discussion.
</p>
<p><a name="d0e202"></a>Ann: need talk about 13250 but not limited to that single document. What can we say now about the nature of those consequences
now?
</p>
<p><a name="d0e205"></a>Newcomb: Lars' contributions have been heroic.
</p>
<p><a name="d0e208"></a><b>Proposed Resolution</b> defer to after lunch
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e214"></a>1.2. prop-scope-structure
</h3>
<p><a name="d0e218"></a>Should it be possible for the scopes of topic characteristic assignments to have internal structure?
</p>
<p><a name="d0e221"></a>Ann: does not want the standard to specify this structure, but a way to use scope with structure.
</p>
<p><a name="d0e224"></a>Elliott: experience with Cyc indicates that this should not be standardized. Have the ability so don't need to add anything.
</p>
<p><a name="d0e227"></a>Newcomb: would be delighted to say why it is pernicious to structure scope.
</p>
<p><a name="d0e230"></a><b>Proposed Resolution</b> Should the scope property continue to be a set of topic items? Answer is Yes for SAM.
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e236"></a>1.3. err-constraint-violations
</h3>
<p><a name="d0e240"></a>Is it acceptable that constraint violations may be reported at any time?
</p>
<p><a name="d0e243"></a>Holger: errors appear in processing, should be mentioned when it happens, add a note has to report the error.
</p>
<p><a name="d0e246"></a>Lars: either constraint is violated or not, in a data model.
</p>
<p><a name="d0e249"></a>Michel: should use the reference model to track this sort of error, defer resolution on errors until we see the reference
model.
</p>
<p><a name="d0e252"></a>Newcomb: this is an editorial problem, say pre-condition and post-condition, say what would happen is you use a particular
algorithm and your results must match as though that algorithm was used.
</p>
<p><a name="d0e255"></a>Lars: will re-write to be in declarative style
</p>
<p><a name="d0e258"></a><b>Proposed Resolution</b> All errors must be reported in processing of a topic map and in a note that specifications building in this one will determine
when such errors are reported.
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e264"></a>1.4. op-modifications
</h3>
<p><a name="d0e268"></a><b>Proposed Resolution</b> Should not define operations and the specification of merging should be re-written to a more declarative style. If declarative
formulation is found to be hard to understand, we may keep algorithm in an explanatory note.
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e274"></a>1.5. reification-effects
</h3>
<p><a name="d0e278"></a>If you reify a topic name, does that affect your allowed type?
</p>
<p><a name="d0e281"></a>
<pre>
<topic ***>
<baseName> id-"a">
<instanceOf>#b</instanceOf>
<topic
subject1 (pointing to "a"
<instanceOf
</pre>
</p>
<p><a name="d0e287"></a>Ann: second one should be required to be name of type #b
</p>
<p><a name="d0e290"></a>Lars: could have two types (also an issue)
</p>
<p><a name="d0e293"></a>Ann: types should be fixed or calculated
</p>
<p><a name="d0e296"></a>Bernard: could use in thesarus construction
</p>
<p><a name="d0e299"></a>Lars: about the limit between the subject and topic world
</p>
<p><a name="d0e302"></a><b>Proposed Resolution</b> defer until after RM
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e308"></a>1.6. term-equal
</h3>
<p><a name="d0e312"></a>Should we speak about 'equality' of items or 'equivalence'? Are we comparing identity of items (equivalence) or equality of
their values?
</p>
<p><a name="d0e315"></a>Holger: means identity is the same. value equality in Java means equivalence: equality to be used for basic types, does evaluation
function matter. Lars: ready to go for equal.
</p>
<p><a name="d0e318"></a><b>Proposed Resolution</b> Stick with equality.
</p>
</div>
</div>
<div class="teidiv">
<h2><a name="body.1_div.2"></a>2. Afternoon Session
</h2>
<p><b>Attendees</b></p>
<ul>
<li><a name="d0e335"></a>Derek Millar
</li>
<li><a name="d0e338"></a>Mary Nishikawa
</li>
<li><a name="d0e341"></a>Patrick Durusau
</li>
<li><a name="d0e344"></a>Motomu Naito
</li>
<li><a name="d0e347"></a>Lars Marius Garshol
</li>
<li><a name="d0e350"></a>Ann Wrightson
</li>
<li><a name="d0e353"></a>Elliot Kimber
</li>
<li><a name="d0e356"></a>Martin Bryan
</li>
<li><a name="d0e359"></a>Bernard Valant
</li>
<li><a name="d0e362"></a>Steve Newcomb
</li>
<li><a name="d0e365"></a>Holger Rath
</li>
<li><a name="d0e368"></a>Sung_Hyuk Kim
</li>
<li><a name="d0e371"></a>Soon_Bum Lim
</li>
<li><a name="d0e374"></a>Michel
</li>
</ul>
<p>
</p>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e379"></a>2.1. Short Tour of Reference Model
</h3>
<p><a name="d0e384"></a>Newcomb: subjects are primarily conferred upon nodes, node has a subject by virtue of playing a role in one or more assertions,
a node's situation in the graph, is what gives the node its subject. assertions yield subjects to the nodes that are role
players. how is the subject represented? nodes have properties, which are values? requires to say a type but does not have
a type system. properties get values from their situationess, assertion type, roles played and applications' definion of what
properties should be assigned to those nodes. Two kinds of properties, SIDP, subject identity discrimination properties. Ann:
linguistic structure is primary and formal structure is secondary. What is merging? 1. have a topic map graph, some nodes
inside assertions, others are not: merging process, looks at every node to determine it situation in the graph, graph must
be well-formed, well-formed may have multiple nodes for the same subject, then calculate all the property values for all the
nodes, all of the assertion types have been defined, 2. then look at values, SIDP, when nodes have the same values, if merger
has occurred, then have to go back and start over again (since the situations have changed) Logic of the application drives
the merging process. Whenever a user defines an assertion type, have defined a subtype of the SAM. Users may want to define
merging rules for cases when certain types of assertions are used.
</p>
<p><a name="d0e387"></a>Newcomb: diagram, the node of interest on left, need to know the notation of the subject indicator, Theory of Resource Identity,
Theory of Subject Indication, Addressing Scheme. What is the subject of the subject indicator node, gets its subject from
its situation, because of its role, address <nmsploc...> Ann: but HyTime would not work with some systems, Elliott: HyTime
addresses a unique physical thing in the entire universe, It has an address as well, (URI not sufficient for a binding point).
URI's are not binding points unless the strings match. Punch line: SAM can bootstrap URI into stronger addressing such as
the moral equivalent of a grove. Without enhancing the web somehow, can't do serious information management. Semantic web
is doing work related to subjects.
</p>
<p><a name="d0e390"></a>Ann: RDF is dealing with names.
</p>
<p><a name="d0e393"></a>Subject-Locator
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e397"></a>2.2. term-subject-address-def
</h3>
<p><a name="d0e401"></a>At what level of interpretation does the topic represent the resource? Does it represent that storage location? The stream
of bytes? The stream of bytes interpreted in some particular way? The standard must either leave the details open or clarify
this. Note that it may be impossible to clarity whnt the interpretation is left undefined.
</p>
<p><a name="d0e404"></a>Elliotte: nothing in the use of resourceRef that licenses you to interpret anything about the subject
</p>
<p><a name="d0e407"></a>Bernard: could use a public subject identifier
</p>
<p><a name="d0e410"></a>Lars: meaning in ultimately something you cannot reach
</p>
<p><a name="d0e413"></a>Lars: resourceRef in and of itself does not tell you what it meant. It is the resource and not the URI.
</p>
<p><a name="d0e416"></a><b>Proposed Resolution</b> The precise definition of the subject address reference is unknown. The subject has something to do with the resource but
we don't know what.
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e422"></a>2.3. locator-reference
</h3>
<p><a name="d0e427"></a>Must locators really refer to information resources? Some URN schemes allow resources tht are not information resources to
be addressed. This affects the definitions of "information resource", "locator", as well as the [subject identifiers] and
[subject address] properties.
</p>
<p><a name="d0e430"></a>Newcomb: Use Fielding approach to subject identity.
</p>
<p><a name="d0e433"></a>Elliotte: create a reliable binding point in the topic map, the whole point of subject identity. was to point to another topic
in the topic map.
</p>
<p><a name="d0e436"></a>Lars: Fielding has a URN scheme, tbd:http://................, semantics are identical to normal subjectIndicator
</p>
<p><a name="d0e439"></a>Elliotte: subject indicator vs. subject address, in XTM syntax can be distinguished.
</p>
<p><a name="d0e442"></a>Lars: can have a topic that does not represent an information resource, use URI that is not resolveable in a resourceRef,
if no byte stream, a reportable error.
</p>
<p><a name="d0e445"></a>Bernard: can't impose strict error condition b/c server may just not reply
</p>
<p><a name="d0e448"></a>Sam: some people may want this level validation
</p>
<p><a name="d0e451"></a>Newcomb: agrees with Lars, have to say in the spec, what you mean when you utter a resourceRef,
</p>
<p><a name="d0e454"></a><b>Proposed Resolution</b> Use Fielding position, ineffable identity referenced by the locator.
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e460"></a>2.4. infinite-subject-spaces
</h3>
<p><a name="d0e464"></a>How should values from infinite subject spaces be represented in topic maps?
</p>
<p><a name="d0e467"></a>Lars: logic usually used with set theory, graph theory builds upon it. set theory, member-of, subset-of, implies that sets
are discrete objects, set theory forces you to be discrete, should be continuous subject spaces rather than infinite subject
spaces. Marriology (sp?)
</p>
<p><a name="d0e470"></a>Newcomb: RM can deal with this, topic map is subset of all topics,
</p>
<p><a name="d0e473"></a>Ann: need a holder for ways for others to work this out for particular domains
</p>
<p><a name="d0e476"></a>Lars: create a topic for a train, and creates a topic for a railway, then wants to say with association, this train was used
on this railway, from 1937 to 1972. How to say this?
</p>
<p><a name="d0e479"></a>Newcomb: here have a section of a timeline, a finite coordinate space.
</p>
<p><a name="d0e482"></a>Elliotte: only has to put in the start and end of the range and not everything in between
</p>
<p><a name="d0e485"></a>Bernard: linked to something lacking from the beginning, people don't want to create topics for every date and don't want
dates to be occurrences. does not fit with normal idea of occurrences.
</p>
<p><a name="d0e488"></a>Martin: date is not an occurrence of a topic, put an axis in to point to, property value pairs associate with topics.
</p>
<p><a name="d0e491"></a>Newcomb: can make an assertion about a topic that has a range in which a topic is valid.
</p>
<p><a name="d0e494"></a>Newcomb: occurrences are really the requirement that a relationship exists between a subject and a piece of information
</p>
<p><a name="d0e497"></a>Ann: cannot go there, user requirement that need imprecise dates, Ex. Oct. rather than a particular date, is it before or
after Oct. 15th? is I don't know.
</p>
<p><a name="d0e500"></a>Newcomb: SAM is the substructure on which other ^A_A/A/applications will be built.
</p>
<p><a name="d0e503"></a><b>Proposed Resolution</b> Should leave it alone.
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e509"></a>2.5. strings-as-subjects
</h3>
<p><a name="d0e514"></a>Should it be possible to create topics that represent strings, and for it to be formally clear that these topics do represent
particular strings? If so, how?
</p>
<p><a name="d0e517"></a>Newcomb: 2.1 requires that a string be read as a particular string
</p>
<p><a name="d0e520"></a>Lars: user's don't want to create a topic for 24 degrees Celius, just to list the temperature. don't know what it is the name
of. If use a name type
</p>
<p><a name="d0e523"></a>Newcomb: strings can in particular locations, or strings that have no context,
</p>
<p><a name="d0e526"></a>Bryan: how to decide what meaning "run" has in a particular case
</p>
<p><a name="d0e529"></a>Lars: [tzy="24C"] [lmg="Lars M"] these are identical in terms of syntax
</p>
<p><a name="d0e532"></a>Newcomb: when you have resourceData, does it have context or not?
</p>
<p><a name="d0e535"></a>Ann: two concepts of string, actually mean the equivalence class of strings, name is one of the possible uses,
</p>
<p><a name="d0e538"></a>Newcomb: node demanders, define a syntax for interchange and deserialization procedure, can define things in the syntax that
you want to point at in deserialization, can point at the node or information item that results from deserialization. can
do the same as the baseNameString.
</p>
<p><a name="d0e541"></a><b>Proposed Resolution</b> Defer
</p>
</div>
<div class="teidiv">
<h3><a name="sc34_8Dec2002-div-d0e547"></a>2.6. Future plans for SAM
</h3>
<p><a name="d0e552"></a>Editor will produce a new draft with all these issues included, probably after Christmas. For comment by national bodies,
Should have ready by London in May. London should be TMCL and TMQL. will also produce a list of proposed solutions
</p>
</div>
</div>
<hr>
<div class="footer">Comments should be directed to the topic map list, <a href="mailto:sc34wg3@isotopicmaps.org">sc34wg3@isotopicmaps.org</a>. Corrections should be send to <a href="mailto:pdurusau@emory.edu">Patrick Durusau, pdurusau@emory.edu</a></div>
</body>
</html>
--------------090907040708010407030306--