/[volute]/trunk/projects/vocabularies/doc/vocabularies-issues.xml
ViewVC logotype

Contents of /trunk/projects/vocabularies/doc/vocabularies-issues.xml

Parent Directory Parent Directory | Revision Log Revision Log


Revision 61 - (show annotations)
Thu Feb 14 17:43:23 2008 UTC (13 years ago) by norman.x.gray
File MIME type: text/xml
File size: 14435 byte(s)
Adjust issues list, following resolution/discussion on the mailing list.
Make matching edits to the document.

1 <html xmlns="http://www.w3.org/1999/xhtml"
2 xmlns:dc="http://purl.org/dc/elements/1.1/"
3 xmlns:dcterms="http://purl.org/dc/terms/"
4 xmlns:grddl="http://www.w3.org/2003/g/data-view#"
5 grddl:transformation="http://ns.inria.fr/grddl/rdfa/2007/05/25/RDFa2RDFXML.xsl"
6 >
7
8 <head>
9 <link href="http://www.ivoa.net/misc/ivoa_note.css"
10 rel="stylesheet"
11 type="text/css"/>
12 <title>Issues list: Vocabularies in the Virtual Observatory</title>
13 <meta name='DC.subject'
14 content="IVOA, Virtual Observatory, Vocabulary"/>
15 <meta name='rcsdate'
16 content="$Date$"/>
17 <style type='text/css'>
18 /* make the ToC a little more compact, and without bullets */
19 div.toc ul { list-style: none; padding-left: 1em; }
20 div.toc li { padding-top: 0ex; padding-bottom: 0ex; }
21 li { padding-top: 1ex; padding-bottom: 1ex; }
22 td { vertical-align: top; }
23 span.userinput { font-weight: bold; }
24 span.url { font-family: monospace; }
25 span.rfc2119 { color: #800; }
26 q { color: #666; }
27 q:before { content: "“"; }
28 q:after { content: "”"; }
29 .todo { background: #ff7; }
30
31 /* 'link here' text in section headers */
32 *.hlink a {
33 text-decoration: none;
34 color: #fff; /* the page background colour */
35 }
36 *:hover.hlink a {
37 color: #800;
38 }
39 </style>
40 </head>
41
42 <body>
43 <div class="head">
44 <table>
45 <tr><td><a href="http://www.ivoa.net/"
46 ><img alt="IVOA logo"
47 src="http://ivoa.net/icons/ivoa_logo_small.jpg"
48 border="0"/></a></td></tr>
49 </table>
50 </div>
51
52 <h1 property="dc:title">Issues list: Vocabularies in the Virtual Observatory</h1>
53 <h2>IVOA Note</h2>
54
55 <dl>
56 <dt>Working Group</dt>
57 <dd><em><a href="http://www.ivoa.net/twiki/bin/view/IVOA/IvoaSemantics">Semantics</a></em></dd>
58
59 <dt>This version</dt>
60 <dd>Issues as of $Revision$, $Date$</dd>
61
62 <dt>Editors</dt>
63 <dd><a href='http://nxg.me.uk/norman/#norman'
64 property='dc:creator' >Norman Gray</a></dd>
65
66 <!--
67 <dt>Authors</dt>
68 <dd>
69 </dd>
70 -->
71
72 </dl>
73
74 <h2><a id="contents" name="contents">Table of Contents</a></h2>
75 <?toc?>
76 <hr/>
77
78 <div class='section'>
79 <p class='title'>Introduction</p>
80
81 <p>This is the list of major issues for the vocabularies work; this
82 does not include minor issues more concerned with the fine details of
83 maintaining and distributing the vocabularies; such minor issues might
84 be better handled using the
85 <a href='http://code.google.com/p/volute/issues/list' >Volute issues
86 list</a>.</p>
87
88 <p>The list is present here as a record of the points at general
89 issue, and (post standardisation) as a reference pointing to some
90 rationale for the design decisions in the standardised document.</p>
91
92 </div>
93
94 <div class='section' id='issueslist'>
95 <p class='title'>The issues list</p>
96
97 <p>The sections below are intended to be a log of the various options
98 and eventual conclusions, rather than a summary of the arguments. For
99 those, see the linked online discussions.</p>
100
101 <p>Some of the issues below have <q>provisional resolutions</q>.
102 These are believed to capture the sense of the on-list discussion,
103 without being final.</p>
104
105 <div class='section-nonum' id='masterformat-1'>
106 <p class='title'>CLOSED: [masterformat-1] Format of the master vocabulary</p>
107
108 <p><em>[Issue summarised here in some detail since it hasn't had much
109 airing on-list]</em></p>
110
111 <p>The distributed (and normative) SKOS files are generated to a
112 greater or lesser extent. This might consist of a conversion from
113 some completely different format, such as the IAUT files which
114 originate in the easily-parsed native format of the Lexicon
115 application which originally managed them (see Shobbrook and Shobbrook
116 <a href='http://www.mso.anu.edu.au/library/thesaurus/introduction.html'
117 >discussion</a>), or a relatively lightweight processing involving
118 adding missing but mechanially inferrable relations.</p>
119
120 <p>Question: what should be the format of the master files?</p>
121
122 <p>Possible resolution 1: nothing mandated in the document -- the
123 format of the master file should be whatever is most convenient, as
124 long as the generated and distributed files are valid SKOS. [This
125 says: there is no need for the IVOA to specify this, as it's purely
126 private to the vocabulary maintainers]</p>
127
128 <p>Possible resolution 2: SKOS, in Turtle notation, possibly requiring
129 some post-processing to add omitted-but-inferrable relations. This is
130 easy to read and write, and it is simple enough that it would be
131 feasible to create from scratch a parser for the relevant subset of
132 it, if that were somehow necessary. [This says: what we're
133 distributing -- SKOS -- might as well be the format we edit, so we
134 mandate that, for the sake of simplicity]</p>
135
136 <p>Possible resolution 3: some more fundamental no-punctuation format,
137 such as that for the Lexicon program. [This says: we want to be
138 <em>completely</em> technology-agnostic, and even SKOS is too hard to
139 parse, post-apocalypse]</p>
140
141 <p><strong>Provisional resolution: option (1) above – nothing
142 mandated</strong>. Only the distribution format is to be specified (no
143 objections on the list).</p>
144
145 <p>Discussion (such as it was):
146 <a href='http://www.ivoa.net/forum/semantics/0802/0670.htm' >2008 Feb 4</a>
147 (+ thread).
148 </p>
149 <!--
150 <p>No 2 seems the best balance between manageability and long-term
151 maintainability.</p>
152 -->
153
154 </div>
155
156 <div class='section-nonum' id='distformat-2'>
157 <p class='title'>CLOSED: [distformat-2] Format of the distributed vocabularies</p>
158
159 <p>Question: in which format should vocabularies be distributed?</p>
160
161 <p>Possible resolution 1: the standard simply mandates that they be
162 distributed in at least one well-known RDF format (which means either RDF/XML
163 or Turtle, which is equivalent to N3 for this purpose). This implies that an
164 RDF parser will, realistically, be required in order to process the
165 vocabulary files.</p>
166
167 <p>Possible resolution 2: the standard requires them to be distributed
168 in a format which is parseable as RDF, but which is also regular
169 enough that it's usefully interpretable as ‘normal’ XML.</p>
170
171 <p><strong>Provisional resolution: option (1) above – distribution in
172 any RDF serialisation</strong>. More rationale added to the
173 document.</p>
174
175 <p>Discussion:
176 <a href='http://www.ivoa.net/forum/semantics/0801/0644.htm' >2008 Jan 21</a>,
177 <a href='http://www.ivoa.net/forum/semantics/0801/0647.htm' >28</a> (+ thread),
178 <a href='http://www.ivoa.net/forum/semantics/0802/0674.htm' >2008 Feb 4</a>.
179 </p>
180
181 </div>
182
183 <div class='section-nonum' id='versioning-3'>
184 <p class='title'>[versioning-3] Identifying vocabulary versions</p>
185
186 <p>Question: do vocabulary users refer to a concept URI with an explicit
187 version, or to a constant URI which always refers to the latest version?</p>
188
189 <p>Possible resolution 1: users always refer to the same concept URI,
190 as for example in <code>http://myvocab.org/myvocab#mytoken</code>, and
191 this refers, either by redirection or server-internal URI rewriting,
192 to the latest version of the vocabulary. The Dublin Core metadata set
193 at <span class='url'>http://purl.org/dc/terms/</span> does this <span
194 class='cite'>std:dublincore</span>.</p>
195
196 <p>Possible resolution 2: users refer to a concept URI without a
197 version; this URL returns a vocabularly with a versioned namespace
198 (this can probably be excluded, since it violates the good practice of
199 having a namespace be retrievable at its own URL).</p>
200
201 <p>Possible resolution 3: users will refer to concepts which have a
202 version explicit within the namespace, as for example in
203 <code>http://myvocab.org/myvocab-v1.1#mytoken</code> (the precise
204 location of the version number or date in the URI is arguably a
205 distribution/maintainance detail).</p>
206
207
208
209 <p>References: see <span class='cite'>berrueta08</span>,
210 <span class='cite'>sauermann07</span>.</p>
211
212 <p>Discussion:
213 <a href='http://www.ivoa.net/forum/semantics/0801/0644.htm' >2008 Jan 21</a>,
214 <a href='http://www.ivoa.net/forum/semantics/0801/0648.htm' >28</a>
215 (+ threads),
216 <a href='http://www.ivoa.net/forum/semantics/0801/0660.htm' >31</a>,
217 <a href='http://www.ivoa.net/forum/semantics/0802/0678.htm' >2008 Feb 4</a>
218 </p>
219
220 </div>
221
222 <div class='section-nonum' id='maintenance-4'>
223 <p class='title'>CLOSED: [maintenance-4] Who maintains vocabularies?</p>
224
225 <p>Question: By whom, and by what process, are vocabularies maintained?</p>
226
227 <p>This is a different issue from <span
228 class='xref'>versioning-3</span>, since that is concerned with how the
229 versions are identified, whereas this is concerned with who it is
230 who manages the changes which are necessary as a vocabulary evolves.</p>
231
232 <p>Option 1: the vocabularies in the standardised
233 document are regarded purely as examples, with no normative force and
234 no specified maintenance process.</p>
235
236 <p>Option 2: the document's vocabularies are normative,
237 and the document should define a maintenance process, possibly
238 modelled on the UCD process <span class='cite' >std:ucdmaint</span>.</p>
239
240 <p>Option 3: the document's vocabularies are normative,
241 but not claimed to be more than merely adequate. They will not be
242 developed as part of this standard's evolution, but instead be
243 maintained by other interest groups, either within or outwith the IVOA
244 process.</p>
245
246 <p>Are there minimal standards of curation which conforming
247 vocabularies must abide by? For example, need we require vocabulary
248 maintainers to use the <code>&lt;skos:changeNote&gt;</code>
249 mechanisms, or just rely on their good sense?</p>
250
251 <p><strong>Provisional resolution: Option 3</strong>. The final
252 published standard will include a number of SKOS vocabularies produced
253 as part of this process. These will be usable and citable, and the
254 community will be encouraged to use them, but they will not be
255 maintained after the standard is complete. Instead, the `owners' of
256 the underlying vocabularies (for example the UCD maintenance group)
257 will be encouraged to maintain the SKOS version alongside their other
258 forms. In particular, the IVOA-T vocabulary will be developed and
259 maintained in a parallel standard to this one.</p>
260
261 <p>Discussion:
262 <a href='http://www.ivoa.net/forum/semantics/0801/0665.htm' >2008 Jan 31</a>,
263 <a href='http://www.ivoa.net/forum/semantics/0802/0753.htm' >2008 Feb 14</a>
264 </p>
265
266 </div>
267
268 <div class='section-nonum' id='vocabset-5'>
269 <p class='title'>CLOSED: [vocabset-5] What vocabularies are included in the standard?</p>
270
271 <p>Irrespective of the resolution to issue <span
272 class='xref'>maintenance-4</span>, there will be a set of vocabularies
273 included in the document, either as samples, or as an initial
274 specification. Question: What should this set contain?</p>
275
276 <p>There are six vocabularies which have been associated with the
277 draft standardisation process, namely</p>
278
279 <ul>
280 <li>the A&amp;A journal keyword list, the IVOA AOIM list, and the 1993
281 IAU thesaurus, whose inclusion is as far as I can see completely
282 uncontroversial;</li>
283 <li>an IVOA Thesaurus based on the IAU-93, which may or may not be in
284 this standard depending on whether people would prefer a completely
285 separate process to develop it;</li>
286 <li>a UCD1+ vocabulary (though this deals with a different set of
287 concepts – namely data types – from the other vocabularies and might
288 arguably connect poorly to them); and</li>
289 <li>a SKOS version of the list of constellations, which is very
290 simple, and which might reasonably find a home in this standard on
291 that ground alone.</li>
292 </ul>
293 <p>In addition, there are multiple informal keyword lists associated
294 with the VOEvent arena (see Roy's message and Rob's). These haven't
295 been SKOSified at all, and Rick's excellent suggestion is that these
296 be left as homework for the VOEvent group.</p>
297
298 <p><strong>Provisional resolution: include all five/six</strong>. The
299 A&amp;A, AOIM, UCD1, IAU-93 and constellations vocabularies will be
300 finished and immediately useable (see the resolution on maintenance in
301 <a href='#maintenance-4'>[maintenance-4]</a>). The IVOAT vocabulary
302 will be developed in a parallel process to this vocabularies standard:
303 it will be referred to, and a snapshot of it <em>may</em> be included
304 in the standard, but it will be clearly marked as a work-in-progress.</p>
305
306 <p>Discussion:
307 <a href='http://www.ivoa.net/forum/semantics/0801/0665.htm' >2008 Jan 31</a>,
308 <a href='http://www.ivoa.net/forum/semantics/0802/0682.htm' >2008 Feb 4</a>,
309 <a href='http://www.ivoa.net/forum/voevent/0802/0654.htm' >2008 Feb 4
310 (VOEvent list)</a>,
311 <a href='http://www.ivoa.net/forum/semantics/0802/0711.htm' >7</a>,
312 <a href='http://www.ivoa.net/forum/semantics/0802/0753.htm' >14</a>;
313 <a href='http://www.ivoa.net/twiki/bin/view/IVOA/VocabulariesWorkingArea'
314 >wiki page</a>.
315 </p>
316
317 </div>
318
319 <div class='section-nonum' id='mappings-6'>
320 <p class='title'>[mappings-6] Inclusion of mappings in vocabularies</p>
321
322 <p>Should mappings between vocabularies be in this standard, and if
323 so, how closely bound should they be to the vocabulary itself?</p>
324
325 <p>The early-2008 editors draft for the SKOS standard <span
326 class='cite'>std:skosref</span> included inter-vocabulary mappings,
327 which were hitherto separate from the intra-vocabulary links in the
328 SKOS core. The question of mappings in the SKOS standard is still
329 (early 2008) in flux.</p>
330
331 <p>Question: how do we accomodate this uncertainty in the IVOA
332 Vocabularies standard? And how do we advise mappings to be published?</p>
333
334 <p>Consideration 1: The mappings spec is still in flux, and likely to
335 remain so for some time after the SKOS core document is
336 standardised</p>
337
338 <p>Consideration 2: Norman would hope to see the situation developing
339 where there are multiple third-party mappings between vocabularies,
340 maintained by specific communities, or which describe mappings at
341 different levels of granularity, or which represent significant
342 (publication-worthy?) labour on the part of individuals, adding value
343 to the network of vocabularies.</p>
344
345 <p>Suggested resolution: include mappings as non-normative parts of
346 this standard, published alongside, but separate from, the normative
347 SKOS versions of the vocabularies, and using whatever are the
348 then-current best mapping practices. In this standard, and in the
349 best-practice guidelines we include, we should proscribe inter-vocabulary
350 mappings being published as part of a vocabulary.</p>
351
352 <p>Crucially, vocabularies and the mappings between them are
353 conceptually separate entities, although they will in practice likely be
354 maintained together.</p>
355
356 </div>
357
358 </div>
359
360 <div class='appendices'>
361 <div class='section-nonum' id='bibliography'>
362 <p class='title'>Bibliography</p>
363 <?bibliography rm-refs ?>
364 </div>
365 </div>
366
367 </body>
368 </html>

Properties

Name Value
svn:keywords Date Revision

msdemlei@ari.uni-heidelberg.de
ViewVC Help
Powered by ViewVC 1.1.26