- From: Thomas Baker <tbaker@tbaker.de>
- Date: Tue, 31 Jan 2006 15:23:27 +0100
- To: SW Best Practices <public-swbp-wg@w3.org>
SWBPD VM 2006-01-31 telecon agenda
Tuesday, 15:00 UTC (16:00 Berlin)
http://www.w3.org/Guide/1998/08/teleconference-calendar#D20060131
Zakim: +1-617-761-6200
Conference code 8683# ('VMTF')
irc://irc.w3.org:6665/vmtf
Recent telecons
-- 2006-01-24: http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0125.html
Next telecons (weekly)
-- 2005-02-07 Tue 1500 UTC http://www.w3.org/Guide/1998/08/teleconference-calendar#D20060207
-- 2005-02-14 Tue 1500 UTC http://www.w3.org/Guide/1998/08/teleconference-calendar#D20060214
AGENDA
-- New version is:
http://www.w3.org/2001/sw/BestPractices/VM/http-examples/2006-01-18/
--- $Id: Overview.html,v 1.6 2006/01/27 04:23:11 ajm65 Exp $ ---
2006-01-27: editorial control passed to Tom B.
-- Title is now 'Best Practice Recipes for Publishing RDFS
and OWL Vocabularies/Ontologies'. Alistair argued for having
the words 'OWL' and 'ontologies' in the title somehow,
to acknowledge the community and improve search hits using
those words; he was, however, tempted by the tantalisingly
snappy 'Best Practice Recipes for Publishing RDF'.
-- Alistair cannot attend the call today. He is in agreement
to advance to Working Draft pending TF consensus on the
issue raised at [1]. The problem is that the content
types "text/xml" and "application/xml", which are used
in the rewrite rules, are ambiguous inasmuch they are in
practice used to label either HTML content or RDF content.
Alistair suggests removing both of the content types as
conditions from the rewrite rules. Does the TF agree?
[1] http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0163.html
-- It has been suggested that the 'testing' sections contain
a script of some sort for automatically running the
tests. For testing purposes, Alistair tried unsuccessfully
to get wget to change the default Accept headers. Does anyone
know of an alternative command line HTTP client for use in a
shell script?
-- Alistair has implemented and tested all the new
configurations, using the tests given in the document.
See http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0154.html
-- We are up-to-date on all actions except:
ACTION: Ralph to test recipes with W3C configuration.
ONGOING PROGRESS REPORT...
1. Editor's Draft
"Best Practice Recipes for Publishing RDFS and OWL Vocabularies/Ontologies"
http://www.w3.org/2001/sw/BestPractices/VM/http-examples/2006-01-18/
1.1. 2006-01-24 - Resolved: To ask the SWBP Working Group
at its Feb 6 telecon to approve publication of a Working Draft.
1.2. New version in preparation:
http://www.w3.org/2001/sw/BestPractices/VM/http-examples/2006-01-18/
Alistair working on this for now, making structural changes.
Should not be edited in-place by others until Alistair explicitly
passes the token.
1.3. Cookbook name: "Best Practice Recipes for Serving
RDFS and OWL Vocabularies" was decided on Jan
17. Since then, further discussion of details.
A weak preference for "publishing" (instead of
"serving"); just "RDF vocabularies"; and "short
is better" - but Alistair has some latitude here
to decide. Alistair would like the title to mention
"ontologies".
2. Default response - "the IE6 problem"
RDF as the default response was discussed in December:
http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0022.html
Alistair has since found that conditional redirects work differently
with Mozilla and IE6 because the two browsers send different
Accept headers with an HTTP GET request. Hence if we want
URIs to be clickable in IE, either we
(a) leave the recipes as they are and ask IE to send more headers, or
(b) change the recipes around and ask RDF toolkits to send more
headers.
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0069.html
Changing the default to HTML could break existing
SW apps, which traditionally behave as if RDF were the default.
And we note that servers are not required to respect Accept
headers. We therefore think that the Cookbook's stance should
be to support either HTML or RDF as the default.
Alistair thinks this might be done most elegantly by including
lines supporting HTML as the default in each of the recipes --
but commented out. If cut and paste exactly "as is", each
recipe would default to RDF. The accompanying text would
however explain how to uncomment lines to set the default
to HTML. Commenting and uncommenting the lines in question
would "flip the switch" between a default of HTML and a default
of RDF. Additional notes in the Appendix could suggest ways
to use User Agent to work around the lack of Accept headers
in picking an appropriate response.
ACTION - DONE: Alistair added commented/uncommented
lines to recipes to support either HTML or RDF as default.
Ralph argued for RDF as default response where content
negotiation is configured, because of dependency of current
Semantic Web applications. We found a way around the
problem in IE6 that allows URIs to be clickable even with
RDF as the default. Ralph argued for this 'workaround'
to go into an appendix. Alistair tried doing this, but made
a decision to incorporate it into the recipes, because
that way the vast majority of people will be able to just
copy and paste from one location to another. It didn't
make sense to recommend RDF as the default, and then
direct everyone who wanted IE6 clickability (i.e. almost
everyone) to an appendix that says how each recipe needs
to be modified. See the new section 'content negotiation'
for an explanation.
3. Responses to reviews
On Jan 10, we decided that in January we would discuss
reviewer comments and formulate responses to reviewers
on the list.
3.1. David Booth review
http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0124.html
-- Global suggestions
G1. To discuss trade-offs between hash and slash URIs
Response: Ralph has addressed this with added text in the
introduction. David has not yet indicated whether he is
satisfied.
G2. To avoid purl.org recipes, which violate TAG resolution
with 302 redirect code.
Problem with purl.org: It is not enough to change all 302s to
303s because 302 is appropriate for most URIs. So the purl.org
maintainers would have to implement a feature for users to
specify that some resource is a non-information resource.
This would require changes to the database. Are there any
options to do a double redirection? I.e. if purl returns a
302 redirect, then my own server does a 303. On Jan 17, decided
to clarify with TAG whether inferences are supposed to be made
already on the initial response code.
ACTION (DONE Jan 17): Alistair drafted the question (i.e., that only the
initial response code matters) for discussion in BPD, then to send to TAG:
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0076.html
This draft note to TAG
-- suggests they coin a URI for class "resource"
(tag:informationResource) so that things like rdfs:Class,
owl:Class, and rdf:Property could be declared disjoint with it.
-- requests clarification on what implication one can draw when
303 is returned as opposed to 200 ("X is a tag:infoResource").
(Note: In follow-up, David Booth suggested
a draft "HTTP URI-Identity-Algorithm",
out of scope for the VM TF per se:
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0116.html
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0165.html)
ACTION - DONE: Alistair put the purl.org material into an Appendix.
-- Specific recipes
Recipe 3. Interpretation of a fragment identifier in the
presence of 303 redirects is unclear, so recipe
should note that browser may or may not apply
fragment identifier to secondary URI.
-- Editorial suggestions
E1. Shorter URIs in the examples would be better.
Alistair would rather leave the longer URIs for now because
a UK server is configured to support them, see
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0034.html.
Ralph suggests using w3c URIs in the final version (with
shorter URIs for the examples).
E2. At the beginning of each recipe, say what the URIs would return.
Alistair proposes to illustrate this graphically, so added images
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0034.html.
David Booth actually intended simply to spell out which URIs
are redirected to. Ralph wonders whether the images really add any
new information.
On Jan 18, Alistair reorganized recipes 1 and
2, adding short description of outcomes as
per Booth suggestion. Added examples with
expected outcomes for purpose of testing.
Wants to organize the rest like this when
IE6 bug resolved.
3.2. Andreas Harth review
http://lists.w3.org/Archives/Public/public-swbp-wg/2006Jan/0004.html
-- The document has too many choices - suggests
cutting down to 3 or 4 covering 80% of the cases.
-- Suggests content negotiation instead of mod_rewrite
modules. Response at:
http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0016.html
-- Suggests mod_alias instead of mod_rewrite.
-- Maybe put purl.org examples into an appendix.
4. TAG httpRange-14 decision [10]
There have been recent comments on the list about the TAG
decision by David Booth: [11] summarizing best practice
issues around the decision, [12] on minting http URIs,
and [13] on 302 versus 303 redirects; by David Wood: [15]
with a use case for RDF; and and by Jacco van Ossenbruggen
[14], reviewing the comments by David Wood and David Booth.
In BPD, the idea arose to write a TAG-like "finding"
to explain the impact of the httpRange-14 decision.
David Booth and David Wood wrote drafts independently.
Probably not enough time in BPD to start a new document,
so could they be incorporated into cookbook? David Booth
agreed to go to propose a short section for us to consider.
General agreement that we should avoid getting too
historical -- just cover practical results. Some of
what David Booth has written could plausibly be recast
as introductory material on how to choose reasonable
namespace names.
5. Testing
ACTION: Ralph to test recipes with W3C configuration.
6. xml:base and hash URIs
On Jan 20, Alistair discovered that if you put a hash at the end
of the base URI, e.g.:
xml:base="http://example.com/foo#"
the hash is ignored when constructing relative URIs. Discussion followed.
7. Recipe covering the case of very large vocabularies
2006-01-24. In addition to the current five recipes,
Alistair sees the need for a Recipe Six -- if only
a placeholder with text outlining the problem -- for
covering the scenario of very large vocabularies, where
instead of serving an entire vocabulary in RDF one might
want to serve bounded descriptions of individual terms.
As the Cookbook would not provide a recipe for this case,
it is an open question how this issue should be handled
editorially.
8. Longer-term issue: alignment of content-negotiation ideas
in the cookbook with TAG:
-- http://www.w3.org/2001/tag/issues.html#namespaceDocument-8
-- Associating Resources with Namespaces
Draft TAG Finding 13 December 2005
http://www.w3.org/2001/tag/doc/nsDocuments-2005-12-13/
[10] http://lists.w3.org/Archives/Public/www-tag/2005Jun/0039.html
[11] http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0055.html
[12] http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0056.html
[13] http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0123.html
[14] http://lists.w3.org/Archives/Public/public-swbp-wg/2005Dec/0085.html
[15] http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0010.html
--
Dr. Thomas Baker baker@sub.uni-goettingen.de
SUB - Goettingen State +49-551-39-3883
and University Library +49-30-8109-9027
Papendiek 14, 37073 Göttingen
Received on Tuesday, 31 January 2006 14:20:48 UTC