IRC log of vmtf on 2005-09-27

Timestamps are in UTC.

12:55:36 [RRSAgent]

RRSAgent has joined #vmtf

12:55:36 [RRSAgent]

logging to http://www.w3.org/2005/09/27-vmtf-irc

12:55:45 [RalphS]

Meeting: SWBPD VM Task Force

12:56:02 [RalphS]

Agenda: http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0111.html

12:58:39 [RalphS]

Previous? 2005-07-19 http://lists.w3.org/Archives/Public/public-swbp-wg/2005Jul/0064.html

13:01:30 [Zakim]

SW_BPD(VMTF)9:00AM has now started

13:01:37 [Zakim]

+Ralph

13:02:03 [danbri]

danbri has joined #vmtf

13:02:12 [aliman]

aliman has joined #vmtf

13:02:18 [danbri]

hi

13:02:20 [RalphS]

rrsagent, please make record public

13:02:31 [Zakim]

+Tom_Baker

13:02:59 [Zakim]

+??P5

13:03:08 [Zakim]

+Alistair_Miles (was ??P5)

13:03:19 [Zakim]

+Danbri

13:03:20 [berva]

berva has joined #vmtf

13:03:32 [berva]

berva has joined #vmtf

13:04:13 [berva]

hi - coming soon on the phone

13:05:15 [Zakim]

+??P6

13:05:30 [RalphS]

zakim, ??p6 is Bernard

13:05:30 [Zakim]

+Bernard; got it

13:07:15 [tomb]

tomb has joined #vmtf

13:07:27 [RalphS]

zakim, who's on the phone?

13:07:27 [Zakim]

On the phone I see Ralph, Tom_Baker, Alistair_Miles, Danbri, Bernard

13:08:55 [RalphS]

DanBri: my response (question) to Alistair didn't make complete sense as I momentarily forgot that SKOS URIs use '#'

13:09:18 [RalphS]

Ralph: question still made some sense, as it leads to Bernard's questions

13:10:43 [RalphS]

-> http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0116.html Alistair's thoughts on SKOS changes

13:10:49 [RalphS]

Scribe: Alistair

13:10:52 [RalphS]

Chair: Tom

13:11:02 [aliman_scribe]

tom: if you click on http://purl.../title you get angle brackets ...

13:11:11 [aliman_scribe]

agreed in madrid this isn't good enough ...

13:11:42 [aliman_scribe]

so far assume this is reasonable practice ...

13:12:02 [aliman_scribe]

options: some sort of content-negotiation...

13:13:27 [aliman_scribe]

or resolve to a web page which contains pointer to RDF schema (e.g. via link attribute, or GRDDL) ...

13:13:54 [RalphS]

q+ to talk about deployed tools

13:14:02 [aliman_scribe]

so by default, go to readable web page, and applications that want to consume RDF can do so. ...

13:14:08 [aliman_scribe]

q+ to comment on SKOS proposal

13:14:34 [aliman_scribe]

might be technically difficult, but practically best. But not clear on best way forward.

13:14:43 [Zakim]

RalphS, you wanted to talk about deployed tools

13:15:17 [aliman_scribe]

ralph: important to use solution that respects compatibility with current tools, which expect to do a GET on prop URIs and expect RDF/XML content. ...

13:15:34 [aliman_scribe]

so if we require they do accept: application/rdf+xml prob ok

13:15:36 [Zakim]

aliman_scribe, you wanted to comment on SKOS proposal

13:16:00 [RalphS]

Alistair: the SKOS Core proposal I wrote does try to do both of these things;

13:16:10 [RalphS]

... requesting text/html returns a human-readable document

13:16:18 [RalphS]

... this is different from what DCMI currently does

13:16:43 [RalphS]

... if the clients asks for RDF/XML it gets redirected to something that is RDF/XML

13:17:04 [RalphS]

... this feature of DCMI -- using redirects -- is a good idea

13:17:23 [RalphS]

... so I'm in favor of using content negotiation

13:17:42 [aliman_scribe]

tom: content negotiation seems clear, other option fuzzy.

13:17:44 [danbri]

q+ to ask re "RedirectTemp /foaf/0.1/Organization http://xmlns.com/foaf/0.1/" in apache

13:18:44 [aliman_scribe]

tom: other option? to have web page with embedded pointer to rdf schema, & if enough peple do it then applications will have to handle it.

13:19:22 [RalphS]

q+ to ask if embedded RDF was talked about in Madrid

13:19:45 [aliman_scribe]

tom: ericm said in madrid more practical to have some sort of redirect ?

13:20:03 [aliman_scribe]

danbri: remember ericm saying something about lnking to RDF

13:20:25 [aliman_scribe]

ralph: published recommendation says something about link.

13:20:37 [aliman_scribe]

danbri: webarch provides for content negotiation

13:20:50 [aliman_scribe]

tom: consensus for content-negotiation?

13:21:04 [aliman_scribe]

ralph: other option is embedded RDF (XHMTL 2.0)

13:21:38 [aliman_scribe]

danbri: previous foaf specs have embedded RDF/xXML at expense of validation

13:22:08 [aliman_scribe]

ralph: support as much RDF as we can as embedded, but not necessarily evrything. ...

13:22:33 [aliman_scribe]

RDF/A syntax is what HTML TF is converging on, is it reasonable to express RDF schema as RDF/A?

13:23:17 [aliman_scribe]

ralph: RDF/A solution does not address deployed tools

13:23:30 [aliman_scribe]

tom: content negotiation raise issues?

13:23:45 [aliman_scribe]

ralph: yes but cheap and tools should change to do this anyway.

13:24:01 [RalphS]

ack me

13:24:01 [Zakim]

RalphS, you wanted to ask if embedded RDF was talked about in Madrid

13:24:06 [RalphS]

ack danb

13:24:06 [Zakim]

danbri, you wanted to ask re "RedirectTemp /foaf/0.1/Organization http://xmlns.com/foaf/0.1/" in apache

13:24:21 [aliman_scribe]

danbri: trying to figure out what do do with FOAF ns ...

13:24:32 [aliman_scribe]

currently wrong redirect 30x ...

13:24:43 [aliman_scribe]

currently goes from term URI to namespace URI ...

13:24:52 [aliman_scribe]

can then get that with html or rdf ...

13:25:08 [RalphS]

Ralph: [re Accept: application/rdf+xml] -- it's a shame it took us several years to tell application developers what content type to request, but oh well

13:25:40 [aliman_scribe]

discussed doing redirect so that e.g. ns/mbox redirects to spec#term_mbox

13:26:01 [aliman_scribe]

... but problem with hash & encoding in apache with rewriting ...

13:26:26 [aliman_scribe]

... reasonable for clients to get redirects dependedt on content type requested ...

13:26:49 [aliman_scribe]

but with above options redirects are content type specific?

13:27:05 [aliman_scribe]

...how do you configure in apache?

13:27:10 [danbri]

current:RedirectTemp /foaf/0.1/Organization http://xmlns.com/foaf/0.1/

13:27:33 [danbri]

possible: RedirectTemp /foaf/0.1/Organization http://xmlns.com/foaf/0.1/#term_Organization

13:27:45 [danbri]

prob 1: # gets escaped somehow (reported by al)

13:28:22 [danbri]

prob 2: this becomes representation-specific: #term_Organization is an anchor only in the HTML version of the ns doc ... means nothing in the RDF version

13:28:28 [RalphS]

Ralph: it doesn't strike me that different redirects based on Accept: is architecturally bad, but if there's no apache support that's an issue for us

13:28:57 [RalphS]

Alistair: current browsers preserve their fragid across redirects

13:29:19 [RalphS]

... so a redirect that includes a fragid would lead to fragid clashes

13:29:49 [aliman_scribe]

ralph: suspect alistair is right ... but wonder if its actually endorsed by http spec ...

13:30:01 [aliman_scribe]

if its not endorsed then not all browsers may do it.

13:30:10 [danbri]

ah, so a link such as <a href="http://xmlns.com/foaf/0.1/Organization#term_Organization">Organization</a> might redirect to right part of the html doc

13:30:15 [danbri]

rrsagent, pointer?

13:30:15 [RRSAgent]

See http://www.w3.org/2005/09/27-vmtf-irc#T13-30-15

13:30:32 [RalphS]

Ralph: perhaps we could ask the TAG if it is legitimate for a server to return a fragid in a redirect

13:30:47 [aliman_scribe]

ralph: ask the tag if legit for server to return uri with hash in redirect?

13:31:41 [aliman_scribe]

tom: consensus on content-negotiation, so we could as a next step write up how this works, leading to issues with the TAG?

13:32:16 [aliman_scribe]

tom: ... talking about a workable solution, which raises questions of compatibility with spec.

13:33:03 [RalphS]

Alistair: apache URI-encodes '#' before serving a redirect to the client but purl.org leaves '#' un-encoded

13:33:19 [danbri]

q+ to note that redirect+conneg option, on a / namespace, means that term URIs can't redir to parts of the html spec

13:33:35 [aliman_scribe]

tom: what's the nnext step?

13:33:46 [aliman_scribe]

q+ to talk about requirements

13:33:50 [danbri]

eg <a href="http://xmlns.com/foaf/0.1/Organization#term_Organization">Organization</a> took me to http://xmlns.com/foaf/0.1/#term_Organization

13:33:58 [aliman_scribe]

tom: or is it too speculative?

13:34:02 [RalphS]

q+ to ask for documentation of FOAF URI resolution process ala Alistair's SKOS Core resolution message

13:34:26 [aliman_scribe]

tom: who could take the lead?

13:34:38 [Zakim]

danbri, you wanted to note that redirect+conneg option, on a / namespace, means that term URIs can't redir to parts of the html spec

13:35:32 [aliman_scribe]

danbri: realising that / ns prevents us getting clickable term URIs getting to correct sections of a doc ...

13:36:18 [aliman_scribe]

tested redirects with hash for FOAF ...

13:36:34 [danbri]

http://xmlns.com/foaf/0.1/Organization#xyz

13:36:46 [aliman_scribe]

... so apache hash encoding makes sense if browsers hang on to frag ids across redirects.

13:37:07 [aliman_scribe]

ralph: above redirect is broken...

13:38:07 [aliman_scribe]

danbri: can document this as one of the tradeoffs for hash vs slash.

13:38:54 [aliman_scribe]

ralph: is there a doc which describes FOAF URI resolution policy now, and what it should be post madrid discussion?

13:39:02 [aliman_scribe]

ralph: can you write this up?

13:39:26 [aliman_scribe]

ralph:... propose you write how it is now, then write best guess on how it should work.

13:39:31 [danbri]

http://rdfweb.org/viewcvs/viewcvs.cgi/xmlns.com/htdocs/foaf/0.1/.htaccess

13:39:35 [Zakim]

RalphS, you wanted to ask for documentation of FOAF URI resolution process ala Alistair's SKOS Core resolution message

13:39:51 [aliman_scribe]

ACTION: danbri to write up how foaf URI dereferencing works now, and how it should be done.

13:40:28 [aliman_scribe]

ACTION: tom to write up current DCMI URI dereferencing works now, and how it should be done.

13:40:34 [Zakim]

aliman_scribe, you wanted to talk about requirements

13:41:13 [RalphS]

Alistair: we can state our [DCMI, SKOS, FOAF] requirements, perhaps more simply than we can describe what currently is implemented

13:43:04 [RalphS]

Ralph: it's fine to use SHOULD and MUST in a requirements document

13:43:13 [aliman_scribe]

tom: e.g. marc relator codes are declared as RDF props, and click gets you to html, but you have to know to go elsewhere for the RDF ...

13:43:28 [aliman_scribe]

nice for humans, bad for computers.

13:43:36 [aliman_scribe]

ralph: ok to say must and should

13:44:21 [aliman_scribe]

ralph: bernard asked provocative question in email...

13:44:30 [aliman_scribe]

bernard: not deliberately provocative :)

13:44:44 [RalphS]

-> http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0119.html Bernard asks about relationship to Published Subjects

13:45:36 [aliman_scribe]

ralph: if we go down this route of making URIs friendly to both humans and computers they start to behave alot like publish subjects ...

13:45:46 [aliman_scribe]

can we characterise the similarities and differences?

13:46:30 [danbri]

q+ to suggest "Namespaces whose URI ends in '/', since term URIs cannot redirect HTML clients to sections of the HTML documentation, SHOULD provide 'table of contents' per-term hyperlinks near the top of the HTML documentation"

13:46:31 [aliman_scribe]

bernard: when we did published subjects, recommendation (might) way that computers would use annexed schemas was not clear ...

13:46:39 [aliman_scribe]

pubsub primarily directed at humans ...

13:46:53 [aliman_scribe]

now with this new state of things we can revisit pubsub ...

13:47:23 [aliman_scribe]

ralph: useful line of investigation ... gives us way to do convergence ...

13:47:30 [danbri]

q-

13:47:49 [aliman_scribe]

but one issue is if there is a disagreement between interpretation of human-readabel and machine-readable, which do yu belive?

13:48:22 [aliman_scribe]

bernard: TMs have no notion of a contradiction, no notion of consistency, just some binding points for bits of information about a subject ...

13:48:33 [aliman_scribe]

but no way to make sure these things are consistent. ...

13:48:46 [aliman_scribe]

which is why AI folks don#t like TMs ...

13:49:03 [aliman_scribe]

but consistency between human and computer descriptions is difficult to control.

13:49:17 [aliman_scribe]

tom: clarification: you're not talking about simple workflow issues ...

13:49:34 [aliman_scribe]

because e.g. DCMI generates documentation from schemas (??) ...

13:49:49 [aliman_scribe]

(scribe may have got that wrong way round)

13:50:03 [Zakim]

This conference is scheduled to end in 10 minutes; all ports must be freed

13:50:07 [aliman_scribe]

... DCMI has not said e.g. 'RDF schema has precendence over the web page ...

13:50:25 [aliman_scribe]

' ... so be defult the web page is the 'authoritative' resource ..

13:50:44 [aliman_scribe]

but this situation could change in the future, esp as we move towards saying more about the terms in a formal sense ...

13:50:51 [RalphS]

q+ to ask Tom to elaborate on "by default, the [HTML] page is authoritative over the RDF schema"

13:51:08 [aliman_scribe]

so we're going to have to be more explicit on this (i.e. which is more authoritative).

13:51:27 [aliman_scribe]

bernard: machine readable is normative, human readable is informative ... go in this direction?

13:52:01 [aliman_scribe]

tom: this is direction for dcmi, but RDF schema presents a subset of the information that's presented in the web pages ...

13:52:14 [aliman_scribe]

e.g. versioning information is only in the web pages ...

13:52:20 [aliman_scribe]

q+ to talk about skos

13:52:42 [aliman_scribe]

ralph: is is a proper subset? are there e.g. domains & ranges in RDF overlooked in HTML?

13:52:42 [danbri]

q+ to note http://www.w3.org/TR/rdf-schema/#ch_introduction

13:52:52 [danbri]

"This specification does not attempt to enumerate all the possible forms of vocabulary description that are useful for representing the meaning of RDF classes and properties. Instead, the RDF vocabulary description strategy is to acknowledge that there are many techniques through which the meaning of classes and properties can be described. ..."

13:53:09 [danbri]

rdfs is (on purpose) agnostic re which forms of vocab desc take precedence

13:53:10 [aliman_scribe]

tom: no, actually right now looking at assigning domains and ranges to DCMI props, to express current semantics ...

13:53:33 [aliman_scribe]

seen as a positive step, but danger of restricting definitions in practice ...

13:53:55 [danbri]

q-

13:54:11 [aliman_scribe]

also need to rethink policy implications, how do we express that same information in web documents, and when do we want to say which resource is 'authoritative'?

13:54:32 [aliman_scribe]

... currently assume they're always in sync, which could be difficult.

13:55:04 [Zakim]

This conference is scheduled to end in 5 minutes; all ports must be freed

13:55:04 [aliman_scribe]

danbri: could say e.g. 'we believe these things are always in sync, if not tell us'

13:55:38 [aliman_scribe]

ralph: you (tom) said that by default human readable web page is 'authoritative' .. expand?

13:56:06 [aliman_scribe]

tom: we have web pages which say 'this is the authoirtative description of DCMI terms' and RDF docs don't say that.

13:56:31 [aliman_scribe]

tom: also historically always precendence for html

13:56:31 [Zakim]

RalphS, you wanted to ask Tom to elaborate on "by default, the [HTML] page is authoritative over the RDF schema"

13:56:36 [Zakim]

aliman_scribe, you wanted to talk about skos

13:57:06 [RalphS]

Alistair: this [which variant is authoritative] is a longer-term issue

13:57:27 [RalphS]

... some constraints will be hard to describe; e.g. in SKOS constraints are currently only described in SKOS

13:57:34 [RalphS]

s/in SKOS/in prose/

13:57:47 [RalphS]

... it may be possible to use SPARQL to write some constraints

13:58:00 [RalphS]

... RDFS doesn't have enough to specify all SKOS constraints

13:58:05 [Zakim]

This conference is scheduled to end in 2 minutes; all ports must be freed

13:58:14 [RalphS]

... e.g. broader/narrower

13:58:30 [RalphS]

... and "there should be only one preferred label per language"

13:58:57 [RalphS]

DanBri: RDFS had a class ConstraintResource

13:59:05 [Zakim]

This conference is scheduled to end in 1 minute; all ports must be freed

13:59:59 [RRSAgent]

See http://www.w3.org/2005/09/27-vmtf-irc#T13-59-05

14:00:05 [Zakim]

This conference is scheduled to end now; all ports must be freed immediately

14:00:09 [Zakim]

The time reserved for this conference has been exceeded. 11 ports must be freed

14:00:18 [Zakim]

-Danbri

14:00:30 [RalphS]

rrsagent, please draft minutes\

14:00:30 [RRSAgent]

I'm logging. I don't understand 'please draft minutes\', RalphS. Try /msg RRSAgent help

14:00:32 [RalphS]

rrsagent, please draft minutes

14:00:32 [RRSAgent]

I have made the request to generate http://www.w3.org/2005/09/27-vmtf-minutes.html RalphS

14:00:43 [Zakim]

-Alistair_Miles

14:00:44 [Zakim]

-Bernard

14:01:06 [Zakim]

-Ralph

14:01:07 [Zakim]

-Tom_Baker

14:01:09 [Zakim]

SW_BPD(VMTF)9:00AM has ended

14:01:10 [Zakim]

Attendees were Ralph, Tom_Baker, Alistair_Miles, Danbri, Bernard

14:01:15 [berva]

berva has left #vmtf

14:20:03 [danbri]

danbri has left #vmtf

16:02:19 [Zakim]

Zakim has left #vmtf