See also: Agenda 9-11 Dec, IRC log
<scribe> ACTION: Stuart to query WebApps whether external access to widgets via URIs is ever expected [recorded in http://www.w3.org/2001/tag/2008/12/11-minutes.html#action01]
<trackbot> Created ACTION-206 - Query WebApps whether external access to widgets via URIs is ever expected [on Stuart Williams - due 2008-12-18].
<DanC_lap> 3 Dec cover note for Self-Describing Web draft
NM: [reviews the review
history]
... TBL suggested using the notion of "grounded in the Web"
... I've attempted to include that in this draft
NM: So where do we go with this?
DC: Take some time to read it?
<DanC_lap> struggling to read, in the abstract: "each representation is provided explicitly within the representation"
<DanC_lap> hmm... this seems to start at the wrong end: "Even though this document is of media type application/xhtml+xml [XHTMLMediaType], which is not a member of the RDF family of media types, an RDFa-enabled user agent can extract RDF from this document."
<DanC_lap> but the right story is told in a para below ("So, taken together, these specifications...") so it's not a critical problem.
<DanC_lap> I was reading too fast in the abstract. never mind.
<Stuart> minor editorial, but suggest that "...use of XHTML with RDFa" should be the other way around "...use of RDFa with XHTML"
DC: Starting with RDFa example - HST, are you satisified that NM's new approach is complete wrt the chain ?
<DanC_lap> - 5.1 Using RDFa To Produce Self-describing HTML
HST: That paragraph (at the end of
section 5.1) looks good to me
... I would like to check the xhtml namespace doc't one more time.
.. .
HST has checked the NS document, and it looks OK as it is at the moment
JR: The mutability of that doct is a bit worrying
TBL: The HTML WG works on behalf of the community, and will only make changes which benefit the Web community
JR: But there is no explicit change policy in that document. . .
TBL: That would be a good idea
<DanC_lap> (I think JR asked rather than stating, but ok)
DC: So, we're happy with the new work in 5.1
<jar> I don't think any reasonable person will be misled into thinking the namespace doc is either less stable or more stable than it is.
DC: Moving on to section 6, Grounded in the Web
TBL: I like the basic exposition in
section 6
... There's an error in the Abstract
DC: Push that as editorial
<timbl> Insert "We say the resource is grounded in the Web"
TBL: Before 'web grounding reduces
ambiguity',
... Turning to section 6
... last para in 6.0, first sentence, should say documents not
representations
<timbl> In last para of 6.0
HST: There's a terminological discussion we need to have of wider scope
<DanC_lap> (I think 6.1 Grounding New Specifications in the Web is good)
<timbl> s/Representations/Documents/ in first line
<timbl> s/represnentation documents/documents/ in te middle
TBL: 6.1, bullet 1
... wrt registries, make this "Currently, these are the ones we
know of"
... And that there a costs to the community to maintain these
<DanC_lap> (this seems like a request for emphasis, hence editorial, but tim introduced it as substantive. hmm.)
TBL: and the cost to the developer to getting through the process
HST: Balanced by the benefit to the community of the review which is part of that process
TBL: Use of the word 'core', i.e. the
core specifications of the Web
... needs a subsection
NM: 4.1 is that subsection
TBL: The core is the bootstrap set
NM: Correction, it's in 2, last paragraph
TBL: That's not the point
... When you talk about an engineer, you say that the URI spec is
the only thing they need a priori
... but turning to a machine we need more, maybe
... Things are different depending on whether or not the machine
has GRDDL, or has RDFa, capability
NM: Actual automatic nose-following takes us further than I was planning to go
HST: I think automatic nose-following is out of scope for this document
DC, TBL, NM: yes it is, it's what 5 is about
<DanC_lap> (aha... there's "nose following" and "automatic nose following". ok, the automatic version is something I'm happy to leave out)
HST: OK, we're in the agreement that automatic nose-following, that is, machines finding machine-interpretable information about how to implement e.g. GRDDL, is out of scope
TBL: But it is in scope to say that a (SemWeb) core-aware agent must understand GRDDL
NM, TBL: discussion about two uses of word 'core'
<Zakim> timbl, you wanted to say that the sepc should point out at the end of the RDFa section "and so the data in te document is uncovered automatically by any client for whom RDFa is in
TBL: Last para of 5.1
... This shows that the specs are all tied together adequately for
humans
... Now consider what a machine needs to tie this all
together
... I don't see a comparable crisp statement for machines to what
you have in 5.1 for people
NM: 5.2 does this
TBL: I think it needs to be more explicit about exactly what's involved: HTTP, HTML, XML, ...
NM: So I missed another level of abstraction: packages of core software/suites which given communities buy into
NM: I.e. there's a community on the web which agrees a set of software that they will use
HST: I like this story because there's a non-semweb parallel, i.e. xml-stylesheet, XML, XSLT, HTML -- specs on the one hand and software on the other
<DanC_lap> (3 ppl just agreed on something...)
NM: OK, I can do that
TBL: Crisper
NM: What terminology?
TBL: 'Client technology', maybe
HST: Crisper means we have in 5.1 a
short characterisation of the registry story, and a concluding
demonstration of how this works for the RDFa case for
for people
... Doing the same for what we're talking about is a brief intro to
the idea of Client Technologies which match the spec. sets, tied
together the way the registries say they should be, and for RDFa a
specific example of what the specific CTs are
TBL: We have the basic stuff written
'by sweaty hands': uri, xml, rdb/xml, http, GRDDL
... then above there it's very different, then things can
happen automatically
... The line above the must-be-built-in bit is really
important
... So e.g. an RDFa/XSLT stylesheet is not below the line, and the
triples it builds likewise
... There are two stories that could be told here
... What do we say to people that they should implement: GRDDL or
RDFa as such?
[NM, TBL discussion, scribe didn't follow]
<Zakim> jar, you wanted to talk about status of 'semantic web' in this doc
JAR: So the messaging of this doc't
is important, and there's a lot I agree about here
... I feel like referring to the semantic web has the effect of artificially balkanising
... Look just before 5.1
<DanC_lap> "Because its model is uniform, because all of its self-description is provided in the same model as the data itself, and because all RDF information is linked into the Web as a whole, RDF provides uniquely powerful facilities for dynamic integration of a self-describing Web."
JAR: That sentence should be the
topic sentence for section 5
... RDF is a language, it's a way to communicate -- it has a
special role, alongside natural language, because we're taking
steps to put RDF in a similar place
... This needs to be made more forcefully--not to change the
technical content, but to reframe it
... The words 'semantic web' shouldn't appear -- this is just the
Web, getting richer
NM: Well, I got asked to add it
explicitly
... Indeed it's not just a bump on the Web, it can talk about more
than its own business
TBL: So I agree with a lot, but why not use the words "Sem Web"
JAR: Because it encourages the view that the SW is a ghetto -- call it RDF and the Self-Describing Web
NM: But isn't this standard W3C terminology/perspective?
JAR: Yes, that's my point, we should be taking a broader/more careful perspective
NM: Well, there's a tension here
JAR: Well, even if there were a
Semantic Web, it wouldn't belong here, this isn't what this
document is about
... We're talking about RDF in the service of the self-describing
Web -- or is it just a use case?
HST: I think so
JAR: Why this one?
HST: He was asked to
TBL: What do you mean 'if' there was a Semantic Web
JAR: A long story
NM: Just as the Web is part of the
Internet, the SemWeb is part of the Web
... buying into these specs grows the Web in a particular way
DC: Break to the top of the hour
NM: Are we coming back to this?
DC: I think so, talk to the chair
<jar> Just proposing removing the word 'semantic web' from the doc, so that RDF doesn't become balkanized. Not saying anything against it - just that RDF is stronger without the implication that the semweb is separate from the web
NM: HST, please help me to ensure that the record shows what I've been asked to do
DC: I read 6.1 because it's the part
I care most about
... I think it says the right thing
... There are two main bullet points: Use standards, and ground new
specs in the Web
... I hoped that we could just ship this, because those two points
are now well enough made
NM: are you still looking to ship, or has the discussion changed that
DC: Still looking to ship
DC: I asked TBL if he has ever drafted his own answer to his request, and he said no over the break
HST: Six points, with some overlap:
HT: Six points are pasted above. Phrase "resource representation" is problematic, Tim said representation -> doc. You should either crossref Web Arch and say terms are used as in Web Arch or use less formal terminology -- either content or document would do just fine. You should not mix the two.
TBL: I agree with Henry.
<Stuart> I often finds document ambiguous as it used to refer to resource (stateful) like things and to representation (message) like things.
<Stuart> I would prefer document for resource like things and content for representation like things.
HT: I was slightly surprised by the ATOM example. It never told me how I knew that this was ATOM at all, or that the rel attribute is a URI.
NM: Would a fix just mentioning served with media type application/whatever-atom-uses do it?
HT: Yes.
... Next example was confusing because you said in 4.2.2 things
about hcard, then vcard in class attr is confusing.
NM: I will change to say hcard spec says use class vcard.
HT: That does it, thank you.
... I'm sympathetic at the high level to saying, at a high level,
in addition to "this is what a human needs", "this is what a
machine needs". E.g. Client technology packages.
... That goes a bit less far than Tim went with the above/below the
line stuff. Modulo the feeling it would be better to include that,
I think this does what it needs to do.
... With respect to major question raised by Jonathan, I guess I'm
content that this document does adopt the same relationships on
semweb terminology that other docs do. I think surgery not worth
doing.
NM: Do you think doing the client technology package stuff has net cost/benefit or just benefit.
HT: Cost/benefit. It should be done.
JAR: I've tried to write something
which gets what I want to say carefully:
... There are order 20 occurrences of "Semantic Web" in the
document. I propose to wordsmith these out of existence, failing
which a definition should be provided.
... I think the document would not only read better but do its job
better if the wordsmithing were done
... Consider the last para of the introduction "how to publish
self-describing Semantic Web data"
TBL: What about 'linked data'?
JAR: Yes, perhaps
HST: Capitalised?
[no reply]
JAR: I think owl:sameAs is not a good example, I'll propose an alternative
NM: I would be glad for help with alternative examples
JAR: My concern about this is that
its use of owl:sameAs is incompatible with Owl-DL
... Section 6, para beginning "In fact according", last sentence: I
disagree with the last sentence -- a) it would be up to a court; b)
I don't think the specifications get us there.
... There are other explanations for how this message might have
come to be sent
... Which the publisher could trot out
HST: I'm sympathetic, but I'd like to understand why
TBL: There's an uninteresting set of
objections (e.g. the spec. prose is ambiguous), but there are more
interesting ones
... In particular, that the publisher has asserted the claim
NM: Something such as "a page with this assertion has been published on the Web"
JAR: A page with the information that the senator is. . .
<jar> "have indeed said the senator is"
<jar> "have indeed delivered the message that the senator is"
<noah> how about: a document has been served including the statement "the senator..."
HST: That's too weak
JAR: It's true, but it's too legalistic
<timbl> "Whatever the other legal arguments about the case, the interpretation of the document cannot be disputed"
<timbl> "Whatever the other legal arguments about the case, the interpretation of the document is well defined"
DC: OK, we aren't closing on
resolving this.
... Ashok?
... Jonathan, what's your net?
JR: The things I'd have trouble signing up to here are fairly small. We could add things, but probably better to get it out.
AM: In section 5.1 that starts "With luck,..." What's luck got to do with it? I think you want to say to "here's what you do"
NM: But, however we say it, what you get back may not helpful.
<DanC_lap> ("With luck," appeals to me)
AM: I think you should spell it out.
DC: Not too sympathetic to the comment.
<jar> I may not have been clear earlier... my problem with the "according to the pertinent specifications" sentence is one of attribution and context - is it the publisher who is saying this, or someone else? Is this an old archived piece of writing, or something they still believe now?
NM: Are we going to resolve?
DC: Maybe later, not now.
AM: Regarding the example in 6. We have spoken about it. I like Tim's sentence starting "Whatever the other legal arguments about the case, ..."
TBL: I am sympathetic to getting rid of the "with luck". Communities will tend to converge and not much luck will be needed. I don't like the style of it.
<Stuart> Noah: FYI just noticed that you are treating foaf as a '#'d namespace, AFAIK it is actually a '/' namespace.
<Stuart> ie. http://xmlns.com/foaf/0.1/#name should be http://xmlns.com/foaf/0.1/name (likewise #mbox etc.)
TBL: "Because FOAF is widely understood...."
NM: I as editor can accept instruction do try and draft something to replace "with luck"
<jar> Attention Noah and Tim: I'm happy with Tim's suggested wording "Whatever the other legal arguments ..."
TBL: Changes should get smaller, and
happy to ship it rather than get stuck
... I don't like talking about "reducing ambiguity"
... instead, say something positive
... e.g. "unifying ... of ..specifications, recursively ... on the
Web"
... remove the "by whom"
... Like the Good Practice note
... Enumerate the registries we knew about, rather that inventing
new ones
... Intro could be a bit sharper, by removing point 1 and the
previous sentence, so point 2 begins the document
NM: Happy to make these changes if others are happy
TBL: "in an important sense" should be removed
DC: How about passing the redlined
document to NM?
... Hmm, too difficult
TBL: Do these and ship it
... Including [the whiteboard-based Client Tech stuff]
NM: That last is real work
... As an individual member (i.e. not editor), I'm concerned that we evidently don't know how to
wrap these things up -- I'm not happy with the work practices
around these things. There's an opportunity cost to continuously
making these things better and better.
... I was convinced that the web grounding change, which was big,
was worth doing.
... I feel less clear that that's the case with the bigger stuff
we've heard today
SKW: I reviewed the draft before KC,
and at that time I thought it could be published
... And I still think that's true
<Zakim> Stuart, you wanted to ask JAR a question and to ask JAR which step in the chain of logic that he is not able to make.
SKW: Coming back to the senator example -- where does the chain break down?
JAR: I put it in IRC. . .
... my problem with the "according to the pertinent specifications"
sentence is one of attribution and context - is it the publisher
who is saying this, or someone else? Is this an old archived piece
of writing, or something they still believe now?
HST: I agree -- there's a tiny bit of case law, but the specs don't say as between the person who wrote the document, or whoever manages the server, or whoever owns the domain name, who is the legally responsible person
<noah> "Whatever the other legal arguments about the case, the interpretation of the document is well defined"
DC: So NM's quote above is from TBL, and we can live with that?
<DanC_lap> everybody seems to like that one.
NM: It is pretty limp
SKW: FOAF URIs are wrong
... You have treated it as if it's hashed, but it's slashed
NM: I thought I copied it out of their Primer
<DanC_lap> e.g. http://xmlns.com/foaf/0.1/#name
<DanC_lap> no #
<DanC_lap> there should be no #
<DanC_lap> the FOAF community uses a URI with no #
TBL: Could the two examples (RDF/XML and RDFa) be the same?
NM: they used to be closer
SKW: The use of the word 'document' in place of 'representation' was suggested, I'm not so happy with that, document to me is like resource, use 'content' for a friendly word for representation
<jar> Small issue: Examples in 5 and 5.1 differ along two axes (1) what triples (2) RDF/XML vs. RDFa. Might be better if they had same triples and differed only in form. (recording conversation between Tim and Noah)
<noah> In that conversation, I said I thought there was benefit to aligning the examples, I'm not convinced net cost/benefit at this point. I'd probably leave it.
DC: OK, so, all this input
<DanC_lap> ok, pls everybody type a critical issue or "no critical issues for me"
<DanC_lap> no critical issues for me
<hst> Client Technology package story for machines parallel to Spec chain story for humans
<noah> I would probably make hyperlinks to Web Arch for terms like representation and resource, as suggested by Henry
<timbl> no critical issues if no one else does but I would really like this text about automatic processing included.
<noah> I accept as near editorial the Tim change relating to whether something has been published in Chapter 6
<noah> I accept several of Tim's smaller edit's like "by whom"
<noah> I accept change to get rid of With Luck.
<jar> critical issue for me: resolving use of phrase "semantic web": choices: (a) add a definition, (b) wordsmith to get rid of it, (c) convince me that there's no problem and I'm just an outlier.
<Stuart> 1) attribution of libel; 2) machine oriented follow-your-nose (at least note the discontinutity).
<jar> I am assuming some of things we seem to agree on are going to be taken care of, e.g. the "liar" sentence we just discussed.
<jar> n.b. there are other issues but I don't call them critical... danc only asked for critical
<noah> None
<noah> I don't insist on changing any. I think the Webarch links are worth doing if group agrees no further review needed.
<scribe> ACTION: Jonathan to try to wordsmith to get rid of 'Semantic Web' and submit for review [recorded in http://www.w3.org/2001/tag/2008/12/11-minutes.html#action02]
<trackbot> Created ACTION-207 - Try to wordsmith to get rid of 'Semantic Web' and submit for review [on Jonathan Rees - due 2008-12-18].
<DanC_lap> ACTION: Noah to work on "attribution of libel" concern with Stuart [recorded in http://www.w3.org/2001/tag/2008/12/11-minutes.html#action03]
<trackbot> Created ACTION-208 - Work on \"attribution of libel\" concern with Stuart [on Noah Mendelsohn - due 2008-12-18].
<DanC_lap> ACTION: Noah to attempt to address critical concerns from end of 11 Dec self-describing web discussion [recorded in http://www.w3.org/2001/tag/2008/12/11-minutes.html#action04]
<trackbot> Created ACTION-209 - Attempt to address critical concerns from end of 11 Dec self-describing web discussion [on Noah Mendelsohn - due 2008-12-18].
<DanC_lap> timbl posts http://www.w3.org/mid/DE7694B0-4998-4B24-9D62-7B28E87D27B7@w3.org as candidate to address the so-called Client Technologies/whiteboard red line issue
<timbl> (a) goes at the end of sect 5.1
<timbl> (B) goes at the end of section 5.2
<timbl> (R) goes at the end of 5.0 but BEFORE good practice note
<DanC_lap> action-209: note http://www.w3.org/mid/DE7694B0-4998-4B24-9D62-7B28E87D27B7@w3.org
<trackbot> ACTION-209 Attempt to address critical concerns from end of 11 Dec self-describing web discussion notes added
HST: Tim's posting embodies such a critical concern.
SKW: Thanks Amy van der Hiel for all her help in supporting this meeting
[Suspended until 1330EST]
[Resumed]
NM: I don't want this to drag on indefinitely.
NM: note TAG elections are in progress... that might affect people's thinking about timing
HT: I expect 2 or 3 telcons is plenty
DC: do we have that many before the new TAG members come on?
JR: I could do mine (ACTION-207) by [missed]
(tim's wording?)
(It's alleged tbl has done action-208 re libel; scribe can't confirm)
HT: I might have some availability to
do some editing
... ...
NM: yes, I'd be happy [if somebody came up with a better RDFa example]
HT: how about in the 1st week of the new year, I check with Noah and then pick up the write token...
[NM notes some doc production/xslt stuff; HT acks.]
HT: yes, let's review tim's comments [in email] this thu [18 Dec 2008]
action-209 due 15 Jan 2009
<trackbot> ACTION-209 Attempt to address critical concerns from end of 11 Dec self-describing web discussion due date now 15 Jan 2009
action-208 due 15 Jan 2009
<trackbot> ACTION-208 Work on "attribution of libel" concern with Stuart due date now 15 Jan 2009
<DanC_lap> action-188 due 18 Dec 2008
<trackbot> ACTION-188 Investigate the URL/IRI/Larry Masinter possible resolution of the URL/HTML5 issue. due date now 18 Dec 2008
<DanC_lap> action-176: part of the trail http://www.w3.org/2001/tag/2008/10/tpac
<trackbot> ACTION-176 Work with Dave to draft comments on exi w.r.t. evaluation and efficiency notes added
<DanC_lap> action-176: TPAC discussion notes http://www.w3.org/2008/10/20-exi-minutes.html#item02
<trackbot> ACTION-176 Work with Dave to draft comments on exi w.r.t. evaluation and efficiency notes added
<DanC_lap> action-143 due 15 Jan 2008
<trackbot> ACTION-143 Review Raman's draft of webApplicationState-60 due date now 15 Jan 2008
<jar> http://esw.w3.org/topic/AwwswHome
JAR: group not going smoothly; I
wonder if a charter would help. "editor" role isn't clearly
filled
... let's look at "Work in progress" ...
... AwwswQuestions is a good list, though I haven't looked at it
recently...
... AwwswVocabulary is something I started in the direction of a
validator for this[?] protocol
... there are some themes under "Individual aims"; never mind
what's attributed to me; I've made some peace with issues I saw
before...
... but in tim's, alan's, and stuart's statements, I see a theme:
capture the semantics of http, i.e.
... when someone sends a response, what can you hold them to?
... I see [some theorems] relating data: URIs and http responses...
e.g. about the mime type
... for data: , you can tell that x is not a representation; not
with http:
TBL: why would anybody try to prove x is not a representation?
JAR: e.g. in an audit, as discussed
in individual aims: "If you were hired to audit a company's site to
see whether they were adhering to Web/Semantic Web architecture,
how would you do so?"
... so I'm close to seeing more clearly what problem I'd like the
group to adopt as its charter
[something about person vs server, which JAR relates to logics of delegation by Lampson et. al.; DanC says it sounds like BAN logic]
(re "no charter", isn't there at least a paper trail back to the TAG meeting whence the taskforce came?)
(found the genesis of the AWWSW mailing list/tf http://www.w3.org/2001/tag/2007/11/05-afternoon-minutes )
JAR: there's some energy for exploring problems around "how to bind names" but I don't think that's where the group should head
action-201?
<trackbot> ACTION-201 -- Jonathan Rees to report on status of AWWSW discussions -- due 2008-12-11 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/201
<timbl> I post http://pastebin.com/f5bef462c as RDFa of the RDF/XML example, crowdsourced to SWIG and supplied by bengee
action-201?
<trackbot> ACTION-201 -- Jonathan Rees to report on status of AWWSW discussions -- due 2009-03-30 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/201
DanC: note that this consumes TAG attention; i.e. competes with other stuff. Are we ok with that? I guess so.
<Stuart> fyi "wget --debug http://www.ihmc.us/users/phayes/PatHayes" yields a redirection (302) to HTTP/1.0
<Stuart> User-Agent: Wget/1.11.3
<Stuart> Accept: */*
<Stuart> Host: www.ihmc.us
<Stuart> Connection: Keep-Alive
<ht> HST does not believe that the class "information resource" does any useful work in GRDDL, as such. That is, its use in the GRDDL spec could not be put at risk by the decision that any particular resource was or was not an information resource. You could induce a partial definition of information resource from the GRDDL spec as "anything which has an XML representation is an information resource", but that a) begs the question of what a representation is and b) only c
<noah> Tim's gotten someone to work up the correct example for converged self-describing Web update at http://pastebin.com/f5bef462c
<noah> Henry, please note that the above may be helpful if you are doing the next updates to the Self-Describing Web finding
<Stuart> doh.... to "http://www.ihmc.us/users/phayes/PatHayesAbout.html"
<timbl> http://pastebin.com/f10e65d55 [Alternative RDFa example?]
HT: I was chatting with Noah about a notion of (X)HTML without document.write
JAR: what is HTML5's story about evolution?
action-145?
<trackbot> ACTION-145 -- Tim Berners-Lee to add public prose around his slides at the AC meeting to make the case for extensiblity and flexible XML -- due 2008-12-11 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/145
close action-145
<trackbot> ACTION-145 Add public prose around his slides at the AC meeting to make the case for extensiblity and flexible XML closed
TBL: yes, ok to finish publishing those TPAC notes
action-7?
<trackbot> ACTION-7 -- Tim Berners-Lee to draft a position regarding extensibility of HTML and the role of the validator for consideration by the TAG -- due 2008-12-11 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/7
close ACTION-7
<trackbot> ACTION-7 draft a position regarding extensibility of HTML and the role of the validator for consideration by the TAG closed
JAR: what's HTML5's language versioning policy?
DanC: no particular technical
mechanism; just think hard at each setp
... that's the sort of implicit policy; meanwhile, it's a formal
open issue
http://www.w3.org/html/wg/tracker/issues/4
SKW recalls 4 main topics from http://www.w3.org/2001/tag/2008/09/f2fkc-agenda
SKW: so how are we doing on those 4 points?
DanC: on HTML and URIs, I have an action (188)
[eek. catch up on what I said]
<ht> I asked Allen Brown about E and the pi calculus and he introduced his response with "E is to the pi calculus roughly as Common Lisp is to the lambda calculus."
TBL: re HTML work parts, there's modularity internal to W3C's work and then module boundaries with other orgs, e.g. IETF/URI/HTTP; the TAG has a bigger role in the latter
<ht> HST asks that we discuss internal module boundaries (i.e. SVG, MathML) with HTML5
issue-27?
<trackbot> ISSUE-27 -- Should W3C specifications start promoting IRIs? -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/issues/27
TimBL: TPAC discussion has persuaded me to change my position on the IRIEverywhere issue
<ht> There are strings out there with %## in them which can't be interpreted as UTF-8, or if so interpreted give the wrong answer
TBL: I learned there are lots of URIs
with %xx stuff in them which either...
... don't convert as utf-8 or
... if they're interpreted as utf-8 encoded, that misrepresents the intent of
the person who created the URI...
... these are the results of bookmarking form results,
where...
... the convention is to encode parameters using the form document's
encoding
... [not always utf-8]
TBL: these aren't just old URIs; more are being made all the time
<ht> HST believes they are mostly arising in cyrillic and sino-japanese language areas
<ht> TBL: Vote of thanks to Stuart for his contributions as chair