See also: IRC log
<trackbot> Date: 02 December 2011
<scribe> scribe:ddahl1
action-70?
<trackbot> ACTION-70 -- Chris Lilley to take the web on tv need for certification to w3m -- due 2011-11-25 -- OPEN
<trackbot> http://www.w3.org/MarkUp/CoordGroup/track/actions/70
debbie: Chris isn't here to update, so we can wait on that
action-71?
<trackbot> ACTION-71 -- Deborah Dahl to organise HCG discussion of media handling once group is public -- due 2012-01-06 -- OPEN
<trackbot> http://www.w3.org/MarkUp/CoordGroup/track/actions/71
action-72?
<trackbot> ACTION-72 -- Deborah Dahl to organise a discovery session after hcg is public -- due 2012-01-06 -- OPEN
<trackbot> http://www.w3.org/MarkUp/CoordGroup/track/actions/72
philippe: there are a wide range
of solutions, most draconian approach is everything is in a
Recommendation, more lightweight is a public wiki, there are
solutions in between.
... depending on how much process you want, XPointer is an
example of a lightweight process
... to register you need a pointer to a specification
... the less resources you have the more lightweight the
process
marc: requirements are that we
need vocabularies of emotions. ideally we would make them part
of the the specification, but there is no agreement
... we wanted to give people a choice among well-defined
vocabularies
... we have an XML format for describing vocabularies
... we want to group a selection of vocabularies in a central
place, but people are free to point to outside
vocabularies.
plh: that makes sense
marc: we want to maintain
snippets of XML, because the atomic entity in our registry is
the vocabulary.
... we have looked at a number of options, IANA was very
tedious, not a lightweight solution.
... can we put an XML representation in a wiki?
plh: you can put pointers to the XML file
marc: we would also like to have control
plh: if the wiki is public you
don't have control
... in HTML5 the wiki can have a status, that is when the WG
approves or not.
marc: we've decided to take the
human description and put it in a document.
... we originally tried to create a Note, but there was an
objection, so we created a WD, but with no expectation that it
would become a Recommendation
... we need a lightweight process for an irregularly updated
item,
... but the Note is supposed to be final
see the second link in this message: https://lists.w3.org/Archives/Member/w3c-html-cg/2011OctDec/0022.html
kaz: there are so many possible solutions and we want to find the best.
marc: which of the two seems appropriate?
plh: who will be in charge of
maintaining the Note in five years?
... I would suggest going for the wiki or the XPointer
registry
... should think about this for the long term
doug: should consider who is the
community who is going to be suggesting changes, whether a
small community or a widely diverse and unknown set of
people
... if you think there are going to be 20-30 items added over
the next five years, then that would point to a registry, if
it's less, you probably just need a document.
marc: probably will be the latter. we just need something that can be updated if necessary.
plh: the wiki would be the most lightweight.
marc: we had a wiki in the Emotion XG, but it was totally spammed.
plh: we got better at that
doug: can make it writable only by WG or someone with a W3C account
marc: can we add an XML link to the wiki
plh: if an XML can't be uploaded we should ask why.
marc: one issue that confused us
was the choice between a dated and an undated URI, in the case
of a wiki it would simply be the link
... are there other repercussions? a link to a wiki can't be
normative
plh: in HTML5 the wiki is normative
marc: didn't know that
plh: look at the definition of the meta-element
marc: we would move everything
from the current WD to the wiki
... is it correct that we have control over who has write
permission?
plh: the WG can be in charge, but
think about what happens in 5 years
... you can lock the page
marc: the thing to do is to try out the wiki
kaz: there is no essential
difference between a Note and a wiki, so fine to use a
wiki
... the definition of the vocabulary set in the current
vocabulary WD is defined by XML notation, but being referenced
to using HTML fragment notation. maybe we should use XPath, not
HTML
marc: we need to be able to point
to a permanent XML in the wiki
... other people should be able to point to a full URI, not
just an identifier
... we can't draw up the ultimate list of vocabularies, so it
must always be possible to point outside of the W3C, so that's
why we need a URI
debbie: is the write-restricted wiki the way to go?
marc: we need to verify that
there isn't any way for it to be harmed, and that we can point
to XML
... but it looks promising
plh: just try to put XML on a
public wiki
... to see if that works
<plh> http://www.w3.org/TR/html5/semantics.html#other-metadata-names
<plh> http://www.w3.org/TR/html5/semantics.html#other-pragma-directives
plh: these are some examples from
HTML5
... the wiki itself document the restrictions
<plh> http://www.w3.org/TR/html5/links.html#other-link-types
marc: it's reassuring that I don't have write permission.
doug: at TPAC we were talking
about the sarcasm element, and we came up with two good use
cases for this, e.g. accessibility, and for non-native
speakers
... if there were some way to annotate tones in written speech
or visual display, or have some kind of markup, but some people
may not understand a tone. that would be a place where a
registry might be useful.
marc: this goes in the same direction that we've been thinking.
doug: machine could be detecting emotion and outputting HTML
marc: we had a comment from the accessibility WG, but out of scope for current work
doug: definite use cases for people in different cultures
marc: do you see how we're starting to come up with a vocabulary, and it's not trivial
doug: if you had a core set of critical information that you could put into tones, that could get traction.
marc: that's on the horizon.
doug: if we had a tone attribute and it represented sarcasm, you could style it as appropriate.
kaz: in SSML 1.0 we tried to extend speech synthesis to various languages, like Chinese or Japanese. EmotionML might need some internationalization variants
<kaz> kaz: e.g., text level vs. actual intention based on the cultures and countries :)
This is scribe.perl Revision: 1.136 of Date: 2011/05/12 12:01:43 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/who is going to be suggesting changes./who is going to be suggesting changes, whether a small community or a widely diverse and unknown set of people/ Succeeded: s/themselves/itself/ Succeeded: s/between a Note and a wiki/between a Note and a wiki, so fine to use a wiki/ Succeeded: s/the definition of the vocabulary set is defined by XML notation, so/the definition of the vocabulary set in the current vocabulary WD is defined by XML notation, but being referenced to using HTML fragment notation. / Succeeded: s/intention/intention based on the cultures and countries/ Found Scribe: ddahl1 Inferring ScribeNick: ddahl1 Default Present: Debbie_Dahl, Marc_Schroeder, Kazuyuki, Bert, Janina, Plh, Doug_Schepers Present: Debbie_Dahl Marc_Schroeder Kazuyuki Bert Janina Plh Doug_Schepers Regrets: Steven_Pemberton Found Date: 02 Dec 2011 Guessing minutes URL: http://www.w3.org/2011/12/02-hcg-minutes.html People with action items:[End of scribe.perl diagnostic output]