See also: IRC log
<trackbot> Date: 02 June 2010
<scribe> scribe: janina
john: let's start with open
actions review
... judy's actions?
judy: have been working on a
comparison to iso, seems orthogonal
... will post my conclusions
... so complete on iso comparison, but not on checking with
other key individuals
... also unsure how to engage external people effectively
without overwhelming them
... while action wasn't recroded, i did some work on the
disability language in the doc, and will still complete
that
john: is frank here?
[silence]
john: john: not sure how to break
our log jam and move forward
... much media will be pushed via the net, whether via
specialized ua, or a general browser, our doc is
important
... we don't have the luxury to continue debating the doc,
though
... our survey was to make sure our doc is complete
... think going through the doc detail by detail is
premature
eric: perhaps i misunderstood
what our purpose was in these requirements
... i understood that these would be required of any ua that
implemented html5 media
john: my understanding, from the a11y side, is to fully participate in consumption of this media
q!
<MikeSmith> I think other implementors are liking to view requirements in the same way that Eric just described -- as implementation requirements
john: for instance, sign-language
differs from country to country
... it'll be a tricky balancing
shawn: i think a lot of our requirements aren't technically difficult
eric: not so sure
... hard to decode on small platform
shawn: but there are other ways to achieve this
janina: we need it in the spec so that it can be covered
eric: but, even on the desktop
smil2 is lots of work
... so not necessarily easy to do
... are these rfc2119 musts?
shawn: not necessarily that we need smil, but that we need the ability to do these things
judy: we specifically separated user needs to avoid this conundrum
eric: but we need to discuss what is required
judy: absolutely agree
eric: my concern is that 'requirement' rfc2119 'must' on the user agent
john: i think everyone here is
realistic
... we should remember the concept of "graceful
degradation"
... yes, going back to the described video example, it's a
balancing thing here too
... some kind of switching mechanism may be needed between a
generalized broser and a more specialized player ua
janina: it's a question of what html5 wants to be, if a rich media platform, all these requirements will be needed somewhere, though probably almost never on any individual content item
<Zakim> Judy, you wanted to say that's why we separated out what the user needs vs how they can be technically achieved, so we wouldn't go into that
judy: we keep hearing the
question of whether we need to prioritize these
... should we discuss prioritization?
john: also my question in some
ways
... a bit disappointed in so few responses
<Zakim> MikeSmith, you wanted to say what we ultimately need in regard to the spec is a specific (sub)set of requirements for HTML UAs
mike: i think eric's response
will be typical of browser developers
... the spec provides conformance criteria
... if there are requirements on something other than browsers,
then it's not about html
janina: i can support rfc2119 should, but the spec needs to comprehend most of these because the intent seems to be a full gammut of media hosted via html
john: so our experience to date
has been that addons and plugins may play something launched
from the browser
... expect the browser will be more limited janina: not sure
that browser may need to drive it all
judy: want to raise a different
point ...
... want to come back to the question of what will help move
forward
... only one person on this call has filled out the survey, so
can we ask a new deadline for survey?
john: so, is this a wrong understanding of where we're headed?
eric: what?
john: html
... is should language overly onorous
eric: seems we want interoperable content on the web, so a spec that all user agents that most people use will implement
eric; so description of any feature in the spec needs to be so well described that it will be implemented consistently
shawn: but it's about how we
capture author intent, not how user agents provide it
... roughly the same way sounds ok, but not exactly the same
way my competitor does
eric: so spec doesn't say how to
do some thing, but what it should do
... so external file format needs specification
john: so, eric, you mentioned
multiple audio streams unsupportable on some devices
today
... so maybe not today, but who knows about a year from now? I
recall touch interfaces were experimental just a few years
ago
... we just don't want today's device limitations constrain our
specification of what's needed
shawn: so there's an issue of
balance between providing the data in a form that something can
be done with it
... so that's authoring--give the data in a usable way
... the must of what the content supports
... and the should of what the ua does
... we need to capture the musts
john: agree
janina well said
john: do we have agreement of where to go with this?
silence -- taken as ascent
john: so, judy ...
judy: i expect so few have completed the form because there's so much to do in completing it
shawn: will do, but want weekend
janina: ditto
kenny: end of today
john: by monday
mike: don't plan to respond
michael: don't feel qualified on this
judy: monday is better for me
john: so can we extend this until
monday?
... and we need to send around another explanation that we want
to capture user requirements
<Judy> +1 to accept input via email as well
janina: extend to 23:59 boston time monday
john: so we should have more complete results next wednesday
<Judy> janina: not sure that's the correct breakdown
<Judy> ...Sean Hayes helped us understand constaints around the authored requirements
<Judy> ...the must part might be on what the spec provides
<Judy> ...maybe the should are what the user agent can do
<Judy> ...if this is going to be a media platform, then we don't want the breakdown to be 'we can do text descriptions but we can't provide for audio description tracks"
<Judy> Janina: if you don't have a way to support how you implement something in the spec then you're nowhere
janina: i'm not clear that prioritizing these is the right way to go
<JF> Janina: if HTML5 is to become the platform for media, then we need to ensure that we have a means to capture all of these requirements, alternatives and requirements specified
john: don't disagree, but wonder
whether we should start with author or 'must and should'
... think the latter because of what the engineers are
saying
... think we should start with user agent requirements
... if our user reqs doc is more or less complete, i suggest
our next discussion is ua requirements
<JF> Sean: next step is to turn user requirements to tech requirements
<JF> Eric: next step is to get clarification of some of the user requirements
<JF> Sean: agrees with Eric
shawn: next step is to turn these into tech reqs, then to decide what's on content and what's on ua
eric: think we first need to agree on the user reqs
john: so that should be our
agenda for next week--item by item discussion and sign
off
... any other business, questions for now?
... thanks to everyone!
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/time/time monday/ Found Scribe: janina Inferring ScribeNick: janina Default Present: +1.650.862.aaaa, Janina, +1.408.307.aabb, Michael_Cooper, John_Foliot, Eric_Carlson, +0154558aacc, Judy, Sean_Hayes, MikeSmith, +61.3.986.4.aadd, Kenny_Johar Present: +1.650.862.aaaa Janina +1.408.307.aabb Michael_Cooper John_Foliot Eric_Carlson +0154558aacc Judy Sean_Hayes MikeSmith +61.3.986.4.aadd Kenny_Johar Regrets: Markku_Hakkinen Geoff_Freed Laura_Carlson Agenda: http://lists.w3.org/Archives/Public/public-html-a11y/2010Jun/0008.html Found Date: 02 Jun 2010 Guessing minutes URL: http://www.w3.org/2010/06/02-html-a11y-minutes.html People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]