W3C home > Mailing lists > Public > w3c-wai-er-ig@w3.org > October 2001

29 October 2001 ERT WG telecon minutes

From: Wendy A Chisholm <wendy@w3.org>
Date: Mon, 29 Oct 2001 15:07:35 -0500
Message-Id: <5.1.0.14.0.20011029145905.02790ac0@localhost>
To: w3c-wai-er-ig@w3.org
Summary

- everyone should review and comment on SBP's EARL 1.0 schema
http://lists.w3.org/Archives/Public/w3c-wai-er-ig/2001Oct/0063.html

- SBP asked if EARL should be on the W3C Rec track. No one was eager to do so, but there would be benefits. WC is going to take the question to the WAI team.

- The two methods that seem most likely to associate EARL and content are as an annotation and via the link element in HTML.  Can't say we will only use one or the other since there are audiences for both.  However, makes retrieving earl a bit more complicated since both methods have to be implemented.  It seems the best thing to do now is to implement and test.  Jim is working on annotations [1].  Perhaps if Nick or someone else were to work on html:link?  Seems that our user scenarios need work.
[1] http://lists.w3.org/Archives/Public/www-annotation/2001JulDec/0033.html

Al has also been talking about a portal, but we all realized we weren't sure we knew what he was saying - Al - help?

- Since there is an issue of costs and attending the calls, we decided to hold the next meeting on IRC.  This will be on 12 November 2001.
clients and other info available at: http://www.irchelp.org/
Once you set up a client here is the relevant info:
server: irc.w3.org
port: 6665
channel: #er

I hang out there when I am connected.  type "wendy" to get my attention. If you have trouble, please send me e-mail and I'll try to help you work out any kinks before the meeting.


Detailed minutes

where does the earl go?
* wendy summarizes discussion
SBP RSS discussions about mime-type. went ahead to register.
.. since so structural.
.. don't think earl needs one - should be the same as the serialization as RDF.
.. if EARL in XML RDF get RDF, if use n3 get whatever is used for that, RDF core currently saying text/plain
.. but I don't agree w/that. it's an app.
HB text/plain sufficiently generic that what we do should work w/it.
SBP Don't see much sense in mime-types. provide a hint, but just process. if it works great, if not, tough.
.. w/earl, could link to it.  Link that says, "this links to earl profile" a portal (from al)
.. not sure if we would want to do that.
HB If several people assess the same site would we need to link among them.
SBP that would tie back into what al is talking about. from portal, link to diff results for diff people for a page.
.. link to profile, profile link to all results.
HB Gain any political benefit by encouraging people to put a link to beginning of portal whether has result for htem or not.
SBP The more tech we introduce, more for people to learn. Languages like RIDL would be perfect for that.
CR Not sure makes sense to link from the page to EARL.  If a center evaluating sites for accessibility, would not be able to modify the page itself.
WC Think both models: as annotation and as linked to from w/in the content would be useful.
.. conscious author and evaluator.
WC SBP how this related to 1.0?
SBP One bug: date. Mostly just making more stable.  NIck has been complaining that there is not a spec,
.. learning from examples.
.. said difficult.
.. bits of it here and there.
.. schema is prime defn. need to knwo rdf schema to read.
WC I was to work on primer, you and Nick to save msgs in www-archive for us to use as reference.
SBP thought we had discussed more.
HB Self-interest, if expect things like Goggle to find it, put a link in if knew there were one.
SBP Nick and Jim said they would write something that would grab from link element.
WC sounds like at least two options, through link in HTML and as annotation.
SBP Discussions don't really seem to be going anywhere w/annotea.
WC Just presented paper. Interesting scenario, class learning about web content.  Teacher annotates w/earl, Students take and fix.
Proposed resolution: get implementations of both methods: earl in annotation and earl linked to from LINK element in HTML content.
... xlink.

EARL 1.0
SBP Nick invented some properties, like line numbers. Should adopt into 1.0?
.. also Len's BNF to XML thing.
.. xpointers. ways of pointing to content that you are evaluating.
.. a case of how much you put in or say "you have to implement this."
.. earl is a framework. doesn't say much beyond, Here's how to annotate something.
.. doesn't get into the details.
.. that's what nick did, by pointing to line numbers.
.. pointing to a site or range of pages w/in a site.
.. you could say this piece of earl is for an entire site except these 3...
.. priority levels of things it excludes, go into detail.
.. depends on how much we want to put in 1.0. how much we expect them to implement.
.. how comfortable people will be inventing properties.
.. NK invented line and col #. He needed properties to point to that.
.. if someone else also invent them, they would have to create a diff transformation.
.. but concept of the semantic web.
.. people are not comfortable. Jim said "earl doesn't have a method to point to a whole site"
.. i said could use a variety of RDF things, and Jim wasn't comfortable w/that. Perhaps should be in EARL.
.. if it's there, people would feel more obliged to implement it.
.. if don't put it in, won't have to implement.

WC annotations...
SBP annotation and earl should be different and separate.
.. if annotaiton is only how to find earl then does not affect what earl says.
.. that's my worry. have 2 page addresses.
HB takin this to QA?
WC Yes.
SBP If specs are to give out EARL evaluations, they need EARL to have an official grounding. 
.. EARL should be on Rec track
.. if used as site-wide thing.
WC At F2F talked about XAG being on Rec but not EARL
SBP prety sure we talked about it w/WL, he said would be daft.
WC Should EARL be on the Rec track?
CR Doesn't matter, would slow it down. I would say no, but not a strong opinion.
HB No strong opinion.
SBP As long as a clear status and know it won't disappear.  As long as people can feel comfortable using. When moves to QA won't get reorged.
WC All ther easons we gave to take XAG to Rec could be applied to EARL.

Action WC: Take questions of EARL to Rec to WAI team.

SBP Test case that ATR has are HTML files, WCAG would be a spec checkpoint (in the spec) rather than bits of tech.
.. thought we needed to properties for that.
WC Will NK have many changes if move to 1.0?
SBP not sure.
12 November - next mtg
WC for future meetings - just irc, mix of irc and phone (including jim's broadcast).
.. do we want to try 12 november meeting as just irc?
HB In lieu of telephone?

next meeting: 12 November, only IRC no voice.
Received on Monday, 29 October 2001 15:03:11 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:10:39 GMT