See also: IRC log
SAZ: Have mentioned this before in meetings.
Looks like WCAG will be going to last call by the end of this month. Will
then want to start developing some tests. Thought of having a joint task
force hosted inside ERT developing WCAG 2.0 tests.
... next on my to-do list is to propose a task force statement and present to the group. So this is a heads-up...
CR: started work on test suite for Legge Stanca - would like to see that public. Wonder if that would help with WCAG, too - can bring up on the list and show people
SAZ: In this task force we will be doing a top-down mode where WCAG gives a stable technique, we collect tests that are available, e.g. from you guys, bentoweb, to try and reduce actual development. Then check those tests to make sure they do the job and send them back to WCAG
CR: OK. The Italian stuff might help as tests forWCAG too.
CR: Could we do tests for the legge stanca here?
SAZ: Think that is out of scope for this group.
But aim is to have things be atomic and flexible so they can be re-used.
... we can't take on further work....
CR: Any problem posting stuff we have done?
CMN: Sounds like a good source of potential tests, so post away... (assuming that these are not protected by some kind of IPR of course)
SAZ: Who would participate?
JK: maybe... probably someone from here
CMN: Will look for someone at Opera, cannot do
... so not sure if we will have anyone or not :|
Niq: Not likely to have available time at the moment.
DR: Makes sense for Segala, too
CI: Not sure at the moment.
SAZ: The work has been done, the systems are ready - waiting for vendors to check their entries, but otherwise will be announcing "very soon" - can we close the task force?
RESOLUTION: The Evaluation Tools Task Force is closed.
SAZ: Please check that these schemes are good...
<niq> .. that'll help with round tuits
the plan is to review these and approve them next week.
<drooks> i have to leave around 3:45
SAZ: Talked during face to face, conclude that
earl:Software doesn't really fit in EARL, looking at other vocabs such as
... discussion with edd, DOAP editor, forwarded to the list. My take is that DOAP is not as mature as I would like, there are some issues...
... we could also publish a note.
... while ETTF was working they developed a schema that describes tools. Extends earl:Software to describe some of the features of tools.
... could publish that and let people use it or something else.
JK: There will be a property pointing to some other vocabulary?
CMN: No really strong opinion on vocabs.
Suggest we keep earl:Software since that schema depends on it anyway, have it
in as an optional thing people can use, and say "do what seems cool" to
... e.g. use earl:Software, DOAP, whatever. Let people play around a bit and see what settles in as the pattern from real use.
SAZ: Have people looked at DOAP?
CMN: Not worried if people do, not sure that I would go for "You should use DOAP"...
JK: Saw that it doesn't like to re-use other vocab.
SAZ: Think we should mention it in the EARL guide, maybe schema, but not rely on it.
CMN: Works for me in guide, schema text...
RESOLUTION: Will refer to DOAP in guide and in the text of teh schema doc, but not rely on it.
SAZ: Should we take earl:Software out?
... leaning to taking it out. we have done that with other things.
CMN: Suggest we leavve it in for prgamatic reason that you are shiipping code that refers to it, and there is no harm done - at the end, which namespace things are in is completely arbitrary. Don't think we should be requiring use of it though.
JK: This is an assertor, right?
SAZ: Can be an assertor or test subject.
... there used to be a subClass of one or the other, was decoupled, so can be both. Seems odd in the earl spec, but it is there for historical reasons.
DR: Don't care either way.
SAZ: Software could be extended - e.g. in the schema I noted above. A resolution could be to keep it there in schema, work with people like Edd and get them to mature their work, re-use it down the line.
CMN: Yeah, leave it as is, don't make it a requirement...
proposed RESOLUTION: earl:Software stays, will not be required
RESOLUTION: earl:Software stays, will not be required
SAZ: We were wondering what the overlap is.
DavidR did an analysis recently and posted -> http://lists.w3.org/Archives/Public/public-wai-ert/2006Mar/0007
... my take is that the CL is focussing much more on the labels themselves - they have a lot of vocab for describing the labels.
... we have barely anything to describe the test requirements itself. On the other side we focus on who said something, and the result, which they don't - it's just a resource that describes this.
... seems to me that both can be used, the use cases are different.
CMN: They have different focus, so there isn't direct overlap in much, but think they would be valuable used together. There is currently one technical issue that would need to be resolved, to do with how EARL and CL describe the earl:testSubject - the current data models are incompatible. We have appointed someone to the group specifically because we want to be able to mix them, and apply them to content.
DR: They can be used together so CL makes a claim, EARL can be used to prove it.
SAZ: CL doesn't have a way of saying a page
doesn't have a label, right? EARL provides more granularity?
... we wanted to see how RDF-CL use the subject - how they address several groups of pages and see if we can re-use that in EARL.
... any ideas on how to approach this?
CMN: Until we have spent more time in the technical aspects don't have detailed suggestions. Think that what we want is prety clear
SAZ: So what does ERT need to do?
CMN: Write up what we want and make sure they understand...
<scribe> ACTION: Carlos to write up ERT requirements on RDF-CL so we can make it clear what we hope they offer. [recorded in http://www.w3.org/2006/03/22-er-minutes.html#action01]
SAZ: Use cases for mixing these. Would like to
have some documentation of this...
... will XG do it?
DR: XG isn't doing it at the moment...
... Think these things should be used together?
<scribe> ACTION: Chaals to rough out use case for mixed EARL/RDF-CL [recorded in http://www.w3.org/2006/03/22-er-minutes.html#action02]
SAZ: Would like to see some examples, with code
CMN: Feel free to take the action item... :)
DR: Have to go now.
SAZ: Would love to have further feedback - I will rough out something....
CMN: I am not in the XG - it will be Kjetil Kjernsmo for opera
<scribe> ACTION: Shadi to write some examples mixing RDF-CL and EARL [recorded in http://www.w3.org/2006/03/22-er-minutes.html#action03]
SAZ: Nothing next week, next call 5? April