See also: IRC log
SAZ: Welcome to the call, and Happy New Year!
SAZ: Tentative meet at the Technical Plenary
... Good for cross WG interactions.
... Maybe meet with WCAG WG and QA-IG folk
... 2 proposals, either in that week, or find another date/venue
... Chaals you said Monday/Tuesday aren't any good for you?
CMN: I'm booked up for the entire week...
CarlosI: Probably Yes.
ChrisR: I'd love to, but no budget.
JL: No problem with any of the dates
Sandor: 27/28 are okay with me
<chaals> [If any days are going to be possible for me they will be Monday/Tuesday]
CMN: Monday and Tuesday are possible.
SAZ: Are there any objections that we hold the meeting?
CMN: I won't object, but it is a little annoying that it's planned so late
SAZ: I'll check with Carlos and Johannes
... okay it's a tentative yes
David Rooks: I'm from Segala mtest, I'm based in London and we're looking at joining the ert group.
CarlosR: Hi, I'm Carlos
CMN: Hi, I'm Chaals
ChrisR: Hi, I'm Chris
JL, Hi, I'm Jim
Sandor: Hi, I'm Sandor
SAZ: Big thanks to Chaals!
SAZ: There are editors questions in the WD, but
the spec is mature, should we wait for the Earl Guide.
... First proposal is we just remove example 9
CMN: The example is invalid, and doesn't really
seem to make sense.
... The use case is covered in example 3 I believe
SAZ: The idea was to be able to link to a
requirement online without having to specify it all, and it's already an RDF
feature and is not needed in the spec
... Remove example 9, all fine?
... okay, let's remove it!
<chaals> RESOLUTION: remove example 9
SAZ: in test mode, we added earl:mixed
... a manual was a human who said it was a pass/fail, automatic is where the computer decides if it's a pass/fail with no human assistance
... earl:mixed is useful for compound assertors using multiple results to get a result.
... We need specific definitions here
<Zakim> chaals, you wanted to propose....
CMN: My recollection was that the use case for
earl:mixed is where some org. has tested some pages, and some was by machine
some by human and don't want to specifically export the test details so want
to say a mix of techniques were using.
... The agreement as I recall was to support this use case to enable responsibility tracking so a manual test which implies a higher standard of correctness
SAZ: We should discourage earl:mixed for more
granular situations, so just for compound assertions
... earl:manual - is the definition clear
SAZ: earl:automatic - no human assistance. What about a use case where a human answers specific questions that aren't directly related to the decision, the judgement more comes from the tool.
CMN: I believe that's a manual test.
Carlos: I also beleive it's a manual as we agreed.
JL: I can see there are some not very manual computer assisted tests so there is a range, but it would be better for a strict seperation,.
SAZ: I agree
... maybe insert fully to make it clear in earl:automatic
... earl:mixed is clearly only really applicaple for compound results, so should we SHOULD things be included.
CMN: I see mixed for low quality results, so
it's likely people not adding more evidence
... we did a load of tests and concluded that your site sucks - that uses mixed, only in more detailed results we'd give more detail
SAZ: What is my test mode if we've got a load of tests we point to and then give a result for them, what's the test mode?
SAZ: so mixed is for crude blanket
... any disagreement/questions?
Carlos: I'm concerned about a lack of a use case of the earl:mixed - but I can live with it.
<chaals> proposal: Mixed - used for statements which do not provide further information about how the test was done - typically a summary statement which is not pointing to the basis for the claim.
SAZ: We can add we want feedback, this isn't carved in stone.
CMN: I can see a use case, there's some general statements that we want to say, but don't want to expose the processes
Carlos: This seems to be more like an UNKNOWN use case, which we don't need as we can turn it off.
JL: I like the idea of having something if it is a popular use case so we can distinguish between genuinely unknown and just no RDF info on it.
CMN: I'm happy to drop it
SAZ: I think we need to withdraw it without
everyone being present.
... I'd rather have it as an open issue item
CMN: I'd like to have a note saying "this property is likely to be removed type note"
SAZ: lets add this as an editors note then.
<chaals> RESOLUTION: The value is under threat of removaL..
SAZ: Seperate this WD and previous WD contributors?
<chaals> CMN: Don't think that is a good idea.
CMN: I've only ever seen a long list, or people in the WD, and that makes little sense, especially here
SAZ: so let's publish?
RESOLUTION: Publish it.
SAZ: next one we want to sync. with earl guide.
SAZ: The EO WG is developing the user interface
for this, and the ET task force is migrating the existing HTML tools list
... We have a lot more tools than we used to have
SAZ: The tools where we've contacted from tools
people are being tracked here
... Some are on hold as they're not really evaluation tools list, they're more authoring tools.
... Some have been dropped for various reasons
SAZ: Carlos has some suggestions for more
... What to do with the tools on hold? We could suggest to the ATAG folk that they take should take ownership
... or we could just chuck them in anyway, as they can be used in ER way although they aren't primarily
... If there's no opinions, let's send it to the EO people to sort it ou
... Carlos's proposals, adding API integration would be good
Carlos: I'm not so sure the "API integration" is the right approach, the more platform approach names is a better idea
SAZ: yep so proposal to change API integration
... there are other proposals, there's a classifications section proposals
... classifications were tough, more searchable criteria were better.
... I'll take it back to the tools folk!
Carlos: I'd like to add another proposal...
SAZ: Our next meeting is next wednesday 18th.