Re: Suggested EARL Changes

 "Sean B. Palmer"
> 03:15:33 <sbp> * Remove testSubject
> 03:15:40 <sbp> * Add reprOf

I'm concerned this moves away from EARL as being general purpose
evaluate/report on anything language, to a language for reporting on
webpages (with different "EARL"s for UA's etc.)

I'm also concerned that by introducing threading it complicates
evaluations which should be unthreaded, which as I see the uses is more
common (multiple tools/people make evaluations but testSubjects change
infrequently.)

Threading only seems to be of value within a single evaluator, ie I only
believe we can trust Tool says resource fails, Tool later says resource
passes - unless my particular EARL tool has defined some sort of trust
relationship between tools.  I can't see how EARL could define how Tool1
says resource fails, Tool2 later saying resource passes as we don't know
if it's a defective tool, or a change of resource, or some "unknown"
content negotiation.

(As to why CC/PP is not sufficient, is it's got no support, and the
current proposals make no sense so I can't see it having any support for
many years - content negotiation based on browser is a very bad idea.)

Jim.

(cc'd to Aaron as the original was...)

Received on Sunday, 9 December 2001 10:29:50 UTC