My mistake - I can't make today (Wed.) at 14:30 UTC - but the show goes on!

I screwed up big time and double-booked myself for today at 14:30 UTC. 
So I can't make the scheduled "second meeting" this week. My apologies 
to everyone.

Instead, let me post an update of where we stand, and then suggest that 
people still meet and some things that need to be worked on:

Query Language:

(1) Implementation report

Needs:
   + to use the conservative algorithm (EricP)
   + to have a stricter mapping from facets to tests (LeeF, but needs 
info from EricP on where to adjust this mapping)
   + to have green boxes from two implementations for each feature *
   + support for syntax tests (EricP)
   + some indication of multiple implementations of REDUCED *

(2) Disposition of Comments

Needs:
   + to be assembled (LeeF)
   + any outstanding comments to be addressed (TBD)

(3) Spec

Needs:
   + resolution of the {{ ... }} optional+filter issue (LeeF, 
implementors, working group)
   + examination of 7 tests in 
http://lists.w3.org/Archives/Public/public-rdf-dawg/2007JulSep/0177.html 
to see if implementations are broken by the fix to the bug in the spec 
(assigned to LeeF in the meeting, but perhaps Chimezie can do this?)


Protocol:

(1) Implementation report

Needs:
   + to successfully run the protocol tests against multiple HTTP 
endpoints (EliasT)
   + to put together a format for the implementation report (LeeF)
   + to figure out how to represent the SOAP testing in the 
implementation report (AndyS)

(2) Disposition of Comments

I've gone through and identified all of the protocol comments since we 
published CR on 6 Apr 2006. There are not many. Most are addressed, some 
need to be.

Needs:
   + to address remaining comments (LeeF)

Results format:

(1) Implementation report

Needs:
   + to have a format akin to the impl report for the protocol (LeeF)
   + to take results from Andy's XSLT (consumer), Eric's software 
(consumer), Elias's python (consumer) in conjunction with the results 
from the protocol testing (producers) to demonstrate interop (EliasT, LeeF)

(2) Disposition of Comments

No comments have (yet) been received since our recent CR publication.


Other - registration of MIME types (IETF, EricP)



I've broken out the *'d items which don't have a clear owner / path for 
more discussion here:


   + to have green boxes from two implementations for each feature *

If we don't have this once we've processed all the implementation 
reports, then we either need to solicit more results, improve 
implementations, or have a quite good explanation for why we don't have 
it. Perhaps Andy and Eric can look at the state of the implementation 
report and determine what the shortest path to green is.

   + some indication of multiple implementations of REDUCED *

We discussed on the call three (I think) weeks ago that REDUCED -- for 
testing purposes -- is mostly syntactic, since the number of results can 
vary. Should we add a syntax test for this? A query test? Can we gather 
two implementations from the WG that handle REDUCED to fulfill our exit 
criteria? Perhaps Andy and Steve can suggest the shortest path forward here?



I think there's enough there that doesn't rely on me to make it 
productive for people who are around to meet today at 14:30 UTC without 
me. Again my apologies for fouling up the scheduling.

Please mail updates on any of the above progress to the list ASAP. We 
really are on a tight schedule now.

And if you're not mentioned by name above but think you can help, please 
do! The above is just suggestions!

thanks,
Lee

Received on Wednesday, 3 October 2007 05:38:42 UTC