Re: [ws-rx-implement] Interoperability and the F2F

[Anish Karmarkar <Anish.Karmarkar@oracle.com>]

2 comments below.
-Anish
--

Doug Davis wrote:
> 
> In terms of presenting something to the TC during next week's
> F2F, what about this:
> 
> Jan 27 - Send a draft scenario document to the TC.
>          This scenario doc should try to cover interoperability
>          for the features defined in the core spec.
> 
> Feb 10 - Final date for feedback from the TC on the doc.
>          TC doesn't need to formally approve it but we request
>          suggestions/comments by this date.
> 
> Feb 24 - Unless we totally messed up the first draft and just need
>          to make minor changes based on the feedback, we
>          produce the official scenario doc that will be used
>          for the interop event.
>          Send out an official notification of the interop event
>          and doc(s).
> 
> Apr 3  - Shoot for an interop event this week.
>          This is 6 weeks after the scenario doc is published.
> 
> Questions:  
> 
> Both specs or just core spec?  Do we want people to support
> ws-policy?  I'm leaning towards just the core spec.
> 

This is tricky, since WS-Policy has not been submitted anywhere and the 
TC will have to abstract out the policy assertion/parameters (whatever 
that means) if WS-Policy is not 'sufficiently further along the 
standards path' (whatever that means).

> What type of event do we want?  Personally, I like the idea of
> a virtual interop but I know that the pressure of a f2f helps
> force people to make the dates.
> (virtual == put up an endpoint (RMS and RMD) on the web)
>  for example: http://wsi.alphaworks.ibm.com:8080/wsrm/index.html
>  so people can test at will)
> 
> Do we need to shift the dates based on the next CD?  With these
> dates I was assuming we'd use the current WDs available.  What
> version of the spec(s) do we want to use?
> 
> What kind of results do we want to publish and how do we want to
> publish them?  Generic: "it went well" - or specific "company X
> passed 1,2,3 but not 4,5,6"  ???  Do we even want to publish
> anything at all?  What does Oasis require?

I don't think OASIS has any requirement in this area, so it is really up 
to us. I would imagine companies may not want to publish publicly that 
their endpoint does not pass tests p, q and r. I would suggest making 
this member visible unless everyone agrees that it is OK to publish all 
the results.

> 
> thoughts?
> 
> thanks
> -Doug