W3C home > Mailing lists > Public > xml-dist-app@w3.org > January 2006

RE: Comments on the "drop response-only" Part 2 draft

From: <noah_mendelsohn@us.ibm.com>
Date: Mon, 30 Jan 2006 17:56:08 -0500
To: "David Orchard" <dorchard@bea.com>
Cc: xml-dist-app@w3.org
Message-ID: <OF7EFC8502.01AC8B8F-ON85257106.007C9B19-85257106.007DFDB4@lotus.com>

David Orchard writes:

> I think that it's now time to call the question on which approach to
> follow.  We have sufficient information afore us to make a decision and
> proceed.

I think that's basically right.  As I mentioned earlier, I must send 
regrets for this week's call since I am at the Schema meeting. 

Also as mentioned earlier, my strong preference is for something closer to 
the draft I prepared before we started reviewing Dave's.   Rewriting the 
state machine descriptions makes some sense in principle, but I don't see 
it as closely related to what we need to do to succeed at what our users 
have asked us to do in this round, or that when members chose their 
representatives (or lack thereof) for this round of work that they did so 
with an eye to this sort of rework.  Still, I don't absolutely object if 
the WG decides they want to rework the state machine presentations as Dave 
suggests.    Please do look at my comment pointing out that Dave's draft 
seems to provide a bit more detail in the HTTP binding than it does for 
the MEP(s);  I think the binding is more or less fine, and the MEP 
descriptions should have the same level of detail as the bindings.

> I can certainly live with keeping the response-only mep

Good, thank you.  I would strongly non-concur with the proposal that drops 
response-only.

> if that would gain people's support.

Well, as I say, I think that fooling with the state machines is beyond 
what we need to do or should do to declare success.  Still if the WG 
prefers Dave's approach (and unless Chris has some other reason for 
objecting on behalf of IBM), I can live with the two MEP framework that 
Dave proposes.   Most of my review of your drafts has been on the "no 
response only MEP" version, so I'd want to do a detailed review before 
finally blessing the one that has the response MEP, but it should be easy 
to see which of my comments apply to both and which apply only to the one 
without the MEP. 

Bottom line:  as a general direction I much prefer the smaller change 
that's embodied in my draft, but I can live with Dave's two MEP draft as a 
framework for moving forward if a majority of the WG agrees with him that 
it's a  superior base.   Chris can speak formally for IBM while I'm gone.

> I think that it's now time to call the question on which approach to
> follow.  We have sufficient information afore us to make a decision and
> proceed.  I'd like to query the WG members to find out who supports the
> state machineless MEP approach for our direction on the
> request-optional-response problem, vs who supports the state machinefull
> (status quo) MEP approach.

Yes, I think it's reasonable to choose an approach around now.  Some 
reluctance that I won't be able to be on the call, but if you get a good 
quorum of others then you shouldn't wait for me.  I would ask that a 
decision like this not be made without a solid quorum, preferably 
including a good group of those who have expressed serious interest.  I'd 
be grateful if you'd consider my email comments and concerns, and also the 
detailed comments that I submitted on Dave's draft. 

Thank you.

--------------------------------------
Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------
Received on Monday, 30 January 2006 22:56:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:21 GMT