RE: Role of intermediary

Michah,
 
I'm glad we share many concerns in common... I need a little more help with
3. and 4. below, perhaps you could elaborate on these two a little.
 
Thanks,
 
Stuart

-----Original Message-----
From: Lerner, Michah, ALSVC [mailto:michah@att.com]
Sent: 25 January 2001 05:46
To: 'Williams, Stuart'
Subject: RE: Role of intermediary


Yes, these are exactly the concerns I share --- see emboldings below ..

1.	If the paths are not equivalent, then where is determinism?  

2.	Can the sender say "nope, I want a better success result than that?"
or should the sender indicate minimal standards with a "requires".  

3.	Can this mandate a specific security context, for example?  

4.	And, how on earth will it actually deliver something like #3 without
a formal protocol plus analysis? In some odd sense it becomes a straw man
for an independent AAA method running over a different 'provable" protocol
(i.e. potentially provable by virtue of simplicity)

-----Original Message-----
From: Williams, Stuart [mailto:skw@hplb.hpl.hp.com]
Sent: Wednesday, January 24, 2001 2:54 PM
To: Lerner, Michah, ALSVC
Cc: xml-dist-app@w3.org
Subject: RE: Role of intermediary


Michah Lerner asks:

Will the  <http://www.w3.org/2000/xp/Group/1/01/15-abstract-model/> abstract
model consider the practical questions of different bindings between
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#g340> initial XP sender and
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#g130> ultimate XP receiver?

I would hope that we can at least use the abstract model to ask concrete
questions about what we want XP do do for the 'entities' that use (or live
on top of) XP. In that regard I would expect it to try to take a position
with repect to the semantic properties of chains of intermediaries, where
there are questions about whether the relative positioning of XP blocks in
messages and the order that messages (and blocks) are processed at
intemediaries semantically significant or not to the overall operation
performed by the chain (and to the partial view of the operation seen by the
intermediaries the message passes through). There are also questions about
partial failure part way down a chain... do we want all or nothing
transactional semantics... do we want some hybrid where XP module can
indicate the failure behaviour required...
 
To the specific question passing through a chain of different bindings, I
hope that we can address things in a way that will address the practical
questions.... although that will become more apparent as a design begins to
'hang' on the bones of the model.
 
Best regards
 
Stuart Williams

-----Original Message-----
From: Michah Lerner [mailto:michah@att.com]
Sent: 24 January 2001 01:14
To: xml-dist-app@w3.org
Subject: Re: Role of intermediary


Mark's response to Marwan (Re: Role of intermediary) confirms the
requirements or specification do not  preclude an intermediary from
"receiving incoming messages using one protocol binding and forwarding them
using another".  R600 <http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#z600>
, R604 <http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#z604>  and R608
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#z608>  make this more
precise, and R612 <http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#z612>
identifies HTTP as the normative non-exclusive binding.  The composability
requirement of section
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#N1136> 4.4 as well as R505
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#z505>  reinforce this
because a priori knowledge should not be required by the endpoints. 

However (there is always a however), the Charter ( 4.6 Protocol
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#N1423> Bindings) warns of
potential "semantic complications") with some protocol bindings.  Will the
abstract  <http://www.w3.org/2000/xp/Group/1/01/15-abstract-model/> model
consider the practical questions of different bindings between initial XP
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#g340> sender and ultimate XP
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#g130> receiver?  Are there
specific use cases for example an SMTP(HTTP) sender with HTTP(SMTP)
receiver?  What about the scenarios of R502
<http://www.w3.org/TR/2000/WD-xp-reqs-20001219/#z502> ? 


What prevents extra complexity (and problems) when endpoints bind to
different protocols? 


Thanks! 


Re: Role of intermediary 

From: Mark Needleman - DRA (mneedlem@dra.com) 
Date: Fri, Jan 19 2001 


*Next message: Mark Nottingham: "Re: Role of intermediary" 


   * Previous message: Marwan Sabbouh: "Role of intermediary" 
   * In reply to: Marwan Sabbouh: "Role of intermediary" 
   * Next in thread: Mark Nottingham: "Re: Role of intermediary" 
   * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] 
   * Other mail archives: [this mailing list] [other W3C mailing lists] 
   * Mail actions: [ respond to this message ] [ mail a new topic ] 


  ------------------------------------------------------------------------ 


Date: Fri, 19 Jan 2001 15:52:16 -0600 (CST) 
From: Mark Needleman - DRA <mneedlem@dra.com> 
To: Marwan Sabbouh <ms@mitre.org> 
cc: xml-dist-app@w3.org 
Message-ID: <Pine.OSF.3.95.1010119154939.30452M-100000@tourist.dra.com> 
Subject: Re: Role of intermediary 


Marwan 


I dont believe there is and i see the ability to do that as a legitimate 
and useful function 
  


Mark H Needleman 
Product Development Specialist - Standards 
Data Research Associates, Inc. 
1276 North Warson Road 
P.O. Box 8495 
St Louis, MO 63132-1806 
USA 


Phone: 800 325-0888 (US/Canada) 
       314 432-1100 x318 
Fax: 314 993-8927 


Email: mneedleman@dra.com 
  


On Fri, 19 Jan 2001, Marwan Sabbouh wrote: 


> I have this question to the group: Is there anything in the spec that
might prevent an intermediary for receiving incoming messages using one
protocol binding and forwarding them using another? 
> 
> Thanks. 
> Marwan 
> 


  ------------------------------------------------------------------------ 


   * Next message: Mark Nottingham: "Re: Role of intermediary" 
   * Previous message: Marwan Sabbouh: "Role of intermediary" 
   * In reply to: Marwan Sabbouh: "Role of intermediary" 
   * Next in thread: Mark Nottingham: "Re: Role of intermediary" 
   * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] 
   * Other mail archives: [this mailing list] [other W3C mailing lists] 
   * Mail actions: [ respond to this message ] [ mail a new topic ]