RE: D-AC007 & D-AC007.1: "Reliable" architecture

I agree with Suresh.  It is important to state clearly that we want a
reliable WSA.

Hao

-----Original Message-----
From: Damodaran, Suresh
To: 'Katia Sycara'; Alan Davies; Hao He; 'Hugo Haas '; www-ws-arch@w3.org
Sent: 6/22/02 6:25 AM
Subject: RE: D-AC007 & D-AC007.1: "Reliable" architecture

 
Here is a thought FWIW:
I think of reliability in this context as the "what"  - we want
"reliable architecture"
The "how" to achieve reliability consists of different things, as Hao
lists below, and I had earlier championed [1].
 
Because we have a "no duplication" rule applied to  the CSFs and Reqs in
the document, we are unable to multiply list all those things
that can be done to make architecture reliable. Now we are just left
with items that deal with precise and unambiguous
definition. So, I do hear why some of you want to call it "precise and
ambiguous/well-defined."
 
I do, however, think that doing this will remove the notion of how we
intend to achieve a "reliable architecture" from the document.
Please consider the following solution as an alternative.
 
Refer to other related CSFs in the document from 7.x and say that those
also are required to make a "reliable architecture":
i.e., add AR007.3 to AC007
 
AR007.3  Reliability of Web Service Architecture is enabled by AC00xx
(we can fill in this if you agree on the solution)
 
On the other hand, if we all agree that we don't need to say "WSA is
reliable," then we can do what you suggest.
But, I am hopeful that is not what you want.
 
cheers,

-Suresh 
Sterling Commerce   

[1]  <http://lists.w3.org/Archives/Public/www-ws-arch/2002Apr/0016.html>
http://lists.w3.org/Archives/Public/www-ws-arch/2002Apr/0016.html


-----Original Message-----
From: Katia Sycara [mailto:katia@cs.cmu.edu]
Sent: Friday, June 21, 2002 1:33 PM
To: Alan Davies; 'Hao He'; 'Hugo Haas '; www-ws-arch@w3.org
Subject: RE: D-AC007 & D-AC007.1: "Reliable" architecture


I agree with Alan about the potential confusion. That is why I suggested
for A-C007 (if it is not moved, or other goals are not moved to it) to
be called "well-defined" architecture rather than "reliable".
 --Katia

-----Original Message-----
From: www-ws-arch-request@w3.org [mailto:www-ws-arch-request@w3.org]On
Behalf Of Alan Davies
Sent: Friday, June 21, 2002 2:19 PM
To: 'Hao He'; 'Hugo Haas '; 'www-ws-arch@w3.org '
Subject: RE: D-AC007 & D-AC007.1: "Reliable" architecture



The potential confusion here is that any reference to a "reliable Web
Service" may cause the reader to assume that this is reliable in the
same way as a reliable Messaging Service is reliable - and I don't think
this is what Hao intends at all...

Regards, Alan Davies.



 <http://www.seebeyond.com/> Go to SeeBeyond website >" hspace=0
src="http://www.seebeyond.com/images/i_seebeyodlogoSmallnotag.gif"
width=180 align=absMiddle border=0 NOSEND="1">

  _____  




Alan Davies
VP Standards
ph:   +1-626-471-6050
cell: +1-626-437-0272
adavies@SeeBeyond.com <mailto:adavies@SeeBeyond.com> 









> -----Original Message-----
> From: Hao He [ mailto:Hao.He@thomson.com.au
<mailto:Hao.He@thomson.com.au> ]
> Sent: Friday, June 21, 2002 10:57 AM
> To: 'Hugo Haas '; 'www-ws-arch@w3.org '
> Subject: RE: D-AC007 & D-AC007.1: "Reliable" architecture
>
>
> 
> Let me explain what I had in mind when I wrote D-AC007.1 [1].
>
> If an architecture is reliable, then people can follow it
> correctly and
> build reliable
> Web Services. A reliable architecture is a pre-requirement of
> a reliable WS.
>
> More specifically, it means the following aspects:
>   1 Correct and consistent so there are no fundamental flaws in the
> architecture. (Solid theory and framework)
>   2 Precisely defined so there is no ambiguity.
>   3 Can be validated against use cases. (practical context)
>
> However, 1 and 3 are already discussed elsewhere.
>
> We have the following options:
> 1. Move D-AC007.1 to other goals.
> 2. Move other goals under D-AC007. (It would involves goals
> in D-AC005 and
> D-AC002)
>
> Personally, I prefer 2 but the changes seem to be big. 
>
> Hao He
>
> [1] http://lists.w3.org/Archives/Public/www-ws-arch/2002Jun/0088.html
<http://lists.w3.org/Archives/Public/www-ws-arch/2002Jun/0088.html> 
>
> -----Original Message-----
> From: Hugo Haas
> To: www-ws-arch@w3.org
> Sent: 6/22/02 1:58 AM
> Subject: D-AC007 & D-AC007.1: "Reliable" architecture
>
>
> There were some discussions in yesterday's teleconference about the
> word "reliable" in D-AC007 and more precisely in D-AC007.1:
>
> |   D-AC007.1 The Web Service Architecture is reliable.
> |
> |   D-AR007.1.1 The Web Service Architecture is precisely defined
> without
> |   ambiguity,
> |       D-AR007.1.1.1 using standard definition languages whenever
> |   applicable and available,
> |       D-AR007.1.1.2 using standard terms, and clearly defined new
> terms.
>
> It seems to me that reliable here is confusing: D-AR007.1.1 suggests
> that what is actually intended is that the architecture is defined
> with precise terms.
>
> If this is the case, I think that "reliable" should be replaced by
> "precisely and unambiguously defined".
>
> Note that such a change would also affect D-AG002 and D-AR007.2.3.1.
>
> If this is not what is intended, which is what Daniel was suggesting,
> then I think that D-AR007.1.1 isn't related to D-AC007.1, and we
> should have a definition of what a reliable architecture is because
> this is unclear to me.
>
> If we agree on this, I think that the whole D-AC007* could be
> approved. It seemed to be the last subject of discussion.
>
> Regards,
>
> Hugo
>
> --
> Hugo Haas - W3C
> mailto:hugo@w3.org <mailto:hugo@w3.org>  -
http://www.w3.org/People/Hugo/ <http://www.w3.org/People/Hugo/>  -
> tel:+1-617-452-2092
>