Re: new editor's draft of WSA available

Hugo,

Thanks for the comments! More below.

Cheers,

Christopher Ferris
Architect, Emerging e-business Industry Architecture
email: chrisfer@us.ibm.com
phone: +1 508 234 3624

Hugo Haas wrote on 10/28/2002 10:38:10 AM:

> 
> Hi Chris (and editors).
> 
> * Christopher B Ferris <chrisfer@us.ibm.com> [2002-10-21 12:44-0400]
> > Anyway, here's the latest WSA:
> > 
> >  http://dev.w3.org/cvsweb/~checkout~/2002/ws/arch/wsa/wd-wsa-arch.html
> 
> The comments below are against revision 1.19. I have marked as
> substantial comments that are not editorial, i.e. that don't need to
> be dealt with right away. "substantial?" means that it may be easily
> fixable, or maybe not (e.g. modifying a diagram).
> 
> => Section 2: "mutually agreed-upon" (substantial)
> 
> Considering the model where the requestor discovers the provider's
> interface and then interacts with the service, talking about agreement
> seems odd to me. More specifically, I am wondering the relationship
> with issue 1[1].
> 
> => Section 3.1: "publishes it to a requestor or service discovery 
agency"
>    (substantial?)
> 
> The document presents service discovery agency as an abstract role,
> that the provider, or anybody else for that matter, can take. This
> description seems to imply that a service discovery agency is a
> concrete third party.
> 
> => Section 3.1: "Intermediaries may processes certain functions"
> 
> Isn't that s/processes/perform/?

fixed.

> 
> => Figure 2
> 
> Unless I missed something, I don't see any difference between figure 1
> and 2.

hmmm... they were supposed to be different:) I'll have to go
back to the sources...

> 
> => Section 3.1.3: "Service Publication Stck"
> 
> Typo: Service Publication Stack. Should probably be a specref.
> 
> => Section 3.2.1: "Features are..." (substantial?)
> 
> The glossary has:
> 
>           An abstract piece of functionality provided by the
>           architecture.
> 
> Rereading it, I don't think that "provided by the architecture" is
> right and I would like to drop it. Do people agree?

agree...

> 
> => Section 3.2.1: "Asynchrony - dynamic and static" (substantial)
> 
> I must admit that this doesn't speak to me. We should add some text
> here explaining what is meant, and eventually put something in the
> glossary.

Doesn't do anything for me either... when in doubt, take it out:)
We can always add it back after we have figured out what it meant.
I've removed it for now.

> 
> => Section 3.2.1: "long running transaction, aka conversation"
>   (substantial)
> 
> The glossary reads:
> 
>    Conversation
>           A logical collection of messages exchanged between
>           communicating parties.
> 
>    Long-Running Interaction
>           A series of operations between a client and a Web service.
> 
> This will have to be reconciliated. I will add some text in the
> glossary to not forget about it.

sounds good to me. I've added in another aka for long-running interaction
in the WSA.

> 
> => Section 3.3.3.1: "For example, business context is described using"
> 
> Continuing my generalization: s/is/can be/

done

> 
> => Section 3.3.5: Overarching concerns (substantial?)
> 
> I think that privacy is something which will show up fairly often,
> whether one invokes a service, looks for one, publishes a description,
> including its policy.

think that privacy is one aspect of policy, no? certainly agree that
privacy is an important aspect that deserves some due consideration.
I've added an ednote.

> 
> => Section 4.2.1: XML Infoset (substantial?)
> 
> Regarding the comment about flexibility in the choice of
> serialization, I think that we could add examples, such as compression
> of an XML 1.0 document.

I've added an ednote.

> 
> => Section 6: Figure 7/8
> 
> The text talks about figure 7, but there is no figure 7, but 8.

fixed, but will likely be harvested and removed anyway.

> 
> => Section 9.1: Normative References
> 
> I think that WDs are informative references, i.e. 2, 7, 8 and 9.
> 
> => Section 9.2: Informative References
> 
> I would add the glossary and the usage scenarios document, and maybe
> add some text pointing people to them. In a next revision, we will
> need to link (and sync) the glossary and the architecture document,
> and the usage scenarios document too, which sounds like hard work.

Agreed. I've added references to req'ts and glossary.

> 
> Regards,
> 
> Hugo
> 
>   1. http://www.w3.org/2002/ws/arch/2/issues/wsa-issues.html#x1
> -- 
> Hugo Haas - W3C
> mailto:hugo@w3.org - http://www.w3.org/People/Hugo/
>