Content of the email, which ended up being messy:
- a set of comments with an attached patch implementing them.
- a request for the deletion of 3 sentences (comment that I had done a
few weeks ago).
* Christopher B Ferris <chrisfer@us.ibm.com> [2002-10-24 07:58-0400]
> If others have also offered feedback and/or suggested changes, I would
> encourage
> them to identify their comments so that their feedback can be likewise
> incorporated.
First set of comments
---------------------
I had been trying to tweak Heather's contributions in order not to
imply that the architecture was registry-centric by commenting on
HKsContribution.triangle.htm[1], but I obviously missed
HKsContribution.discoveryagency.htm[2] in the email storm.
HKsContribution.discoveryagency.htm is now section 3.3.4[3], so I will
comment on this.
Considering the time constraints, I will try and not start a long
debate, just suggest simple, yet important for the sake of
generalization, edits.
First, a purely editorial comment: there are some redundancies in the
text. In the 3.3.4 introduction, the end, from "The simplest, ..."
until 3.3.4.1 is redundant with what is explained later, and what was
said in the previous paragraph.
The following is very UDDI-centric, and I think that we could add a
few "such as", or "for example".
3.3.4.1.2, which lists examples, should IMO list publication on a Web
site (WSDL document sitting at a URI).
The following:
| For service description repositories that span hosts within an
| enterprise, a service provider would publish to a private UDDI
| registry.
suggests that UDDI is the only answer. I would change "would" by
"may".
In 3.3.4.2.3, I would change "WSIL, unlike UDDI, is not suited for
many/most of the dynamic discovery/bind scenarios" to "Inspection,
unlike registries, is not suited for many/most of the dynamic
discovery/bind scenarios" to generalize concepts. I am sure that this
very sentence can spark an interesting debate, but I just want to
focus on editorial changes here.
I would also remove the "index.html" example: index.html isn't a well
known specified place. "/w3c/p3p.xml" is a better example.
Attached is a patch against wd-wsa-arch.xml, revision 1.19 dated
$Date: 2002/10/24 15:37:40 $, which implements my suggested changes.
I apologize for any weirdness in the patch encoding: the file was in
some weird Mac encoding that Emacs & diff were fighting over.
[ Having just edited the document, I see that some of my comments went
away when Joel's edits were incorporated. ]
Second comment
--------------
One last thing: I had suggested[4] to remove mentions of SOAP 1.1 in
what is now section 3.3.1.2, since SOAP 1.2 is near completion, and I
don't think we need to behave as historians here. I would remove "The
current industry standard [..] input specification for the XML
Protocol Working Group [XMLP 00]." (3 sentences)
I will do a closer reading of the draft in the next few days (I just
looked at the discovery section for now).
Thanks.
Regards,
Hugo
1. http://lists.w3.org/Archives/Public/www-ws-arch/2002Sep/0223.html
2. http://lists.w3.org/Archives/Public/www-ws-arch/2002Sep/0219.html
3. http://dev.w3.org/cvsweb/~checkout~/2002/ws/arch/wsa/wd-wsa-arch.html#IDAGCH2E
4. http://lists.w3.org/Archives/Public/www-ws-arch/2002Oct/0040.html
--
Hugo Haas - W3C
mailto:hugo@w3.org - http://www.w3.org/People/Hugo/