W3C home > Mailing lists > Public > www-ws-arch@w3.org > July 2002

RE: [hst] harvesting subteam

From: <michael.mahan@nokia.com>
Date: Fri, 12 Jul 2002 11:49:03 -0400
Message-ID: <5C76D29CD0FA3143896D08BB1743296AB89A68@bsebe001.NOE.Nokia.com>
To: <www-ws-arch@w3.org>

[this is a repost to the public list of a message sent earlier
to the member list sent Wed 07/03/2002 12:04 PM EST]

Hi All,

Here's our status as I see it, broken down into items.


Item 1: Approach
It seems like we have concensus on the approach. On to the identification 
process of candidate systems/specs. Hao suggested that we trim the list back 
at the end via a vote. That is fine with me.

The current list:
1. SOAP/XMLP
2. WSD
3. OMG/Corba
4. REST
5. Meerkat
6. XML-RPC
7. UDDI

Please send other high relevance candidates. Lets finish this step immediately
so we can distribute the work this week.


Item 2: Scope of harvesting effort. 
Should the boundaries be:
a. Components and connectors i.e. SOAP node, SOAP message, etc.
b. All system/spec features which have arch relevance or visibility i.e.
   MEPS, processing model description, etc.

Please send your vote preference. I think we have Mark Baker's and Chris's
vote.

I personally vote for (b). While realizing this is more challenging to make
our deliverable date, I think our first duty is to the harvesting role and
the analysis/recommedation role is secondary.


Item 3: Output format
Mark Jones suggests following the TAGs lead by partitioning our output
into identifiers, formats, protocols. What do people think about this? My 
question is: would this format cover the 2 outputs I suggested, the raw
findings and the analysis/recommendations? (Both items 3&4 below)


For reference, here's the original approach:

1) List sources or web service arch. artifacts of web architecture models 
   that have direct relevance to WSA (large-scale, distributed systems).
   Everyone pitch in here with suggestions.
  a. XMLP
  b. WSD
  c. OMG
  d. REST
  e. ....
  f. your input here

2) Divide/conquer. Distribute each source to a volunteer. 
   For each source:
    a. summary of WS architectural entities found
    b. describe the notation, modeling method used: MEPS, block and line 
       diagram, etc. (We may need to ignore use cases?)

3) Consolidate the findings in some fashion
4) Analyze the finding and make recomendations on:
  a. Architecture modelling method/notation to use
  b. Most compelling/defined architecture elements
Received on Friday, 12 July 2002 11:50:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:01 GMT