W3C home > Mailing lists > Public > www-ws-arch@w3.org > January 2003

Re: Summing up on visibility(?)

From: Mark Baker <distobj@acm.org>
Date: Thu, 9 Jan 2003 12:49:17 -0500
To: Miles Sabin <miles@milessabin.com>
Cc: www-ws-arch@w3.org
Message-ID: <20030109124917.M529@www.markbaker.ca>

On Thu, Jan 09, 2003 at 04:48:19PM +0000, Miles Sabin wrote:
> If the question is simply "Is information in HTTP headers more easily 
> accessible to intermediaries than information in XML request 
> entities?", then my answer is: Yes, obviously (so long as those 
> intermediaries aren't already parsing the entities anyway).

Not just headers, the whole PDU; request line, headers, body.  But sure,
HTTP headers are more visible to HTTP intermediaries than SOAP headers,
for obvious reasons.  But ignoring that issue, do you see any improved
visibility there?

If Mike wants some text, I would phrase it like so;

"Web services wishing to take advantage of the visibility[ref]
properties of the REST architectural style, should avoid placing method
names within the body of the message being sent (when sent over HTTP),
since doing so has the effect of hiding the meaning of the message to
intermediaries, thereby reducing visibility."

Mike wrote;
> I'm lost, as usual.  Where precicsely does "visibility" fit into the WSA
> framework, in your opinion.  Discovery?  Or are you asserting that REST
> doesn't require the distinction between discovery and invocation if messages
> are "visible"?

Visibility is just a property of an architectural style.  It has some
beneficial qualities, IMO.  I don't think it relates to discovery.

http://www.ics.uci.edu/~fielding/pubs/dissertation/net_app_arch.htm#sec_2_3_5

Where it fits into the WSA framework, I'm not quite sure.  Hmm ...
I do note that the word "connector" isn't yet used in the WSA document,
except in the harvesting material.  Visibility is impacted most strongly
by constraints on the connector elements.  So perhaps that's why it
doesn't seem to fit yet.

Also, the architecture document doesn't list the desirable properties
it's aiming to achieve.  If it did, I would recommend that visibility be
on that list, for the reasons Roy lists at that URI above.

>    Or is using SOAPAction a best practice that you suggest we
> advocate?

Yes, I do, but I can only talk about its role in REST, since I don't
believe it plays a (non-redundant) role in the current WSA.  Would you
be interested in text about that, even if it were REST-specific?

MB
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Web architecture consulting, technical reports, evaluation & analysis
Received on Thursday, 9 January 2003 12:48:44 GMT

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