W3C home > Mailing lists > Public > www-ws-desc@w3.org > November 2004

Re: Summary, 9-11 Nov 2004 WS Description WG FTF: two objections

From: Amelia A Lewis <alewis@tibco.com>
Date: Thu, 18 Nov 2004 12:42:46 -0500
To: www-ws-desc@w3.org
Message-id: <20041118124246.0d8e9d08.alewis@tibco.com>

Following up to myself.  How tastelessly egotistical ....

On Wed, 17 Nov 2004 13:07:09 -0500
Amelia A Lewis <alewis@tibco.com> wrote:
> On Tue, 16 Nov 2004 12:51:25 -0800
> Jonathan Marsh <jmarsh@microsoft.com> wrote:
> >   Issue LC59a
> >     RESOLUTION: In-Optional-Out and Out-Optional-In will be marked
> >     at
> > risk 
> >                 when entering CR and will be removed unless we see 2
> >                 
> >                 interoperable implementations
> 
> Objection: TIBCO Software, Inc., requests that this be modified to
> "All MEPs are to be marked as at risk when entering CR and will be
> removed unless we see 2 interoperable implementations."  This MUST
> include in-out, among others.  Only if this is done will anyone bother
> to define"interoperable implementation" in a testable fashion, as the
> client/server computing crowd will have their toys perfectly protected
> in advance.

TIBCO won't lay down in the road if any particular MEPs are removed, so
long as the mechanism for defining new MEPs remains as it is now, with
no preference given to the MEPs published in part two or used in part
three (apart from namespacing and likelihood of widespread
implementation, sigh).

I'd like to suggest a rule for whether a MEP ought to be included or
dropped.  This is, in fact, something like the less formal measure that
the WG adopted some time ago, when it became more open to additional MEP
definitions.

A MEP should be included in part two after CR if:

1) some other specification makes reference to it
2) that specification puports to establish interoperability measures

This is fairly clear, and any of the MEPs referenced in part three
automatically get in under these rules.  So would the MEPs specifically
targeted in LC59f, if, for instance, TIBCO Software were to publish a
binding specification using out-optional-in, with guidelines for
measuring interoperability (well, in-optional-out is another story, but
someone else is the narrator there).

Amy!
-- 
Amelia A. Lewis
Senior Architect
TIBCO/Extensibility, Inc.
alewis@tibco.com
Received on Thursday, 18 November 2004 17:43:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:33 GMT