• From: Michael Rys <mrys@m...>
• To: Gary Stephenson <garys@i...>, xml-dev@l...
• Date: Tue, 19 Jun 2001 08:32:10 -0700

A specification normally has a formal foundation that covers the
semantics much more than a test suite. Thus, defining the final
recommendation as a sequence of tests would be a really bad idea.
However, it would be a good idea to require that the implementations
presented as proofs to go to proposed recommendation go through a test
suite to show the requirements to go out of the candidate recommendation
phase...

Best regards
Michael (speaking for myself)

> -----Original Message-----
> From: Gary Stephenson [mailto:garys@i...] 
> Sent: Tuesday, June 19, 2001 3:28 AM
> To: xml-dev@l...
> Subject: Re: XML 1.0 Conformance Test Results
> 
> 
> ----- Original Message -----
> From: "Francis Norton" <francis@r...>
> 
> > [2] Specs precede implementation better than test packs do - it is 
> > easier to define a complete spec than it is to define a 
> complete test 
> > suite, and many test questions cannot reasonably be 
> anticipated unless 
> > you make assumptions about how the application will be implemented.
> 
> 
> "Completeness", as Goedel informed us, is an impossible goal. 
>  (We might navigate by the stars without actually hoping to 
> reach them.)
> 
> Certainly the spec has to precede the test suite - but given 
> that there is a well-defined process that a spec has to be 
> frog-marched through to reach "Recommendation" status, why 
> could it not reasonably be expected that the final 
> recommendation should be expressed as a sequence of 
> executable tests in some agreed upon language (presumably 
> Java at this stage).  If, as I am led to believe, a spec can 
> not actually attain W3C recommendation status unless and 
> until it has actually been implemented, then that should 
> mandate that there be a valid test suite to determine whether 
> it truly _has_ been implemented or not.
> 
> my  $A0.014  (and falling)
> 
> gary
> 
> 
> 
> >
> 
> 
> ------------------------------------------------------------------
> The xml-dev list is sponsored by XML.org, an initiative of 
> OASIS <http://www.oasis-open.org>
> 
> The list archives are at 
> http://lists.xml.org/archives/xml-dev/
> 
> To unsubscribe from this elist send a message with the single 
> word "unsubscribe" in the body to: xml-dev-request@l...
> 

• Prev by Date: RE: almost four years ago....
• Next by Date: Re: xml and z39.50 protocol
• Previous by thread: Re: XML 1.0 Conformance Test Results
• Next by thread: XML Schema: is it possible to get a list of ALL the errors in an XML?
• Index(es):
  • Date
  • Thread