- From: 傈辆全 <hollobit@etri.re.kr>
- Date: Fri, 27 May 2016 18:00:14 +0000
- To: Philipp Hoschka <ph@w3.org>
- CC: Dave Raggett <dsr@w3.org>, "Kovatsch, Matthias" <matthias.kovatsch@siemens.com>, Takuki Kamiya <tkamiya@us.fujitsu.com>, Public Web of Things IG <public-wot-ig@w3.org>, "J. Alan Bird" <abird@w3.org>
Thanks Philipp, 2016. 5. 27. 20:35 Philipp Hoschka <ph@w3.org> 累己: > [jumping in after having read Jonathan's latest mails on this - I think there's been some confusion] > >> On 5/25/2016 10:30 AM, Dave Raggett wrote: >> Just to note that while WGs are required to produce test suites and to publish implementation reports for transitioning from Candidate Recommendation to Proposed Recommendation, WGs do not normally work on interoperability testing. > > They do - the "rule of thumb" is that the implementation report has to show "two interoperable implementations of each feature" in a spec - that's "interoperability testing" (at least to some extent) > That's what I want to said point. We should carefully use the terminology - "interoperability testing". And also, we should clearly define its scope in the WG / IG charters. > What WGs really don't work on is *conformance* testing - because that requires a much more extensive test suite > >> The idea of the test suites is more oriented towards demonstrating that the specifications are implementable, and thus the test suites are required to cover all of the testable features in a spec. > > That's all correct and part of the demonstration of 'interoperable implementations" for each feature > >> This is why the diagram shows the WG producing specs and test suites. > Best Regards, --- Jonathan Jeon
Received on Friday, 27 May 2016 18:00:58 UTC