- From: Doug Schepers <schepers@w3.org>
- Date: Thu, 12 Jan 2012 11:06:34 -0500
- To: Arthur Barstow <art.barstow@nokia.com>
- CC: ext Glen Shires <gshires@google.com>, olli@pettay.fi, public-webapps@w3.org, public-xg-htmlspeech@w3.org, Dan Burnett <dburnett@voxeo.com>, ext Satish S <satish@google.com>, Peter Beverloo <peter@chromium.org>
Hi, folks- On 1/11/12 9:40 AM, Arthur Barstow wrote: > On 1/10/12 11:25 AM, ext Glen Shires wrote: >> Per #4 Testing commitment(s): can you elaborate on what you would like >> to see at this point? > > At this point, I think a `warm fuzzy` like "if/when the spec advances to > Candidate Recommendation, we will contribute to a test suite that is > sufficient to exit the CR" would be useful. I agree with this general sentiment, but I'd like to offer a different priority to tests. Technically, the W3C Process does not require a test suite, but pragmatically, it's the best way to indicate conformance and implementability, and to promote interoperability. Modern expectations (e.g. in the last 4-6 years) about the specification process include early prototyping and implementation feedback well before CR phase, with real-world webapps depending upon these early implementations, so we need interoperability fairly early on. The infrastructure and methodology for creating and maintaining tests (at W3C and in implementation projects) has improved dramatically in the last few years as well, so it's easier to do. As such, the creation of tests should not be left to CR... there should be a plan in place (e.g. a person, and a loose policy, like "as we implement, we'll make tests and contribute them to the WG"), and a person responsible for collecting and maintaining the tests (i.e. making sure that the tests are adapted to meet the changing spec). Regards- -Doug
Received on Thursday, 12 January 2012 16:11:02 UTC