- From: Andy Seaborne <andy.seaborne@talis.com>
- Date: Mon, 28 Jun 2010 18:49:11 +0100
- To: Axel Polleres <axel.polleres@deri.org>, SPARQL Working Group WG <public-rdf-dawg@w3.org>
On 28/06/2010 17:48, Axel Polleres wrote: > 3) Go through test cases proposals... I contected all owners of respective actions individually, it would be extremly helpful to have some example queries with expected answers per category at the least before the call... > 4) Discuss responsibles for the test suite... My proposal would be to have one overall editor who is supported by "category responsibles" at least to start with? Last time we checked "working" tests cases into W3C CVS to collect the material. "working" = manifest linking to parsable queries and data. We had test cases for proposals (so they didn't parse or execute) in the early stages of DAWG but we had syntactically correct queries and data for the later stages. This all helps who ever is the overal test case editor and coordinator. If they end up translating/trasnscribing test cases from email then the test case may not reflect the original intent and also it could be quite a burden on the coordinator. Better to crowd source in the correct formats. It also helps if questions about the spec go to the email list separately to avoid point getting missed or test case material and discussion getting mixed and something missed as a result. Once we get into the flow of creating test materials, this should avoid too many bottlenecks but we do need to make sure we are distributing the work somehow. Andy
Received on Monday, 28 June 2010 17:49:45 UTC