- From: Philippe Le Hegaret <plh@w3.org>
- Date: Mon, 19 May 2014 23:41:41 -0400
- To: public-test-infra <public-test-infra@w3.org>
- Cc: Erika Doyle Navara <Erika.Doyle@microsoft.com>
On 28/04/14 16:19, Philippe Le Hegaret wrote: > I note that it is now mid-May :) Besides the question from Erika that Robin addressed, I didn't see any reaction or concern. To complete the answer from Robin on coverage, I would add that the current thinking is to use a combination of techniques to establish the coverage, such as: 1- what amount of the browser code isn't triggered by the tests? 2- how many tests do we have for each html element, DOM function, etc? 3- Use metadata that comes on top of WPT For 1, we would need to get the information from the browser vednors themselves. For 2, the difficulty is to figure what the scripts are doing in the tests. For 3, there is a high cost maintenance and I think we would need to figure the purpose and value of the information first before investing in it. At best, all those methods would be used to get us closer to a valuable answer. Other ideas to establishing the coverage without paying a cost maintenance are welcome. Philippe
Received on Tuesday, 20 May 2014 03:41:52 UTC