Re: Fwd: WebIDL Testing Plan: who is doing what (and why) by when?

On 05/14/2013 01:29 PM, Arthur Barstow wrote:
> * Web IDL parser: there are at least two [idlharness.js] and
> [webidl2.js]. Which one should be used for CR testing; has anyone
> committed to maintaining and completing the parser; how is it used
> vis-a-vis the CR exit criteria?

idlharness.js is not a WebIDL parser; it's a script that takes the 
output of webidl2.js and generates tests based on that.

I would argue that we should have idlharness.js test as much as 
possible, because it makes it much easier to test multiple APIs without 
the work duplication involved in manually writing tests for each API. Of 
course, not all requirements in WebIDL can be tested that way, so we'll 
need specific tests as well.

HTH
Ms2ger

Received on Tuesday, 14 May 2013 13:48:47 UTC