W3C home > Mailing lists > Public > public-html-testsuite@w3.org > August 2010

RE: Philip Taylor's Canvas Tests - call for feedback (fallback, type, size)

From: Kris Krueger <krisk@microsoft.com>
Date: Thu, 5 Aug 2010 05:13:45 +0000
To: Philip Taylor <pjt47@cam.ac.uk>
CC: "'public-html-testsuite@w3.org' (public-html-testsuite@w3.org)" <public-html-testsuite@w3.org>
Message-ID: <FA9085650D2F8B4A9D501ED9D9706E9A12603E04@TK5EX14MBXW651.wingroup.windeploy.ntdev.microsoft.com>
At the last conf call we agreed that the current status of WebIDL indicates that indeed this should not be delete able.
We'll have to watch to see if the WebIDL spec changes, if it does we can create a bugzilla bug and change this testcase.
Though at this time this test and all the fallback, size and type tests have been 'approved'.

See -> http://test.w3.org/html/tests/approved/canvas/
Run them at http://test.w3.org/html/tests/harness/harness.htm

-----Original Message-----
From: Philip Taylor [mailto:pjt47@cam.ac.uk] 
Sent: Tuesday, July 27, 2010 2:49 AM
To: Kris Krueger
Cc: 'public-html-testsuite@w3.org' (public-html-testsuite@w3.org)
Subject: Re: Philip Taylor's Canvas Tests - call for feedback (fallback, type, size)

Kris Krueger wrote:
> OK my only feedback on these first set of tests is for 
> http://test.w3.org/html/tests/submission/PhilipTaylor/canvas/type.dele
> te.html  It's really testing the WebIDL draft, that should align to 
> ECMA5 eventually.  Which in theory will make DOM object constructors 
> behave the same as native JS builtin objects like function, object, 
> regex.
> So then DOM objects can be deleted just like JS objects.

Currently HTML5 normatively references WebIDL for the definition of what its IDL means. If WebIDL changes in the future then we can easily update the tests to match, but it seems unlikely it'll change very soon, and in the meantime I think we should test what the specs say now.

If the specs conflict with what implementers are willing to implement, hopefully the failing tests will remind them to put more resources into defining what the specs ought to say instead.

But it looks like all the browsers match the current spec/test anyway, so the behaviour is interoperable and there's no problem now.

Philip Taylor
Received on Thursday, 5 August 2010 05:14:21 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:14:29 UTC