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

RE: Reflection tests

From: Kris Krueger <krisk@microsoft.com>
Date: Wed, 29 Sep 2010 23:06:09 +0000
To: Aryeh Gregor <Simetrical+w3c@gmail.com>, Anne van Kesteren <annevk@opera.com>, James Graham <jgraham@opera.com>
CC: Maciej Stachowiak <mjs@apple.com>, "public-html-testsuite@w3.org" <public-html-testsuite@w3.org>
Message-ID: <FA9085650D2F8B4A9D501ED9D9706E9A14B196FE@TK5EX14MBXW651.wingroup.windeploy.ntdev.microsoft.com>
The error will go away - give it some time...

-----Original Message-----
From: public-html-testsuite-request@w3.org [mailto:public-html-testsuite-request@w3.org] On Behalf Of Aryeh Gregor
Sent: Wednesday, September 29, 2010 2:58 PM
To: Anne van Kesteren; James Graham
Cc: Maciej Stachowiak; public-html-testsuite@w3.org
Subject: Re: Reflection tests

On Wed, Sep 29, 2010 at 6:40 AM, Anne van Kesteren <annevk@opera.com> wrote:
> (Personally I think it s fine if not all JavaScript tests use the same 
> framework. Especially if it is basically a self-contained test suite.)

I don't think that's fine if an eventual goal is for browsers to integrate the W3C tests into their internal regression testing.  I think this is what's really needed to get a comprehensive set of tests.  At least Gecko and WebKit (maybe IE and Opera too, don't know) have extensive collections of tests for standards compliance, since they require automated tests for all new features.  Like this nice .dataset test from WebKit:

http://trac.webkit.org/export/66582/trunk/LayoutTests/fast/dom/dataset.html


Ideally I think we should settle on a standard API for W3C tests, and get implementers to write standards-based tests to the W3C API instead of their own, and sync the implementer tests to a central W3C repository as much as possible (in both directions).


That doesn't have to block short-term acceptance, of course, but I don't think it's ideal in principle for every test to make up its own API.

On Wed, Sep 29, 2010 at 7:35 AM, James Graham <jgraham@opera.com> wrote:
> This is not a blocker. If we decide to port it to the existing 
> framework, we can do that and make whatever changes are necessary to 
> the framework to accommodate those tests later rather than waiting for it.
>
> . . .
>
> Right, so it is not sensible to approve the tests yet. I don't think 
> this is a big problem either.
>
> . . .
>
> In my opinion, yes. The only question is whether to use 
> tests/Submitted for this or add e.g. tests/InProgress (feel free to 
> bikeshed the name). I am happy with either.

Okay, I tried checking them in, but I'm getting an error:

$ hg push
pushing to https://dvcs.w3.org/hg/html

searching for changes
http authorization required
realm: W3C Mercurial Repository
user: agregor
password:
abort: HTTP Error 500: Internal Server Error

The username and password I used work for <http://www.w3.org/2002/09/wbs/40318/>, unless I mistyped the password four or five times in a row but got it right on the first try for the members' area.


Received on Wednesday, 29 September 2010 23:06:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 29 September 2010 23:06:45 GMT