Subj : Re: W3C DOM ECMAscript test suite errors -- already reported?
To   : Bob Clary
From : Ian Graham
Date : Fri May 27 2005 10:33 pm

Bob Clary wrote:
> Brendan Eich wrote:
> 
>> Ian Graham wrote:
>>
>>> I recently ran the W3C's DOM ECMAscript (JSUnit-based) test suite on 
>>> Mozilla 1.7.x and Deer Park Alpha 1, and noted a bunch of failures / 
>>> errors.
>>>
>>> Are these already logged in Bugzilla somewhere? I couldn't find any , 
>>> but figured they must be there somewhere...
>>
>>
>>
>>
>> I believe Bob Clary has done the same.  Bcc'ing him to spare his email 
>> some spam, but I bet he will followup here.
>>
> 
> Yes, I had actually written a reply to his post but after I realized he 
> was using a munged email decided it wasn't worth it.
> 
> We have a number of bugs filed on issues which are related to 
> conformance to the various DOM Specs. You can find them (along with alot 
> of others) by querying bugzilla for Core:DOM*. Unfortunately we don't 
> have an easy way to flag which bugs are strictly related to conformance 
> issues.
> 
> Regarding the DOM TS, be sure you are using an up to date cvs version of 
> the W3 DOM TS. In addition, there are many errors in the W3 DOM TS which 
> cause us to either fail tests or experience errors depending on the 
> loader and the mime type used. Before filing new bugs, please make sure 
> that a) a bug hasn't already been filed and b) that the test itself 
> isn't wrong.
> 
> I have a partial list of failures of Mozilla Firefox (Deer Park) from 
> yesterday for DOM 1/2 Core/HTML/Events for each of the loader types and 
> mime types if anyone is interested. It is not complete due to bug 
> 292717, but is a good starting point. In the near future I will be 
> making full test results for the DOM TS available on a regular basis.
> 
> I would also recommend that anyone interested in filing bugs against 
> Mozilla (Firefox)? get jst and others in the DOM Community involved to 
> see what they would like done.
> 
> Bob

Thanks -- it;s a bit disconcerting to have to verify that a conformance 
test is accurate, but I somehow suspected that might be the case.  Bob 
-- is your partial list of failures available online somewhere??

.