W3C home > Mailing lists > Public > www-dom-ts@w3.org > October 2001

Re: Assert functionality

From: Dimitris Dimitriadis <dimitris@ontologicon.com>
Date: Wed, 31 Oct 2001 22:16:26 +0100
Message-Id: <200110312117.f9VLHcn23744@mail.24-7webhosting.com>
Cc: "'www-dom-ts@w3.org'" <www-dom-ts@w3.org>
To: "Arnold, Curt" <Curt.Arnold@hyprotech.com>
Good thing that we seem to have resolved the various licensing issues.


On Wednesday, October 31, 2001, at 09:40 PM, Arnold, Curt wrote:

>> After some clarifications, it is in fact ok to use IBM Public
>> License or GPL softwares for the DOM Test Suites. As long as
>> we only use them, there is no issues.
>
> We are not modifying JUnit, Testlet, JSUnit, Nunit et al and I do not 
> anticipate that we will.  When we have identified problems with one of 
> the frameworks, their maintainers have been willing to
> address the issue in a timely manner.

>> If we need to modify
>> them, then neither the GPL or the IBM Public license would do
>> it. The Apache license is fine. So we don't really need to
>> develop special harness for the DOM Conformance Test Suites
>> in ECMAScript and Java if we don't modify the *Unit
>> harnesses. As Dimitris pointed out, we still need to be sure
>> that the tests can be reused with an other harness.
>
> The Java tests have run (and I'll double check the CVS to make sure 
> that I have put the code out there) with both JUnit and Avalon Testlet, 
> so I feel pretty confident that the test jar could be run
> with any arbitrary test framework after somebody has modified the 
> adapter classes.
>
[dd] Good. This is exactly what we want.

> Using JUnit would be the preferred approach for Java testing, since you 
> would assume that the target audience would be parser developers who 
> want to integrate DOM tests into their build process or
> external developers who want to determine compliance of a particular 
> parser.
>
> JSUnit while usable, isn't ideal for a relative novice who wants to 
> determine how well a browser adheres to the recommendations.  Ideally, 
> they could point their browser at a URL and run the tests
> without any setup.
>
[dd] It is for this reason that there has been a wish for a harness 
online that can serve to run the ECMA tests. Only browsers can be tested 
using this approach, though.
Received on Wednesday, 31 October 2001 16:19:51 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 6 April 2009 12:58:45 GMT