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

RE: [General] Need for a JavaML to DOMTSML transform

From: Arnold, Curt <Curt.Arnold@hyprotech.com>
Date: Thu, 16 Aug 2001 13:57:08 -0600
Message-ID: <70E215722F6AD511820A000103D141D40AA4DC@thor.aeathtl.com>
To: "'Dimitris Dimitriadis'" <dimitris.dimitriadis@improve.se>
Cc: "'www-dom-ts@w3.org'" <www-dom-ts@w3.org>
Sorry, I inadvertantly replied off list.  Dimitris's off-list reply to my off-list reply appears later to provide context for the following comment.

[dd]
>Hopefully we will end up in either b or c, as 
> we've agreed on having all tests in a language-neutral form 
> in order to allow for other bindings to be generated eventually.
> 

[ca]
Option a explicitly states that if we accept tests written in a general purpose programming language we would only use it as a pattern for creating an equivalent DOMTSML test.  However, we wouldn't be
obligated to accept such tests and would only do so if we thought the tests were a valuable resource and it would be better to take them how they are than to not have them.

This far in, I'm definitely committed to the XML based representation.  At the start my preference was test development in general purpose languages first, language independent infrastructure second,
but now that the infrastructure is mostly in place, it makes no sense to try to integrate tests in general purpose languages.

While we are on this subject, haven't heard from the Microsoft guys for a long time.  I assume that they are not the party that you are talking about, but they might have a similar situation, a
substantial body of tests in a general purpose language where requiring the contributor to make the conversion to DOMTSML might be a barrier.

> -----Original Message-----
> From: Dimitris Dimitriadis [mailto:dimitris.dimitriadis@improve.se] 
> Sent: Thursday, August 16, 2001 2:31 PM
> To: 'Arnold, Curt '
> Subject: RE: [General] Need for a JavaML to DOMTSML transform
> 
> 
> (if this wasn't intended to be off-list, feel free to forward 
> it to the DOM TS mailing list)
> 
> I ecpect anyone who has tests in Java to submit them and show 
> us what they have. Also, I think it's plausible that they 
> will conrtibute in writing a transform. 
> 
> In order to find out which of your listed alternatives is the 
> more appropriate one, I've asked them to send a few examples 
> of code to the DOM TS mailing list, then we could take it 
> from there. Hopefully we will end up in either b or c, as 
> we've agreed on having all tests in a language-neutral form 
> in order to allow for other bindings to be generated eventually.
> 
> /Dimitris
> 
> [dd] Good that we agree, we definitely need to be able to 
> convert tests, either from code or from, say JavaML, to our 
> language. Perhaps not trivial, but I can't imagine it's impossible. 
> 
> The request came from someone who has Java tests and can 
> convert them to JavaML, so the question is if we can provide 
> a way to convert them to DOMTSML.
> 
> Is our "little language" [...] "esoteric"? :) Remind me to 
> put that in the documentation.
> 
> [ca]
> Are we agreeing?  I don't foresee the process of conversion 
> being automatic and reproducable, I don't think that we will 
> have enough samples to make anything general.
> 
> If the someone has tests (currently in Java) that they are 
> interested in submitting but want or need help, there are a 
> few options:
> 
> a) they could submit the tests as is in Java.  We could put 
> the tests in the CVS in something like 
> 2001/DOM-Test-Suite/submissions/acmesoft and then proceed to 
> convert them to DOMTSML and put the converted tests in the 
> appropriate tests location.
> 
> b) they show one of their tests and we could show them the 
> corresponding DOMTSML and they could go and convert the rest 
> of their tests and submit them in DOMTSML.
> 
> c) they could convert them on their own and submit them in DOMTSML
> 
Received on Thursday, 16 August 2001 16:00:33 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:34:03 UTC