W3C home > Mailing lists > Public > public-mwts@w3.org > December 2009

Re: Write up of test assertion extraction methodology

From: Marcos Caceres <marcosc@opera.com>
Date: Mon, 14 Dec 2009 18:18:44 +0100
Message-ID: <4B267374.1020703@opera.com>
To: Dmitri Silaev <Dmitri.Silaev@Sun.COM>
CC: Dominique Hazael-Massieux <dom@w3.org>, public-mwts@w3.org, Robin Berjon <robin@robineko.com>

Dmitri Silaev wrote:
> Hi Dom,
> Here is my comments for "Extracting Test Assertions from a
> Specification" document. The document looks very good. I have only minor
> comments.
> I. "3. Mark-up conventions for test assertions"
> "Each assertion is uniquely identified through the id attribute on the
> paragraph element; the unique identifier starts by convention with ta-,
> and its uniqueness is ensured by the HTML validity requirements of the
> document."
> Is it useful to explain the best practice how to get the assertion the
> unique assertion id? For example, "its uniqueness generated randomly is
> ensured..." It saves reader from solving the riddle what "RRZxvvTFHx"
> means.

Agreed. We need to change this to just use one or two characters or a 
number; Something easy to use... assigning the long random string was a 
big mistake on my part, which caused a few headaches.

In the Widget Interface spec, I've been working with short identifiers: 
just two letters. This makes life a lot easier.

> II. "4. Extracting automatically test assertions"
> "The original extraction of test assertions was made through an XSLT
> style sheet, that allowed to generate a static list of test assertions
> that served as the first basis for the review of the testability of the
> specification."
> I believe it need to be explained where the extraction was made, e.g.
> "When "Widgets Packaging and Configuration" specification was marked up
> by convention said before, the original extraction of test assertions
> was made through an XSLT style sheet,..."

I'll just note that the only reason we didn't end up using it in the end 
was just a matter of preference (I personally don't know XSLT and get 
very confused every time I try to use it. Coming from front-end web dev, 
I'm more at home with CSS Selectors which is the reason why we remade 
the extraction using JQuery... actually, I believe my original version 
of the extraction tool used the Selectors API ).

> III. "5. Test assertions and test cases"
> "To maintain the association between test cases and test assertions, a
> simple XML file was set up:
> It seems the example should match to marked assertion in "3. Mark-up
> conventions for test assertions". It simplifies the understanding of the
> document.
> VI. "5. Test assertions and test cases"
> "its content is integrated in the test plan with JavaScript to attach
> test cases to the previously extracted test assertions;"
> It may be useful to add link to [WIDGETS-TESTS] one more time here.
> V. I've found strange symbols in document: "—", ... It may be not a
> problem in final document, but now it made a little difficulties to read
> document for me. I've used "download" link from cvs page:
> http://dev.w3.org/cvsweb/~checkout~/2008/dev-ind-testing/extracting-test-assertions-pub.html?rev=1.1&content-type=text/html;%20charset=iso-8859-1
> Thanks,
> Dmitri.
> On Dec 10, 2009, at 12:53 PM, Dominique Hazael-Massieux wrote:
>> Le mardi 08 décembre 2009 à 11:09 +0100, Dominique Hazael-Massieux a
>> écrit :
>>> Here is (attached) what the document would like once made publication
>>> rules-ready.
>> Based on Marcos’ kind offer to take an editing pass at the document,
>> I’ve put in CVS at:
>> http://dev.w3.org/cvsweb/2008/dev-ind-testing/extracting-test-assertions-pub.html
>> Dom
Received on Monday, 14 December 2009 17:19:21 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:08:03 UTC