FW: test assertions guidelines under review (OASIS)

["Jacques R. Durand" <JDurand@us.fujitsu.com>]

 Forwarding this comment.

-----Original Message-----
From: marcosscaceres@gmail.com [mailto:marcosscaceres@gmail.com] On
Behalf Of Marcos Caceres
Sent: Wednesday, May 12, 2010 9:01 AM
To: Jacques R. Durand
Cc: dom@w3.org; liam@w3.org; dennis.hamilton@acm.org
Subject: Re: test assertions guidelines under review (OASIS)

Hi Jacques,

I've only just had a chance to review the Test Assertions Model (I will
try to review the other documents when I find a bit more time).
Overall, I found the model to be consistant with our own document (and
experience). It was also good to see the model captured as a class
structure and diagrammed - it makes everything quite clear. I also found
the document quite accessible and easy to read/understand.

A minor comment:

Section 3.1

"This unique identifier facilitates tools development and the mapping of
assertions to specification statements. It is recommended that the
identifier be made universally unique.1"

This implies a UUID. People without experience may end up choosing
complicated identifiers which can become difficult for people to work
with. I think it should be made clear that the universality of the
identifier should revolve around the document that contains the test
assertion (e.g., via a URI and fragment idetifier:
http://example.com/tests#ta-5)


Marcos Caceres
Opera Software ASA, http://www.opera.com/ http://datadriven.com.au