some $0.02 considerations related to tagging " assertions" vs.
writing " test assertions" :
(a plea in favor of more than just
tagging...)
Regards,
-Jacques
As I suggested in a recent post, I don't believe that it is always
necessary to "reword" spec requirements in order to turn them into test
assertions. If the spec is written in appropriately precise and normative
language, it ought to be possible to treat text within the spec itself as
assertions.
in some
cases, the spec may appear crystal-clear on what needs to be tested and gives
enough hints on how to verify the requirement But shouldn't we still
recommend to extract a test assertion for the sake of consistency with the rest
of TAs that are associated with this spec?
I am aware
that this is more a methodology argument than a technical one, but my concern
would be: often what is " crystal clear" for spec writers, may not be so
for test suite writers, and spec writers may be prone to be content with "
assertions" tags where a little more structured " test
assertion" would really be more helpful.
For example
WS-I " profile definitions" already seemed to be stated quite
clearly (each requirement is a separate, numbered item etc.), but when we
wrote TAs for these in a methodical way (with a test approach in mind, e.g.
stating more clearly which were the artifacts under test, etc.), we were
surprised of how non-obvious that exercise was, of the questioning and
challenges this exercise caused to the original statements (helping increase
quality) and ultimately of the difference it made for people writing ultimate
test cases.
A data-point:
when the W3C QA Working Group was working on the Specification Guidelines [1] we
wanted to make sure that this document (which was itself a specification)
conformed to itself (that is, that this document conformed to the
recommendations contained within it).
" eat
your own dog food" as some say... I guess we' ll have to, for our
credibility...
Since one of our recommendations was to use
test assertions, we made the effort to create an assertion list. It turned out
to be a largely cosmetic exercise, involving making slight but non-critical
changes to text that was already present within the spec.
This is why we
defined a test assertion as "a measurable or testable statement of behavior,
action, or condition [that is] contained within or derived from the
specification's requirements". (In retrospect, I think "the specification's
requirements" is redundant, and we could simply have said "the
specification".)
[1] QA Framework:
Specification Guidelines:
http://www.w3.org/TR/qaframe-spec/Patrick
Curran
Manager, Java SE Conformance Test Development, Sun Microsystems
--------------------------
Durand, Jacques R. wrote:
Serm:
Ah, one more lurker coming out in the open :-) Thanks for detailed
feedback - that reinforces the need to make sure we all agree on the
terminology. So far I see agreement on this list on the general
intuitive meaning of TA: a TA is a rewording of a spec requirement in a
way that is "testable" within an assumed mode of deployment for an
implementation.
Divergences seem to be in the details:
- Test Assertion (TA) is in fact Test Requirement (TR) plus Test
Profile
I must confess that I never really made a difference between TA and
TR... from a modeling viewpoint, isn't this "test profile" matching
something like a conformance profile? i.e. grouping a selected set of
TRs. In that case no real difference TA/TR. Or, do you see TR as closer
to spec narrative, and TA possibly narrower, designed for a specific
conformance level or profile (e.g. may cover only partially some spec
requirement, if the "profile" does not need more)?
On the scoping of a TAG TC activity:
- I also favor a "descriptive" guideline (and data model I guess) as a
decent first objective, that has value in itself and less controversial
for broad adoption.
- even if all agree on such a 1st step/deliverable, we still need to
decide whether the TC charter should plan for [a possible] next phase
that involves an adjunct for a mark-up / representation of TAs. (note:
charter does not need to be precise on which purpose this mark-up would
serve... could be just wrapping the descriptive for publishing)
Cheers,
-Jacques
-----Original Message-----
From: Serm Kulvatunyou [mailto:serm@nist.gov]
Sent: Saturday, December 02, 2006 4:00 PM
To: tag-discuss@lists.oasis-open.org
Subject: RE: [tag-discuss] Re: Test case metadata...
Hi, have been following the discussion for a while.
Yes. There are a lot of terminology mismatches around here.
A number of us on this list are familiar with the terminologies in OASIS
ebXML IIC Test Framework spec and the OASIS conformance definition.
I saw that the W3C published note has different terminology,
particularly of
confusion here is the "test case metadata" (on the W3C note web page, I
think, it is called just "test metadata". I will use the term test case
metadata in this message since it has been used in this discussion
group).
Having read some software testing literatures, the terminology in the
IIC
specification seems more common.
My understanding so far as having read all the messages in the TAG
discussion is as follows.
- Test Assertion (TA) is in fact Test Requirement (TR) plus Test Profile
(test profile from the functional module perspective, not from usage
profile
perspective) in IIC, and maybe a few more additions (e.g., relationships
b/w
TRs). Note that TA is an element of the Test Requirement. In the IIC
specs,
only descriptive info is contained these entities, which is, perhaps,
why a
number of us cannot quite grasp the idea of auto-generate something from
here including test cases or implementation code.
- In the W3C note on test case metadata, there seems to be overlaps into
the
Test Case as defined by the IIC. This is because some elements in the
test
case metadata are defined to be machine processable (e.g., the Inputs
and
Expected Results. On this basis, perhaps it is why some of the list
participants are thinking of some automations based on the TA, if it
includes such elements. Note that in IIC, a test case contains
everything
from relationship to the TR, machine interpretable test execution steps,
inputs, and verification conditions/expected result, and more.
My opinion is as follows:
- The big question is how far does a TA or TR spec goes, whether
including
just descriptive info or also machine interpretable instructions.
- I think if the (business) objective of the TC is to create something
that
a lot of Standard TC can chew on, the TA or TR spec should include only
descriptive info, i.e., it should not be much different from those
already
included in the IIC spec for TR. And 'yes' as one of Dave Marson's email
suggested, let the test expert and/or implementer with the help of the
target standard expert translates that into executable TC (as defined by
the
IIC). I personally think that such TA or TR spec is useful and of
sufficient
value b/c although it is unlikely to provide value in automation in
subsequent steps it is divide the target standard into atomic
features/functions that make test coverage analysis (matrix as termed by
Dave Pawson) more clear. In addition, it would definitely make test case
writing easier as well.
- I think for standards that are very well defined via some sort of
formal
models, autogenerating TAs/TRs as well as TCs maybe at least partially
possible. But I think that should not be the subject of concern for this
TC
(it should be an interest of research labs). The produce of this TC
would
provide a standard medium for the auto-generation algorithm to publish
result. Ie. This TC focus can be only defining TA/TR data structure such
that it facilitates test coverage analysis, test case authoring, and
understanding of functions within the target standards.
I have witnessed values and how people appreciate the availability of
such
descriptive TA/TR.
Having said that I would recommend that we used the terms defined in the
IIC
spec for discussion (in which case I think we should have used the term
TR
in place of TA), and the length of work for this TC shouldn't be more
than 6
months.
Thanks for reading this long message...I think I deserve one since I
have
read tons of messages before sending one :).
Serm Kulvatunyou
National Institute of Standards and Technology
-----Original Message-----
From: Dave Pawson [mailto:dave.pawson@gmail.com]
Sent: Friday, December 01, 2006 8:59 AM
To: tag-discuss@lists.oasis-open.org
Subject: Re: [tag-discuss] Re: Test case metadata...
The terminology issues is very rife round here.
Please could the TC take some of the jargon and perhaps develop a
glossary.
Simple statement.
My definitions are contra those of David M's and I find it hard
to comprehend his terminology.
regards