Pseudo-code Tests in the specification

[David Nuescheler <david@day.com>]

Dear TC members,

as we discussed during the f2f we are not going to provide a tck
(technology compatibilty kit).

If I remember correctly though we agreed based on Ethan's suggestion
to include almost
pseudo code testcases in the specification. Given that I think this is
something that is crucial
do avoid ambiguities in the specification I would like to trigger the
conversation on this very
subject.

I think there are a number of questions around this topic that need to
be addressed.
What would such a "in-line test" look like formally"? ...and also to
what level of granularity
would it go? Do you think we could get an example/draft of a testcase
so we could
see how this works in practical life and the TC could discuss it?

If I remember correctly the intention was to have a large number of
these pseudo-code tests
in the specification. I definitely agree that having a lot of tests
definitely helps clarifying
the specification. But that also means that we should practice
producing those tests rather
earlier than later...

Would such test examples be something that we could include in the .52 draft?

regards,
david

-- 
Visit: http://dev.day.com/