Ant,
In my opinion, not checking that you're getting the right exceptions will be a false economy and
will simply come back to bite. Especially as one of the test designers is taking part in all this.
I have fixed several bugs in Tuscany when I examined the exceptions actually coming back from the
runtime where they were clearly the wrong ones. In some cases, I extended the tests to catch the
underlying problem....
Yours, Mike.
ant elder wrote:
On Wed, Aug 12, 2009 at 12:34 AM, Raymond Feng<[email protected]> wrote:
The purpose of this exercise is to make sure that Tuscany behaves as the
specification says under various situations.
Thats the same as what Simon said previously about "accurately
reflects the content of the SCA specifications", I don't think that
should be the purpose of this exercise. Not to say we shouldn't also
continue to do that as well, but its a much bigger task and will take
a lot more work, and if thats what we focus on it could be well in to
next year before we can say we're passing the conformance tests.
IMHO it would be good to work as a team to fix the otest that are
failing first, but its Apache so we're all free to work on what we
please so if others want to first look at the ones passing for the
wrong reasons then would a reasonable compromise be to not "activate"
any new Tuscany specific error validation to cause new tests to fail
until the Tuscany problem is fixed? That way it wont scupper any
attempt by some of us to get all the tests passing sooner.
...ant
