On 9/27/07, Rupert Smith <[EMAIL PROTECTED]> wrote:
>
> Rajith,
>
> I we are in agreement with you about the unit tests. The code will benefit
> from pure unit tests, and IMO, some of the dodgier aspects of locking
> AMQSession and the like, would certainly have found better solutions, if
> that code had been unit testable.


Rupert, I felt that too. Sometimes the pain that we go through in writing
these tests upfront can help us a lot down the line.

I'm all for sorting this stuff out, and
> looking a code coverage on unit tests, etc. To give you an example of what
> a
> huge fan of the unit test + coverage approach I am, I have an AI library
> that I wrote, that is 100% covered (but of course still not 100% bug free,
> but getting to that coverage level uncovered so many little bugs it was
> worth it).
>
> Moving the non-unit tests out of the way, will let us see more clearly
> what
> is needed.


I simply made a suggestion. We should all agree about how it is done and the
timing.
I am quite happy about the way this discussion is progressing.

Rupert
>
> On 27/09/2007, Rajith Attapattu <[EMAIL PROTECTED]> wrote:
> >
> > Thank you very much. This kind of cynicism helps this project a lot.
> > I give up trying to explain it anymore. We can definitely continue to go
> > our
> > merry ways.
> >
> > Regards,
> >
> > Rajith
> >
> > On 9/27/07, Robert Godfrey <[EMAIL PROTECTED]> wrote:
> > >
> > > >
> > > > It is a problem as people continue to write and run the wrong kind
> of
> > > > tests
> > > > under client and broker.
> > >
> > >
> > >
> > > I think we should all be grateful when people contribute tests, and if
> > we
> > > believe there are gaps (i.e. in micro-tests of individual emthods)
> then
> > > one
> > > should write them onself rather than complaining that other people
> > haven't
> > > :-) :-)
> > >
> > > -- Rob
> > >
> >
>

Reply via email to