[Bioperl-l] Re: make test

Hilmar Lapp hlapp@gmx.net
Mon, 08 Jan 2001 10:51:16 -0800

Joseph Insana wrote:
> > The warning is intentional, but I agree it looks alarming to anyone
> > installing bioperl. Test code uses a value outside existing positions.
> > Can you think a way of rewriting the test so that it does not print it
> > out?
> Ok, I will change that test not to create the warning.
> But the whole point of that test was to get that warning and see it was
> working as expected.

As I understand from your and Heikki's replies in your test you
wanted the overriding thing to happen, be accepted (even though a
warning was triggered), and the code be able to handle it.

I'm not sure what you did by your change of the test, but it looks
like you simply don't test that feature anymore. If you do want to
keep the warning in the code (and not turn it into an exception,
which means to me that the call itself may indicate an error on
the client side, but in some cases may be totally sensible), what
if you print a message before the test that a warning should be
expected? If you feel confident with removing the warning message,
what if you test afterwards that your code dealt with the
overriding thing as you expected it to do?

Just my two pennies. I didn't want to suggest that anyone turns
off a test of his code. I just think that a warning message being
printed is not really a measurable test result (i.e., it should be
either 'passed' or 'failed').

Hilmar Lapp                                email: hlapp@gmx.net
GNF, San Diego, Ca. 92122                  phone: +1 858 812 1757