On 07 June 2001, Chuck Esterbrook said:
> At 03:13 PM 6/7/2001 -0400, Andrew Kuchling wrote:
> >I'd like to avoid a replay of the unittest situation, where
> >an inferior solution wins out because no one is aware of our code.
>
> What was the unittest situation? You had a better version in Quixote or
> some other project that got passed on (for the reason you stated above)?

Andrew is referring to a testing framework that I hacked together for
internal use by the MEMS Exchange.  (Neil subsequently hacked it up some
more to gather code coverage data.)  It was origially inspired by PyUnit
-- I think I read the PyUnit mini-paper from IPC9, thought, "Hey, that
sound like a good start -- but I'd do this differently, and add that
feature, and ...", and started coding.

It's an undocumented mess (it doesn't even have a name apart from
unittest.py, which is unfortunate since that's the name that PyUnit has
in the standard library now!), but it has worked quite well for us for
the last year-and-a-bit.  We made a half-hearted attempt to foist it
onto the rest of the Python community (even arguing for it in Guido's
living room!), but never really made our case strongly.  Never got
around to cleaning up the code, documenting it, making a real release
(it was squirreled away in the test directory of the first two Quixote
releases, but Quixote doesn't use it for anyhting), or any of that.

So now PyUnit is in the standard library.  I don't think that's a *bad*
thing, it's just that my impression of PyUnit is that it's not quite as
convenient to use as our unittest.py.  Maybe it needs to steal some of
our good ideas, just as I stole the basic idea from PyUnit in the first
place.

        Greg
--
Greg Ward - software developer                gward@mems-exchange.org
MEMS Exchange                            http://www.mems-exchange.org