On 28 Jul 2005 at 18:25, dhbailey wrote:

> Darcy James Argue wrote:
> 
> > I can't speak to their in-house testing, but installers often get
> > short shrift.  (They shouldn't, because then stuff like this
> > happens!)
> > 
> > Beta testers were given an installer to test, but it did not have
> > this problem.
> > 
> 
> Why would a company give their beta testers one installer and then
> ship with an untested one?  . . .

I'm sure they tested the shipping installer. But the kind of problem 
exhibited is one that is not going to be detected on most testbed 
machines.

> . . . Who's in charge over there?  Change
> something, and then test the darn thing.  It's not rocket science to
> see if something works, even though it might be rocket science to fix
> a problem when it shows up.

Test results are only as good as the test.

I would bet my next paycheck that MakeMusic *did* test the installer, 
and several times, but they didn't examine the areas in which the bug 
turned out to happen.

That's a description of EVERY BUG IN EVERY PIECE OF SOFTWARE THAT'S 
EVER EXISTED. The bugs you find, you fix (insofar as you're able). 
It's the bugs you don't find that bite you, and you can only find 
them by adequate testing, and the test has to be the *right* test.

Unfortunately, sometimes the "test" that reveals the bug doesn't 
emerge until the code is installed by the users.

I would actually blame the makers of the installer, myself, as an 
installer shouldn't allow the removal of existing folders without 
taking some kind of special action to make that happen.

-- 
David W. Fenton                        http://www.bway.net/~dfenton
David Fenton Associates                http://www.bway.net/~dfassoc

_______________________________________________
Finale mailing list
[email protected]
http://lists.shsu.edu/mailman/listinfo/finale

Reply via email to