From: "Aaron Ardiri" <[EMAIL PROTECTED]>

> > > 1) Really bad advice!
> > >         A clean Gremlin's run is essential to having quality
> > >         software, and customers do pay for quality software!
> >
> > This is just an assertion. There are lots of things customers want. If
you
> > were to ask one "Do you want me to spend this time fixing a bug that
> you'll
> > never see, or do you want me to add that feature you requested?" you
might
> > be surprised at the answer.

I was interested in what most customers would choose. Given the choice
between: 1) more features, 2) quicker delivery, or 3) fixing an assertion
failure that only occurs under Gremlins, I maintain that #3 would not be
most folks' choice. :)

> codewarrior, pilrc, prc-tools they are tools too *g*.

Well, of course they are. What did you think?

> put yourself in the clients position. they spend $XXXX on developing an
> application that you squeezed the perfect amount of time to get finished,
> but, heck.. gremlins barfs. but, it doesn't matter - when you delivered,
> it worked, right?

Gremlins aren't the only testing facilities. I write unit tests for
"everything that could possible break" (I'm an Extreme Programming advocate,
if it weren't obvious). I don't rely on monkey-bashing to test my code
(though I'm glad to have it).

> six months later, a new device comes out - and, woah.. the application
> they paid good money for no longer works? should the re-contact you to
> fix these problems? surely, thats a good business model for you isn't it?

You're trying to predict the future, and that's not possible.

> > Have you ever heard of a guy spending a week fixing an obscure bug?
> > Has that guy ever been you? :)
>
> yes. and, yes. :)

I maintain that there are situations where that bug should not be fixed.

> a *good* programmer, knows where to look for bugs. it helps having
> access to the os source code (yes, sign up for it).

No duh. How can one make sense of the assertion failures without the OS
source??

> it doesn't take
> long for you to fix such a small obscure bug. i do it daily for all
> my other developers (yes, i am mr. bug fixer in my full time work) :P

Better to teach them to fix their own bugs, it seems to me.

> > In any case, I'm still looking for answer to my question about the
> > severity of the report. Is "Invalid insertion point" a real bug that
> > would affect my app's performance in the real world?
>
> well, it probably isn't *that* severe.

That doesn't answer the question, which indicates to me that the answer
isn't known.

> how about making the height of the field a bit larger and see if the
> message goes away? :)

Please read my messages more carefully.

> but, take this as a lesson, you really shouldn't
> ignore gremlin messages.

Ditto. No one's talking about ignoring anything.


-- 
For information on using the Palm Developer Forums, or to unsubscribe, please see 
http://www.palmos.com/dev/support/forums/

Reply via email to