Re: About WARNING of ibischk2

>A model can pass the parser and still get WARNINGS. That does
>not mean that the model is **Bad**. We have devices whose V/I
>characterisitcs are non-monotonic. So, the parser will ALWAYS
>give you WARNINGS. I do agree that understanding the WARNINGS
>is a good thing to do.
>
>..and if the model never passes the parser...throw it away.
>
>Regards,
>Syed
>National Semiconductor Corp.

I never suggested that a model that has warnings/errors is
bad. Only that such models require more study to understand
the reasons for the warnings/errors. And this is, in my
opinion, "not a desired thing". It is at worst, an issue of
model accuracy and at best, an issue of model confidence.

 -- Paul Gregory

   phone: (208) 396-5086 USmail: Hewlett-Packard
     fax: (208) 396-4122 M/S 143
   email: paul_gregory@hp.com 11311 Chinden Blvd.
                                               Boise, ID 83714
Received on Mon Apr 6 16:36:05 1998