RE: Purposeful Parity control

["Bruce Young" <bayoung@bigfoot.com> (by way of Daniel Weaver <dan.weaver@znyx.com>)]

As one of the original members of the PCI development team at Intel,
I would like to comment on the discussion of why we put parity
into the PCI spec. 

Politics!

We actually had a fairly large debate about what kind of error 
detection/correction we should include with most of us falling into
one of two camps. One camp wanted full single bit error correction 
with double bit detection. In other words a full robust error 
mechanism. The other camp wanted nothing at all given the PC systems
that we were targeting. The minimalist camp basicly won out with
only a single parity bit thrown into the spec to give the 
appearance of having some error detection. Of course I have always 
subscribed to Seymour Cray's opinion -- "Parity is for Farmers!"
(You probably need to be from the rural Midwest region of the USA to 
fully appreciate the quote but Cray did not believe that Parity
was very useful)

-Bruce Young