Xenix/386 and 386 clones

[Joe Bob Willie]

In article <6956 at bigtex.uucp> james at bigtex.UUCP (James Van Artsdalen) writes:
>Is the step D 386 part in distribution yet?  It should be if I recall
>the timetable.  Even the double-sigma parts may be out of date.  Maybe
>an Intel person could enlighten us as to how to tell the chips apart.
>
>I find it interesting that Intel won't replace the chips.  There is
>this thing called "implied warranty of fitness" that applies to
>manufacturers in every state in the US (except maybe Louisiana): it
>would appear that the older 386s fail this test of fitness for the
>claimed usage (ie, mathematical processing).  Has anyone attempted
>to get a part of Intel by threating to file over this?

i was of the impression intel replaced all of the 32 bit multiply bug
chips (or offered to) when they admitted there was a bug in the
silicon.  so it would appear that intel is willing to replace some
chips which they sold as being non-defective.

i heard that the 16-bit-only 386's were sold as 16 bit only parts to
cut-rate hardware houses for use in clones that were never intended
to run native 386 code.  if this is the case, the beef is with the
system builder, not intel.  if the system is built with a 16 bit only
part and advertised as a 32 bit system, i would venture a deceptive
advertising or deceptive trade practices suit is in order.

as for lousisiana and its strange laws, you might have better
protection in louisiana than the rest of the country.  there is a
legal thingy in louisiana called "demand performance" which means
that once you have paid for something and they have accepted the
money, they have to make it work or give you back your money.  this
also extends any implied or limited warrantees to last forever if
the manufacturer knew about the problem when purchased the goods.
-- 
 jfh at rpp386.uucp	(The Beach Bum at The Big "D" Home for Wayward Hackers)

                              "ANSI C: Just say no"
                                                -- Me.