Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!decwrl!ucbvax!EMUNIX.EMICH.EDU!stretch
From: stretch@EMUNIX.EMICH.EDU (Brian Stretch)
Newsgroups: comp.sys.ibm.pc.hardware
Subject: 386 problems kinda resolved
Message-ID: <9008112237.AA08032@emunix.emich.edu>
Date: 11 Aug 90 22:37:36 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Lines: 30

Welp, after 1-1/2 months digging, I have finally figured out the problems
with my pre-production Intel 386 motherboard and Intel's policies towards it. 

It seems that the early 80386 chips (B-Step?) had a serious bug.. which,
amoung other things, prevented the use of the 386 Protected mode.  Basically,
this bug makes it impossible to run 386-specific software.  Intel had a
policy in place to replace those chips--free--for three years.  All you had
to do is know about it.  Many (most?) dealers ignored the problems, tho, as
there wasn't any/much 386-specific software at the time, and didn't bother
upgrading their stocks.  Now that 386-specific software is popular, these
B-Step chips are popping up all over the place--alas, over a year after the
upgrade policy expired.  One third-party company does have a hardware fix--
for $150--but for that price, I'd just assume ditch the motherboard and
replace it with a new 20Mhz 386SX. 

What lessons have been learned here? 

 o. If you beg, plead, nag, and kiss ass long enough, you can get good info
    out of Intel.  But it'll be too late to do you any good. 

 o. Intel, as a company, is a disorganized mess. 

 o. Pre-production boards can be a royal pain.. altho my problem is present
    in several thousand full production chips. 

 o. It's no wonder why several major PC software developers are happy as
    hell to be supporting a certain new non-Intel-based multitasking,
    multiuser machine that'll be hitting the markets in the next few months.
    (Stay tuned on that one, folks..)
    (Okay, more reasons than the above for that, but..)