An oldie but goodie...

From: Paul Koning <pkoning_at_equallogic.com>
Date: Thu Feb 26 13:12:22 2004

>>>>> "John" == John Allain <allain_at_panix.com> writes:

>> The MMJ connector is an example of the
 John> Seemed like a successful product, back then. What's been
 John> Olsened' with it?

The DECserver-100 -- the first LAT terminal server from DEC -- was
almost ready for release. It came in a fairly small box, about the
same dimensions as two fairly large laptops stacked on top of each
other. It had terminal connectors with snap latches on the side.

Ken came in at the last minute and decreed that (a) it had to go into
a larger box, and (b) it had to use different connectors, the MMJ
connectors. The result was that the initial release of that product
came in a large box with a small PC board inside, and 8 jumper cables
to connect from the original terminal connectors on the board to the
MMJ jacks in the new cabinet. Needless to say, this caused a
significant product delay as well as some highly annoyed engineers.

As a result, Ken was at times referred to as "the chief connector
architect".

This was not by any means the only example of Ken meddling in the
detailed design engineering of a project. That's perhaps
understandable in a startup (but you have to do it at the right time);
it makes no sense in a 50,000 person company.

>> bad habit of setting up several competing projects inside

 John> FWIW, internal competition is practiced at all market leading
 John> companies that I've seen, IBM, Seiko, HP, to name a few.

 John> Last minute changes as dictated by a team of one sound like a
 John> recipie for disaster, if that in fact was what happened. The
 John> only person I heard of getting burned in DEC was Cutler, but
 John> there must have been others.

There were whole competing organizations: the networking products
group in Littleton vs. the Low End networking products group in
Maynard. Those ended up both shipping product, which wasted resources
(money) and confused customers. Various VAXen had the same issue, for
example the 8800 vs. a watercooled variant thereof, or the 8800
vs. the next lower series (7000? 6000? I forgot).

Cutler's case may have been a bit different, though he did end up in
at least one major project that went a long way and was then
canceled. That was the second RISC project at DEC -- it was the THIRD
that finally made it to product release (the Alpha). In the meantime,
so much time was wasted that MIPS had to be brought in to save the
day.

>> ...where's Ken? {from another questioner}

 John> Not dead, just totally retired. I don't think his obituary
 John> would've made it past me. Sounds like from just this short
 John> transaction that the man has some enemies. Too bad, I'd like
 John> to hear him talk.

I hope that "enemy" wasn't applied to my comments. I have no personal
problems with Ken, I don't know him well enough and what little
contact I had with him was certainly pleasant. My issues are around
management inability. Not that his successor was any better, just the
opposite actually; he actually set out to dismantle the company and
succeeded, while the worst Ken ever did was to botch up what should
have been a lot more successful.

     paul
Received on Thu Feb 26 2004 - 13:12:22 GMT