Licensing Issues Aren't DEC's Fault

[Casey Leedom]

  Ok, Yoseff.  A very nice response to a very inflammatory article from
me.  You deserve accolades for your restraint and discretion.  I'll try
to follow suit and keep this discussion above boards.

> In article <51997 at felix.UUCP> (Yoseff Francus) francus at pernod.dec.com writes:
> 
> Just because other companies choose to ignore their license from AT&T
> does not make it right for DEC to do so.  So don't start knocking DEC
> about the licensing issue because DEC follows its license, and everyone
> else doesn't.  How would Sun and all the other companies react if someone
> violated their license with their products. 
> 
> As regarding pricing/performance issues I'll be happy to take this
> off-line with you. Please remember that MIPS is a somewhat bogus
> benchmark, and there are other things to consider besides the box and the
> operating system (i.e. support).

  First, I'd be glad to take this off-line, but I think the issues are
relevant to this group.  As I said above, I promise to try to keep my
comments less inflammatory.

  You're right that N wrongs don't make a right, but AT&T has chosen not
to pursue their license on this point.  That's their problem till they
do, and when they do it will be everyone's problem.  But till that point,
I really wish DEC would spend more time working on Real Issues rather
than random points like these.  I hate to harp, but the point is telling:
4.3BSD has been out for over two years now; WHEN IS ULTRIX GOING TO
UPGRADE?

  As for hardware price/performance vs. support ...

  In my view the initial software that comes with the system and the
fundamental reliability of the hardware are the most important items of
support a company can offer for their system.  This is much more
important than the performance of the hardware.  If the software is bad
or inadequate to begin with, or the hardware won't stay up, no amount of
wonderful hand holding from a company or incredible CPU or I/O speed is
going to make life any easier.

  On going hardware and software support is of course also important since
no significant piece of hardware or software is without bugs no matter
how good it is.  I would rate price/performance and third party support
as least important (within reason).

  And I think that's basically what most people look for in a system:
base quality of hardware and software; on going hardware and software
support; and finally ``MIPS/$'' [sic] and availability of third party
support.

  I think it's pretty obvious to almost anyone on the in that DEC is
falling down pretty badly in that last category.  Their systems are very
low performance and high priced and they're doing their best to eliminate
third party support by making their busses proprietary and only licensing
to companies producing non-competing products.  But since those are the
least important features of a system, so what?

  And that's exactly what it would be (apart from grumblings about how
slow things were), except: first, we really use a linear function of our
parameters, not strict priority to measure systems, and, second, there
are companies out there that are gaining reputations for good
hardware/software systems and fair ongoing hardware and software
support.  And worse (for DEC), while not many would complain about the
reliability of their hardware and I view DEC's on going hardware support
as some of the best, their software (Ultrix) just isn't keeping up and
DEC's on going software support is poor at best.

  My current feeling is that DEC needs to shake up the Ultrix team and
revamp the team's priorities.  DEC could also learn something from Sun on
software support (as well Sun could learn a lot from DEC in hardware
support).

  They also need to give up on the proprietary nature of their busses.  I
think a large reason for the success of the PDP-11 and early VAX line was
the availability of third party products.  DEC's current policy is a
misguided attempt to garner 100% of the market for their machines.  You'd
think that DEC would learn from recent marketing history which seems to
show that open systems are more successful than closed.

  Finally they need to retake some performance ground.  The low
performance and incredibly high prices are embarrassing.

  I've heard rumors from a number of sources that these are all concerns
within DEC, but there seems to be some difficulty in getting off the dime.
Don't get me wrong please, I'm taking issue with DEC from a purely
technical basis.  I have nothing personal against DEC and I would in fact
welcome them back to the field of real contenders.  It's to my advantage
as a customer to have as many companies as possible producing good
products.

  In any case, I've dribbled on long enough here.  Sorry to waste your
screen bandwidth ... :-)

Casey