[comp.os.misc] More on the Modest proposal

eugene@eos.UUCP (Eugene Miya) (02/25/89)

A Modest Proposal, Part II.

Some years ago.  I worked on a somewhat peculiar architecture. It had
a rather small address space for a big machine.  They had an OS
tuned for their machine.  Right...  Sure....  When I moved to Ames,
where they did not have one of these machines, sales people from said
company came by.  "Most the other NASA Centers have some of our machines.
We would like to place one at Ames" as if this was justification
enough to buy one of their machines.  3 people showed up for a presentation
in a room which held 200 people.  An embarassment to them.  Of course,
they are now trying to make it into the Unix market.  Now do these people know
how to sell machines into our market?  Do we want these people
"in the fold?"

Yes, I admit, I get Datamation (from my IBM days).  The current issue is on
OS/2 versus Unix.  Do you remember when it was VMS versus Unix?  Or MVS
or VS2 versus Unix? Or OS/32 versus Unix [you must have worked at "research"
because I don't ;-)].?  I think it was Muuss who pointed out at one of the
Santa Monica Unix User Group meetings that if you wanted to see where MS/DOS
would go, just look toward the U world.  This was before DOS had directores,
etc.

Consider that rather than openly attack our competition, we try to
subvert them like the Afgans.
We could make a more malicious attack on our competitiors.  We could go
on the offensive to "crush our enemies (remember Conan?)."  Rather than
take arrows from them.  We just subvert them in a different way.
We should adopt hit and run tactics by we burdening them.

Our enemies (the loyal opposition?) fall into three basic categories:
competing hardware manufacturers, competing universities running
different OSes, competing applications users: lab versus lab, site versus site.
How do we insure the good survive while pruning the bad?
The question is who should we target first?  They probably don't
read the net.  We should start small.  Put some small computer company
out of business, shut them out of business, place their people into
unemployment.

Maybe we can do the same to some of our competing National Laboratories?
Empire building to the max?  Get rid of one of these other networks
(DECnet: SPAN or HEPnet) in favor of TCP?

How should we do it?
Consider if we request features, fixes which are totally bogus.
Have them waste their manpower.
Many of our sites have highly heterogeneous hardware.
So do the sites of our competitors.  Consider what we could do to
those sites if we burdened them with work?  Make them busy for unused
features.  Maybe we could bring down one of the weaker mainframe
manufacturers?  Meanwhile their regular business falls off.

We shut out universities and departments which aren't aligned with
us by not hiring their grads ("Sorry you have experience with THAT
non-Unix OS.").

Consider EDT running on a 3270, or SPF running on a vt52. (Oh that's
a different religious argument, sorry!).  Owls, H-P 2600 series,
Tek 4010 series.  We do it as separate but coordinated burdens.
We have to watch out for lookalikes (like vt100 clones).  That only
benefits them.  Make certain they never think of anything like termcap.

Ask for ports of COBOL or PL/1 (compilers) to different machines.  Do it
in such a way that they don't think to write a common back end but rather
write a completely new compiler for every machine. ("No, that's
not efficient enough!")

We can create phony requests for information.  Set up demos in vacant lots
which are never attended.  Like spy organizations.  Flood them.
Make certain the amount of work we put into this does not exceed the
waste of time on our side. ;-)

After a while they would get suspicious and start filtering requests.
Considering the economic and human effort to expended filter.
They would have to evaluate whether there is a serious need for
utilities ports.

We could never bring down an elephant as large as IBM: they have far too
many programmers for all of us to keep them busy.  It has been said it will
collapse under its own weight.

So what computer companies and end users don't have network access?
Who do we target?  Maybe some small manufacturer of 16-bit minis
somewhere (and I don't mean DEC, they're too big to start on).  And
DEC are friendlies right?  And maybe we'll do a squirrel, or two
while poisoning pigeons in the park.

I am reasonably content editing 40 MB data files using vi or emacs
on my Cray.  Even if I have to share it with 200 users. 8-)

Me? I side with the FSF.

Another gross generalization from

--eugene miya, NASA Ames Research Center, eugene@aurora.arc.nasa.gov
  resident cynic at the Rock of Ages Home for Retired Hackers:
  "Mailers?! HA!", "If my mail does not reach you, please accept my apology."
  {uunet,hplabs,ncar,decwrl,allegra,tektronix}!ames!aurora!eugene
  "Post follow ups.  Contribute to network noise."

The pattern is the same. Flash a few lights, interfere with their machines,
disrupt their way of life.  Soon they will be at each other's throats.

--Monsters are due on Maple Street.