Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!hobyah.cc.uq.oz.au!bunyip.cc.uq.oz.au!munnari.OZ.AU!spool.mu.edu!howland.reston.ans.net!vixen.cso.uiuc.edu!newsfeed.internetmci.com!in2.uu.net!news.artisoft.com!usenet
From: Terry Lambert <terry@lambert.org>
Newsgroups: comp.os.linux.development.system,comp.unix.bsd.386bsd.misc,comp.unix.bsd.bsdi.misc,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc,comp.os.linux.advocacy
Subject: Re: Historic Opportunity facing Free Unix (was Re: The Lai/Baker paper, benchmarks, and the world of free UNIX)
Date: Fri, 26 Apr 1996 14:10:22 -0700
Organization: Me
Lines: 210
Message-ID: <31813BBE.467CAA89@lambert.org>
References: <4ki055$60l@Radon.Stanford.EDU> <4lfm6m$d2c@rigel.tm.informatik.uni-frankfurt.de> <4lftn3$olb@solaria.cc.gatech.edu> <317C8F0F.307E@curtin.edu.au> <4lgo4c$6fa@solaria.cc.gatech.edu> <317E07F1.441D@curtin.edu.au> <317DAA16.42C86E59@lambert.org> <4lqpu1$j7q@news.missouri.edu>
NNTP-Posting-Host: hecate.artisoft.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.01 (X11; I; Linux 1.1.76 i486)
Xref: euryale.cc.adfa.oz.au comp.os.linux.development.system:22479 comp.unix.bsd.386bsd.misc:864 comp.unix.bsd.bsdi.misc:3553 comp.unix.bsd.netbsd.misc:3409 comp.unix.bsd.freebsd.misc:18279 comp.os.linux.advocacy:47310
Justin Rhys Thuryn McNutt wrote:
] Terry Lambert (terry@lambert.org) wrote:
]
] : ] Great warrior doesn't fight: he sits on his dorstep waiting
] : ] for the body of his enemy to be carried down the street.
] : ] Tao Te Ching
] : ] (not an exact citation: I read it long ago and in
] : ] different language :-)
]
] : An inappropriate quote... it refers to the wisdom of defeating
] : your enemies by outliving them, instead of placing yourself
] : at risk. If your enemies fight and you do not, you are all
] : the more likely to outlive them.
] :
] : What is your estimate of the life expenctancy of Microsoft?
]
] Perhaps not entirely. Unix has been around longer than Microsoft,
] and really, with the way Microsoft is going, I think Unix will
] easily outlast it. In my opinion, Microsoft is going to get
] too big for its breeches in an awful hurry, and then implode
] when product quality goes to hell (more than usual), tech.
] support costs run them into the floor, and they spiral downward
] into zero.
Entropic death of Microsoft predicted.
8-).
Follows: a rough mathematical model predicting company growth
potentials at each stage based on management and organizational
models. Novell and Microsoft SEC filings used as examples.
=========
Companies deriving from an entrepeneurial process have growth
"management thresholds" which they need to cross; as a gross
oversimplification, we have:
1) Startup (~1-3 people)
Person decides to start company
2) Proprietership/partnership (~3-8 people)
Owners wallet is company wallet; each dime spent
is felt leaving personal wallet.
3) Small business (~6-18 people)
Control is still centralized; main characterizing
factor is micromanagement (finger in all the pies).
[
Aside: these numbers come from minimal to maximal
strain on ability to assert personal control.
They are based on the entrepeneur having a strong
sense of identification with the business as "self".
]
4) Business (~16-50 people)
Control has been delegated to trusted lieutenants;
main characterising factor is control of lieutenants
(finger in all the lieutenants pies).
At this time, hire/fire may or may not devolve to
the lieutenants; usually it is a single lieutentant
(evolution of presonel department).
A small business generally becomes a business
when incorporation goes through. Typically, the
lieutenanting process is the result of outsourcing
accounting and the realization that some control
can be safely given away.
5) Small Company (~45-350 people)
Lieutentants are trusted. They now have submanagers
(no central direct reports). Lieutentnants may or may
not delegate hire/fire .
[
Aside: where I'm getting my numbers now (4, 5, 6)
According to a Bell Labs research study, humans can
keep, at most, 5 to 9 items in active memory at once
(hence phone numbers are 7 digits). Ignoring the
ramifications for user interface design (;-)), this
range is based on management depth multiplication:
7*7*7 = 343... minimum 125, maximum 729 for three
tiers of control.
The 16-50 (7*7 = 49) comes from minimum 25, maximum 81.
]
6) Company (~250-2500)
Divisions. Division heads have lieutenants. Overall
view of lower eschelons rarely changes from "Company"
as far as the central figure (if still present).
[
Limits: 525/2401/6561
]
7) Big company (~1500-16500)
Because of personal quality control breakdown, you will
rarely see an average order 7 "level of management";
it is statistically improbable. An order 6 limit is
7776 at this level. Consider: My manager + 6 managees
equals 7 items to juggle.
According to Novell's form 10k filing on their web page,
they had 7272 employees. This is consistent with a big
company order of 5.92.
http://corp.novell.com/strategy/10k/10-k95.htm
If you look at their form 10-Q for January 1995 (also
on their web site), they had 7808 employess, or an
order of 6.005.
http://corp.novell.com/strategy/10q/10-qjan.htm
So, they are hovering at near the computed expectation
value for a Big Company.
Novell is a Big Company.
[
Limits: 3125/16807/59049
]
===
So what is there after "big company"?
There is Microsoft.
According to Microsoft's 10k filing, they employ 17801 people.
http://www.microsoft.com/msft/html/item1.htm
You *could* have a well managed (order 7.08) big company and
have that many people (Novell, if it tightened it's management
quality controls could basically double their growth expectation).
I don't think that's what's going on. I think Bill is more
astute than that.
*I* think that Microsoft has transitioned to the next stage of the
game:
[
Calculation is for *one* holding depth:
Limits O1/O9: 15625/117649/531441
]
The next stage of the game is "holding company". It's where a
company realizes that it doesn't have to centralize all its
infrastructure resources to be profititable. Consider: if you
owned all of the companies producing PC's today, you would be
making money. It doesn't really matter that you would be
paying seperate personnel departments or seperate accounting
departments, etc.... you *would* make money.
Novell is currently struggling with this limit. They want to
be a bigger company, but they are loathe to face the pseudo
loss of control and the false ineconomy of non-centralized
infrastructure. So they attempt things like putting all of
their people at one location to increase out-of-band
communications, etc..
But every time Novell absorbs a new company, they reduce their
total headcount back into the 6.0 index range.
With skilled management (which I think Bob Frankenbur may be),
Novell as the growth potential to double. There are also
potential "spinoff" companies, which can be used to drop their
limit.
Microsoft, on the other hand, either has *zero* growth potential
(and damn fine management, from the top all the way to the
bottom, in every category -- *very* had to believe), or it has
a growth potential to go up by a factor of five (assuming it
doesn't increase its holding depth or use spinoffs or increased
management skill requirements.
===
Conclusion:
Entropic death of Microsoft unlikely.
You will have to find another way of beating them, because
outliving them just isn't going to work. 8-).
Regards,
Terry Lambert
terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.