Date: Mon, 20 Jan 1997 10:52:52 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: jkh@time.cdrom.com (Jordan K. Hubbard) Cc: terry@lambert.org, dennis@etinc.com, hackers@freebsd.org Subject: Re: Commerical applications (was: Development and validation Message-ID: <199701201752.KAA15603@phaeton.artisoft.com> In-Reply-To: <17107.853740593@time.cdrom.com> from "Jordan K. Hubbard" at Jan 19, 97 10:09:53 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> > One of my main talents is the ability to see order in apparently
> > chatotic systems. But further, I see orders of order in these
> > systems. This is what makes me a good systems engineer.
>
> This has also enabled you to pick what was probably the most tediously
> unreadable method of conveying a series of points that it has been my
> displeasure to come across so far this year! Congradulations! ;-)
I admit that following pattern-flow logic requires the ability to
follow patern-flow logic. If you've got it, you've got it; if you
haven't, it's "tediously unreadable". Like Clifford Algebras or
Chebyenchev Polynomials.
> > OBSERVATION: Other OS's have achieved these tasks.
>
> "OBSERVATION: NO OS has acheived these tasks without paying a
> full-time staff, and none of the free OS camps has enough clued-in
> bodies to do even half the things they'd like (and possibly even need)
> to do. Terry is peering into alternate universe again due to local
> field effect of as-yet undetermined nature."
1) Linux has ELF.
2) FreeBSD does not.
3) ELF is desirable
4) Linux is doing something right that FreeBSD isn't.
1) Linux has a large number of willing bodies
2) FreeBSD complains of a dearth of bodies.
3) Allocation of bodies to the projects is based on the
interaction of the social organism with the larger society.
4) Linux is doing something right that FreeBSD isn't.
> California translation: "Awesome. Totally dude. We should, like, do
> something."
>
> > QUESTION: Why is is that the adoption of ELF is categorized
> > as premature, when it works?
>
> Defense lawyer: "Objection!"
>
> Judge: "Yes, Mr. Selachii?"
>
> Selachii: "The actual statement was ``a premature move to elf'', the
> context of which makes it quite clear that any assumption of
> maturity, or lack thereof, refers entirely to the action of
> movement, or in this case the merging of code, rests entirely
> with the speed or pace at which this action is carried out
> and does not, in fact, make any assumptions or claims
> concerning the actual maturity level of the ELF software
> itself."
>
> Judge: "What!?"
Judge: Overruled, Mr. Selachii. During discovery, you agreed
with Prosecution's posit that "ELF was a good thing".
You did not attach conditions then, and I will not
allow you to attach conditions now.
> Since there's little disagreement where it
> comes to security issues or general improvement of the code base, why
> not have people from all segments of "the industry" working together?
There *is* disagreement about general improvement of the code base;
it is in the form of "evolution vs. revolution". The evolutionists
want to take an incremental step to the next stage, because they are
afraid to sail their ship out of sight of the shore. The revolutionists
want faster progress, as they believe that FreeBSD progress has slowed
to the pace of the rest of the industry, and if this keeps up, there
will be no valid reason for the project as an entity seperated from
industry. Other than its own continued momentum (what you call your
"customer base" and what I call your "legacy market").
> > OBSERVATION: The Linux camp has hardware donated by major vendors
> > and other interested parties. So have the MACH,
> > NetBSD, and OpenBSD camps. I have received personal
>
> Cool, please pass on these contacts then. I haven't received any such
> email myself, that's all I can say, and anyone with free hardware they
> want to give away is MORE than welcome to contact me at any time!
> They can even call me collect if they have to! :-)
You miss the point. There is a reason you have not been approached:
you are not approachable, through self-selection. Your lips may
say "yes", but your organizing principles say "no".
[ ... ]
> We've been over that. You refuse to acknowlege that making one person
> change is a lot easier than making 15 of them do so, and you'd prefer
> to turn the iceberg than the ship. Sorry.
It's not an issue of changing people; how many times do I have to
say this? My recognition of the examples of NetBSD/OpenBSD should
*prove* it's not an issue of changing people.
You're missing the point that the social structure, as it exists,
has certain limitations inherent in the process of its operation.
People are members of data sets on which organizational processes
operate. The contents of the data set are irrelevent. If you
were to consider evolution, you are arguing that Methane breathers
could spontaneously arise and thrive in an Oxygen atmosphere. The
organizational processes disagree with you; they hold precedence.
A self-organizing system in a chaotic environment will suffer
increased chaos at its edges as the order in the center increases;
this will occur in inverse proportion to the mathematical order of
the organizational model (the highest order encompassing the whole
of the chaotic environment in which the system is organized). The
lower the order of the organizing principle, the higher the resulting
countervailing force from the environment it is organizing.
Individual molecules of a gas may not be predictable, but the whole
gestalt *will* obey gas laws, allowing the whole to be predictable.
It is not a question of changing people; it is a question of changing
the framework in which they operate. You don't need to change people
to do that; you only need to overcome the inertia or momentum of
the component members who act to sustain and perpetuate the existing
framework.
> Yadda yadda yadda. Go back to business class again and pay special
> attention to the part where they talk about *small* business models
> and making different tradeoffs at different times in the company
> lifecycle, from 2-guys-in-a-garage to multinational corporate empire.
Yadda yadda yadda. So you want to be a small business forever? Then
continue to organize and react as a small business.
I don't know about you, but I've personally worked in industry for
nearly 14 years, and I've help take an organizational O(1) company
("2-guys-in-a-garage") and turn it into an organizational O(2) company
("22-guys-in-an-office") that could never make O(3) because the O(0)
entrepeneur couldn't change his thinking, and O(n) thinking is
incapable of scaling past O(n+2), no matter what controls you attempt
to impose.
I've also worked in an O(5) company (my current employer) and an O(6)
company that just could not make it to O(7) because they were
unwilling to accept the effect of economies of scale on their work
model, and change the model accordingly (Novell). I tend to
recognize the places I've worked when I see them again. Linux is
an O(4). FreeBSD is an O(3) that is an O(4) wanna-be.
The largest organization in the Internet implementation space is O(4);
that's Linux. For comparison, Microsoft is O(7).
> I've known more than a few garage companies who were too proud to make
> any sort of concessions in order to win an initial customer base, and
> you know what? They're still in their garages, wearing dirty
> tee-shirts and railing at the industry for failing to recognise their
> genius! :-)
Take that up a fractal order (social, not organizational), and you have
any group of people producing a UNIX or UNIX-clone system, including
FreeBSD. FreeBSD is too "proud" (I would claim ossified, but we'll use
"proud") to make the structural concessions in order to win a
large-as-Linux market share.
Take that up another fractal order, and you have the UNIX community
as a whole, who have adopted standards as a means of obtaining
short term competitive advantage rather than as a means of
commoditizing themselves into homogeneous components of the larger
whole. That's why a pure-POSIX system isn't sufficient as a UNIX
platform, and why the POSIX standard is *intentionally* deficient
in this way. The UNIX community as a whole is too "proud" (I would
say it was placing O(n-1) goals ahead of O(n) goals, but again,
we'll use "proud") to allow it to take the long view and provide a
unified front.
Let me end with a question: What is the current prioritized list of
goals for the FreeBSD project?
Regards,
Terry Lambert
terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199701201752.KAA15603>