Subject: Re: *precisely* a commons (with tragedy)
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: 04 Jan 2002 19:11:19 +0900

>>>>> "Tom" == Tom Lord <lord@regexps.com> writes:

    Tom> A software engineering infrastructure is a consumable
    Tom> resource.

    Tom> A software engineering infrastructure is a renewable
    Tom> resource.

I don't understand what you're thinking about here.  Infrastructure
normally connotes a capital asset.  Your example (which basically
boils down to renaming "maintainer labor" to "patch bandwidth")
doesn't help much.  All you really did is to rename "management" as
"infrastructure", which surely can't be your intention.

So an infrastructure may depreciate but is not (as such) directly
consumed.  The infrastructure may help to coordinate services
("maintainership"), but it is the service, not the infrastructure,
that is consumed.  Engineers are not cannibals who _eat_ maintainers;
they use their time.  The maintainer survives to evaluate patches for
8 hours (or whatever) again tomorrow.  :-)  The infrastructure is the
mailing list/Bugzilla/whatever that embodies the patch queue;
maintainership is ... just labor.

Also, normally in economics a _renewable_ resource connotes an
exhaustible resource with a _negative depreciation rate_ (ie, positive
"fertility") over some range of stocks (eg, a deep-sea fishery).
However, you seem to be talking about "infrastructure" as a produced
good.  I don't care if you want to use "renewable" in that way.  But at
the same time, you avoid using words like "producible", and so seem to
want to define "renewable" in some third way.  What is it?

    Tom> Unfortunately, the predominant FSB practice has been to
    Tom> exploit the public infrastructure without making a
    Tom> correspondingly large investment in its renewal.  They make
    Tom> some investment, but not enough.

Investment implies the right to extract a return.  FS by definition
forgoes that.  We would like FSBs to contribute more, but you can't
really call it an investment.  The investment part is to construct the
tools they need.  Then they GPL and publish them.  Doing any more than
that is not investment, it's fiduciary irresponsibility.  :-)

Yes, I understand where you're coming from.  There _is_ an aspect of
social investment here, because "society" by definition captures all
returns.  The problem is, VCs don't care about _social_ investment.

    Tom> What would we predict that an over-grazed public software
    Tom> engineering infrastructure would look like?

20-20 hindsight, based on a really dubious analogy of software
_development_ to resource _extraction_.  For example, your "patch
bandwidth" example doesn't fit any of the examples of over-grazing
that you give.  (Well, it's sort of the inverse of "post-apocalypse
gang warfare".)

I don't think that your infrastructure is what is overgrazed in the
examples you give.  It's the ideas (application software) themselves.
That's where the "foraging" comes in.

I think the infrastructure is actually in pretty good shape.  The
tools that support developers (compilers, editors, OS kernels, web
servers, application libraries, etc) are awfully good bang per buck,
both in terms of customer payments (median about $0) and in terms of
developer resources used to create them.  The projects that maintain
them tend to have pretty good management and good processes.

It's going beyond that to develop revenue-generators that isn't
working.  But that's not a problem of lack of infrastructure, as far
as I can see.  It simply a lack of (venture) capital to pay developer
salaries and other _project-specific_ expenses.  But contributing to
OSS development is not an investment, since there is no way (in known
models) to capture enough of the benefits as a revenue stream to pay
back the venture capitalists.

    Tom> FSBs have proven the efficiency of open source processes,
    Tom> especially surrounding the kernels and web servers.  Those
    Tom> parts of the software infrastructure are holding their own at
    Tom> the moment and do receive good levels of renewal from FSB.

Do they?  The web servers seem to be supported by companies with 9
toes firmly planted in various kinds of proprietary markets.  They are
not supported by FSB revenues AFAICS.  The kernels, I dunno, maybe.
Depends on how much development Red Hat is doing, and how much revenue
wasabisystems is getting.

    Tom> I think the niche successes of FSBs justify very large
    Tom> injections of cash now

I see a fair amount of cash injection heading in the direction of free
software.  All the big Japanese computer companies are scrambling to
put together Linux divisions, for example.  (What they'll do with
them, I dunno, but they're hiring more than most Japanese employers
are these days.)  Just not into free software businesses.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.