RFC soon on essay "Does Free Software Production in a Bazaar obey the Law of Diminishing Returns?"

Jacques Chester thunda at manor.downunder.net.au
Wed Aug 11 01:45:09 UTC 1999


Hello there;

I'm sending this email to a number of people. The majority of you
have already been asked in the past about what I am going to
describe. Namely, I have asked most of you to perform a
peer review of an essay about economics and free software.

Some of you have *not* been asked before. This may have
been because I have not thought before to contact you, because
I have been unable to reach you, or because you are receiving
this message on a public forum. If this irks you or if you feel I
have unnecessarily wasted your time, I would like to apologise,
and invite you to email me saying "don't email again, dammit!" :)

As a student in the International Baccalaureate program, I am
required to complete an "Extended Essay"; a mini-thesis of sorts.
The body of this essay is limited to 4000 words. I dithered around
with a number of topics in planning for this essay, many of them
wishy-washy and poorly defined. I've arrived at a point where I
am combining two of my great interests - Economics and Free
Software - into a clearly defined, quantitative research paper.

The topic of this essay is "Does Free Software Production in a Bazaar
obey the Law of Diminishing Returns?". Essentially, the essay seeks to
establish a chain of reasoning that links some of Eric S. Raymond's
qualitative hypotheses with standard Economic theory to allow
for quantitative testing.

First, Brook's Law is introduced. Through comparison of statements,
analogical comparison and graphical comparison, Brook's Law is
taken to be *equivalent* to the Law of Diminishing Returns. In
Economics, the LODR has a very specific meaning, and has been
thoroughly analysed. There are key traits which are easy to identify
in appropriate sourse of data, which makes it an ideal subject of
study in relation to OSS.

Secondly, Raymond's Bazaar hypothesis is introduced. The Bazaar
itself is outlined, a list of "signature traits" (which may be
incomplete or incorrect) was extracted from the C&B paper. Raymond's
assertion that a project following a Bazaar methodology can break
Brook's Law is introduced. It is explicitly reasoned that this means,
in turn, that Raymond's hypothesis implies that Bazaar projects can
break the LODR.

Thirdly, the GNOME project is identified as the source of data. It is
established as being a large-scale Bazaar-style project.

Fourthly, data extracted from GNOME's CVS system is used to generate
a series of graphs for analysis. These graphs can help to demonstrate 
the possible conclusions.

Fifthly, the possible conclusions so far are:
* That ESR is completely correct, that Free Software *does* break the 
  LODR and that it represents a new economic phenomenon in production
* That ESR is completely wrong, that Free Software *does* obey the
  LODR, it just happens to be far more 'scaleable' than the traditional
  "Cathedral"
* That ESR, unaware of the economic framework in which this can be
  viewed, does not realise that Brook's Law is the same as the LODR;
  and has therefore failed to distinguish between production in the
  Short Run (where the LODR applies) and the Long Run (where it does
  not).

On a preliminary viewing of the data, there is significant ambiguity.
As I complete my set of graphs in the next few days, I may find that
one or the other alternatives is true (or indeed, come to a completely
unexpected conclusion), but I suspect that the ambiguity will remain
because the GNOME project is too small to test the limits of
Bazaar scalability.

My own personal view - and this cannot yet be backed by numbers, so
caveat emptor - is that the second and third conclusions are true. This
is because marginal costs *do* exist in Free Software, because they
*do* rise steadily. Also, the LODR does not apply in Long-Run
situations, but frankly, defining the difference between the Short and
Long Run in Free Software is difficult, as it relies on the mapping
of the Economic categories of Land, Labour, Capital and Enterprise
to the resources being manipulated. These issues will be shunted into
an appendix for deeper discussion, in order to help reinforce the
necessary economic framework.

So what does this have to do with *you*?

Well, if you've read *this* far without snoozing, thanks :).
I feel that as well as helping me complete my course, this
paper will act as a first step towards more rigorous, solid and
dependable research into the free software movement. In
particular, it will do a first, rudimentary test of ESR's
wholly qualitative theories. Several of those assisting me
or who have agreed to review this work expressed interest on
those grounds alone!

What I have done so far is to brief you on the essay itself.
You now know what to expect, so hopefully you will have some
time for the core ideas to bounce back and forth in your
heads. I also suspect that you will be prepared to answer
many of my claims and chains of reason with good critiques.

I am hoping that, when you receive my "beta draft" next
week, you will be prepared to examine it well, without having
to both get the gist of it *and* analyse it all at once.

If I may? A wishlist :)

I do not expect everyone to perform the same critique, and
in fact I think it would be sad if this were to happen.
But I would really like to see certain things commented on:

* The link between Brook's Law and the LODR.
* The identification of Bazaars
* The use of the GNOME project
* The collection, use, and displaying of data.

In particular, I am preparing a "caveats and issues"
appendix to accompany the final documents, a lot of which
I hope to source from my reviewers and the community at
large.

If you feel like you're up to critiquing the larger part
of a 4000-word essay, along with some of its appendices
(some of them substantial in themselves), I'd really,
*really* appreciate your help. If not, please email and
ask not to receive any more mail from me. I know it sounds
spam-esque, but there you have it!

I hope to email the essay out this weekend, Australian
Central Standard Time (Darwin, in your clock lists).
I can only allow 7-8 days for your replies, so if you
cannot offer even a basic reply in this time, please
feel free to disregard the work.

The final work will be sent to the International
Baccalaurate Organisation for their assessment. After it
has been assessed by them and I have graduated, I will
post a final draft to the net. I will *not*, repeat,
*not* post a final draft, or the final submitted version.
The IBO does a text-search of the internet for matches to
any work it assesses as part of a standard procedure, and
I do not want my assessment to be delayed as a consequence
of an allegation of plagiarism.

Once it has been assessed however, and pending permission
from the IBO, I will publish a final copy online, possibly
under the OPL.

Thank you all very much in advance, and thanks for your
help;

JC.




More information about the License-discuss mailing list