Subject: Re: Competition by internal expertise for F/OSS vendors
From: "Ben Tilly" <btilly@gmail.com>
Date: Wed, 3 Sep 2008 12:26:15 -0700

On Tue, Sep 2, 2008 at 8:44 PM, Thomas Lord <lord@emf.net> wrote:
[...]
> I think that there is a conceptually simple (but socially difficult)
> solution to that problem.   Here is one of the weird ways
> I conceptualize the problem, followed by a proposed solution:
[ ...long essay deleted... ]

This may vary by software community, and probably does, but I simply
don't buy it when it comes to the community I know best.  Which is
Perl.

Let's take the best-known example of a Perl luminary being hired for
that reason.  A number of years ago The Perl Foundation hired Damian
Conway for a year.  They had a number of hopes for this, and Dr.
Conway certainly exceeded expectations.  However I clearly remember
that the way that started was that Schwern noted that as a university
professor Conway got paid about half the going rate for a Perl
programmer in New York City.  Which made him cheap.  Now I don't know
what Conway was paid in the end.  I certainly hope it was more than he
had been making, and I suspect it was because he has continued to make
money doing training seminars.  But I suspect that it was less than I
made that year programming Perl in New York City.

At least in Perl my impression is that the luminaries, if they get
hired specifically to work on Perl, generally don't get paid that
much.  There are a number of well-known people - Mark Jason Dominus,
Randal Schwartz, and brian d foy come to mind, who make their living
doing consulting and running teaching seminars.  For these people the
value of contributing to the community is clear - it certainly serves
an advertising purpose.  A number of key people, for instance Allison
Randal and chromatic, are hired by O'Reilly and are given a lot of
leeway to do open source.  However I can virtually guarantee you that
Tim O'Reilly, idealistic businessman that he is, finds ways to get
full value from them as employees.

So what about the hordes of Perl programmers who contribute their work
for free to Perl or (more often) to CPAN?  If you talk with them I
expect you'll find, as I have, that they have a wide mix of reasons
for what they do.  Many of them find their open source work to be
valuable training, which helps them find better day jobs.  (That has
certainly been the case with me.)  Many do it for fun.  And when I see
that someone has open source work on their resume, I look at it as a
significant plus in a hiring decision.

It may be a selection effect, but I don't have the impression that
people who have contributed to the Perl community, whether those
contributions are large or small, have regretted that contribution.

Now perhaps Perl is different.  Perhaps the fact that there is a
supply of decent jobs for Perl programmers makes that community
different from others.  But I don't think so.  A decade ago ESR used
to ask rooms of programmers how many of them produced software for
shrinkwrap.  And how many of them produced software for internal use
within an organization.  About 90% produced software for internal use
within an organization.  I don't think that ratio has changed.  Which
means that there is probably *NO* effective business model,
proprietary or open source, where most programmers can get paid to to
produce code that large numbers of people will use.  It ain't going to
happen.  That is not to say that there is a shortage of paying
programming jobs.  However, as is the case with Perl, most of them are
going to be for internal consumption within some organization.

Now Tom drew an analogy between open source and the old Hollywood
system.  I would draw an analogy between open source and the
university system.  Most people cannot make money from participating
in academia.  Those who do make money that way have to be doing it as
a labour of love.  But academia produces a lot of things that are
valuable to the outside world, it transmits knowledge and skills
fairly effectively, outsiders value the evidence it provides of
competence, and a great many participants consider themselves to have
been well rewarded for the time they spent in academia.  Even if they
wound up doing something else.

So just because people are being compensated differently than you'd
want does not mean they are not being compensated sufficiently for
what they do.  And I think participation is more likely to wind up
working out well than poorly.

Cheers,
Ben