Subject: Re: Returns to service professionals (was Re: New ESR paper: The Magic Cauldron)
From: "Tim O'Reilly" <>
Date: Fri, 02 Jul 1999 08:04:09 -0700

Sorry to be coming back in after the better part of a week...this
thread may have been beaten to death, and I'm just catching up
(and haven't yet seen all the messages, since I'm reading on a
place messages from a last-online connect on Tuesday.):

On Mon, 28 Jun 1999, DJ Delorie <> wrote:

> > I fear that these companies consider their applications as their
> > competitive advantage.
> Why is that a fear?  *Every* company has to have a competitive
> advantage.  Companies like AOL provide a service, and that service is
> managed through software.  Giving away their software also gives away
> their service and their advantage.  They'd be nuts to do that.

Agreed.  Obviously, at O'Reilly we don't make everything free
either.  I have no problem with companies keeping things for
competitive advantage.  My concern is that they don't necessarily
get the balance right.  Companies like Cisco have, for instance,
been extremely proprietary, but they've also been big supporters
of the IETF and open networking standards, because they realize
just how closely their proprietary business success is entwined
with the continuation of the open internet.  My worry is that
some of the internet content companies don't realize how
important it is to continue to support the communities and style
of development from which the technologies they depend on have
> Now, I'm not saying that *all* their software is part of that service.
> For example, I wrote some software at Cabletron (which is a software
> company, sort of) that didn't affect their products nor their profits.
> *That* would be a candidate for GPL'ing (in fact, they're considering
> it).  But I wouldn't expect them to open the CGIs they use on their
> e-comm site; those are part of their competitive advantage.
> But, of course, getting such companies to help the community in
> *other* ways is a good thing.  I just don't expect them to give away
> the company jewels.

> > I guess I see the Internet still in one of the "robber baron"
> > periods, where everyone thinks of what they can take out of the
> > ecosystem--after all, it's free, a land of opportunity--and
> > doesn't worry about the long-term consequences.
> Not quite a valid analogy, since the "robber barons" hurt their
> victims (and the economy) directly, whereas copying software doesn't
> detract from the original one whit.  The ratio of users to
> contributors is pretty high these days; expecting all users to
> contribute is folly.  There are zillions (meaning I don't know how
> many ;) DJGPP users; very few actually contribute, but who cares?
> Those that want to contribute do.  Those who want to have a say do.
> Those who are content to quietly accept the status quo don't bother
> us.  Life goes on.

I'm not suggesting that the problem is a variation of "people use
but don't contribute."  I'm suggesting that the problem is that
people are building things that are going to be extremely
important in the future, things that are effectively proprietary
extensions to open standards.  Right now, they may be doing it
unconsciously.  Later, they may end up doing it consciously, but
in any case, chances are good that we'll need another free
software foundation to reinvent stuff that is being done now. 

I'm talking about things like data interchange formats or
"business interaction API's" for particular industries.  A good
example of this came up the other day when talking to our
webmaster.  We supply content to all of the online booksellers: 
cover art, descriptions, tables of contents, indexes, sample
chapters, price, etc.

Each online bookstore has its own format for accepting this
data.  Allen Noren, our guy, is trying to get some communication
going, so this becomes more standardized.  But already Amazon has
so much market power that the question is "do we do it the right
way, or the amazon way."  Amazon hasn't done anything to suppress
alternatives, but neither have they created a framework in which
people can fix "bugs" in their process.  It's effectively "closed

What I'd love to see is some conscious work to make this kind of
problem more susceptible to open source efforts.  For example, if
Amazon published an explicit API, and worked with other online
bookstores to agree on it, people could be writing software that
streamlined the process.   So the question for Amazon, as for
others, is where do they benefit by opening up some of this stuff
more to the outside world.

For example, right now, we can't get them all of our indexes or
some sample chapters, because their current upload process barfs
on files of more than a particular size.  

Sure, they could fix it, but if the code they used was something
other people could work on, we could fix it for them...
> It's one of the things that the masses don't understand about open
> source - if someone wants to use the software privately (or doesn't
> want to use it at all), you can just ignore them (unless you have
> visions of grandeur, of course).  What's the difference between AOL
> not opening their software, and AOL not existing?  In neither case do
> they contribute directly to the community, but their mere existence at
> least brings more people into the possibilities of contributing.  Net
> gain just because more people are on the Internet, if nothing else.

I don't argue with this point at all.  It's really orthogonal to
the concern I was trying to raise.  I apologize if the lack of
precision in my original posting raised a red herring.
> > Think strip mining, clear-cutting, commercial farming with heavy
> > use of chemicals, etc.
> Another bad analogy.  Copying software doesn't leave the original
> author with less software, nor does it prevent the original author
> from continuing with their original development plans.  Your examples
> are exclusive of other options for that land.  I don't mind analogies,
> but please pick better ones.

Actually, I think this is a very good analogy, in that it gets
across the idea that I really do believe that a lot of companies
have, that the internet as they now encounter it, is a limitless
resource, when in fact it was created by a cooperative spirit,
and the kind of innovations that gave them their springboard will
get "used up", and they are going to have to start doing more of
the heavy lifting themselves to get to the next level.  My point
is that like sustainable agriculture, it's better to start before
you've depleted the soil.

I'm really not just talking about free software development here,
but about open standards, and an attitude about cooperation.  As
we go into a more competitive mode, there are unexpected
"friendly fire" casualties, and I think we may all wake up at
some point to an empoverished environment.

> > All of which is to say that I think that Nick (as reported by Frank)
> > is right: the money isn't in software, it's in providing net-based
> > services, and the question we ought to be spending some time on is
> > how open source plays in that space.  What kind of licenses, and
> > what kind of business models, make sense when software is key to a
> > company's success, but *software distribution* isn't a significant
> > revenue vector for that company?
> If a company consumes software but never contributes software back,
> and uses that software to provide a service that they charge for, are
> they even part of the equation?  If so, enlighten me.  I mean, if some
> company secretly uses free software internally and doesn't tell
> anyone, who cares?  They aren't part of the community.  They don't
> effect development.  They don't affect our options and futures.  I
> wouldn't even consider them anything other than a potential marketing
> point, and even that's only if they're willing to put their name on
> something they have no control over.

On the one hand, I agree completely with your argument.  In a lot
of ways, this is an essential element of the disagreement between
BSD license/GPL license advocates.  GPL'ers tend to feel that
someone building any proprietary extensions on top of their work
is evil; BSD people feel, as you do, that nothing is taken away,
even if nothing valuable is given back.

I am actually of that camp, and it's one of the things I admire
most about the university-style licenses.  They are (to use
another analogy) simply enriching the topsoil, the environment
from which further free OR proprietary efforts can grow.

That isn't my point at all.  My point is that if you take
something free, and grow something proprietary on top of it, and
are really successful, you can in fact change the rules of the
game in such a way that makes the free people have to run harder
to catch up.  Now, I don't have some kind of hardwired moral
outrage at this.  TANSTAAFL.  What I am trying to make is an
observation that if you do this consistently enough, you make it
harder for the kind of volunteer innovation that you benefited
from in the first place to flourish, thus impoverishing your
environment, and hurting yourself in the long run.  

That's why I think that "sustainable agriculture" is in fact a
good analogy.  If you take more from a community than you put
back in, you can weaken that community, and limit the amount that
you can take out later.

Note also that "put back in" to me doesn't just mean "contribute
free software."  Even though O'Reilly books aren't free, I think
they contribute heavily to the community.  What would be an
equivalent example of "taking but not contributing" is if we
started publishing books only on the leading products, where we
could make the most money (as many of our competitors do) and
ignoring the up and coming projects that are important.  So, for
example, if we said "we make too much money on sendmail books to
do a book on qmail" that would be putting short term revenue over
long term good.  (We have a book on qmail in the works, btw,
being written by none other than our host, Russ Nelson.)

I also want to address the point that "what happens outside the
community doesn't matter." Case in point:  Microsoft, a huge
consumer of internet protocols and software, takes that software,
incorporates it into their products, and changes the rules of the
game.  Talk to all the web site developers who have to do twice
the work on their sites to accomodate dueling MS and netscape
versions of html, and you might see what I mean.

Now, you can say "But Microsoft isn't a content provider of the
sort you were mentioning earlier" and that's true.  But the same
point, that someone who controls data formats controls a lot of
what software needs to do, does in fact apply.  The protocol for
submitting info to Amazon, as mentioned above, or sites to Yahoo,
or the ways to get listed in search engines, all can have
tremendous impact on the free software community, and the kinds
of problems we have to solve.

Factors outside the community always matter more than we think.

> Open source is about community.  If someone doesn't want to be part of
> the community, oh well.  Let's give them the freedom to choose that
> path at their own peril, and worry about the people who *do* want to
> be part of the community.

Well, part of my job as a publisher and all around loudmouth is
to say "there's danger over there" and so I prefer not to let
them "choose that path at their own peril" but instead try to
warn them about it, and as a result, to get them to add something
back for their own long term good.

Tim O'Reilly @ O'Reilly & Associates, Inc.  
101 Morris Street, Sebastopol, CA 95472   
707-829-0515 ext 266, Fax 707-829-0104,