Subject: Re: appliance-based business models
From: Brian Behlendorf <>
Date: Tue, 25 Oct 2005 17:45:20 -0700 (PDT)

On Tue, 25 Oct 2005, David Kaufman wrote:
> I think your herring and Stephen's cost-curves both missed the mark. The 
> "other appeal to the appliance model", to which I think Brian was referring 
> was its appeal to us, to the free software *developer* (and free software 
> business person wannabe).  The appliance model appeals, to me at least, as a 
> way of making money ethically.  Yes, Free Software also appeals somewhat to 
> clueful customers as high quality but, as a way of increasing perceived value 
> and price, I think free as in beer is low on the features bullet-list. 
> Selling a box however, that is truly *worth* ten times the sum of its parts, 
> *because* of the software that you installed on it (and, perhaps, just 
> coincidentally you also wrote or helped to write) is quite appealing to 
> developers, is it not?  Is that not a decent way to spend your days (even if 
> nights are still spent up toy the eyeballs in debuggers, compilers and bug 
> tracking data?

I think there are actually two significantly different camps that 
developers fall into when it comes to receiving money for their work, that 
can be broken down to the labor theory of value versus the solution theory 
of value.

Those in the labor camp are happy to receive payment based on some healthy 
consideration per unit of time they spend working on the code itself. 
They'd always be happier with more than that, of course - bonuses and 
stock options are cool - but they consider it inethical, or at least 
slightly greedy, if someone else makes more money for what is perceived to 
be less effort.  They might extoll the amount of time they spent 
developing their product as a virtue to potential customers.

Those in the "solution" camp feel that they should be paid some 
substantial fraction of the money either saved or made based on their 
works, aggregated across all potential customers.  It may be the case that 
an overnight hack or publicly-funded PhD thesis forms the basis for a 
multibillion dollar company within 4 years, or it could lead to to 
nothing - that risk is itself a source of energy for the creative process.

Silicon Valley was built much more upon the latter than the former. 
Techies were usually well regarded not just for the brilliance of their 
work, but the ability to translate it into something bankable by the 
company.  Lotus, Thinking Machines, Apple, Sun, Oracle, all had engineers 
who were immensely proud of their ability to spend their weekdays cutting 
code to do things no one else had done before, and their weekends sliding 
along Highway 1 in their sports cars (or more pragmatically, affording a 
good house in a neighborhood with good schools).  Engineers used to 
proudly display their patent awards in their cubicles or offices.

It's no wonder that this culture had severe cognitive whiplash when first 
faced with Open Source; even those who had built companies upon 
pre-existing open source code like BSD, or open protocols like TCP/IP. 
But this is not the suits versus the geeks, as it might sound like at 
first.  I have met many, many, many engineers who fall into that 
"solution" camp rather than "labor" camp.

In my opinion, neither frame of mind has an ethical advantage over the 
other, but I do have my bias.  I don't sit around and wonder whether Larry 
and Sergey's search algorithms should have made them billionaires, because 
jealousy isn't very productive, and it'd be futile anyways.  At the same 
time, when someone tries to sell me something, I'm going to compare their 
offer to what I can accomplish using my own talent and free software.  So 
your solution-value (as a vendor) better still be better than what I think 
my own labor-value would be.  :)

> See the appeal?  Like poprietary software licenses and source-secret security 
> solutions, I detect in our community (and share) a strong repulsion towards 
> high-pressure sales, slick marketing, and evil corporation in general.

This is one of the biggest impacts by Open Source upon the IT industry.
People no longer feel, in general, like they need to acquire an Oracle 
database because there's some secret sauce that they can't otherwise get. 
Some recent survey (I'm sure someone here can come up with the reference) 
polled companies who had started to use open source software, and ask them 
why.  "Price" was not the number one reason, nor was "flexibility".  It 
was that they could acquire technology while avoiding needing to deal with 
a vendor, cited over 40% IIRC.  Software sales teams are notorious for 
their arrogance, the products underdeliver on the hype nearly universally, 
and customers are finally saying "enough".

But that's orthogonal to some of these issues - I could still have a 
pay-per-user-per-month sales approach and solve the above problem.

> But the point is that we distrust software for profit.

I think writing software for money almost necessarily perverts it in some 
important ways - and that the same thing is true for any creative 
profession, whether it's filmmaking, music, interior decorating, or the 
world's oldest profession.  Writing software to meet someone else's 
expectation of schedule and functionality is a pain in the ass.  I'm most 
comfortable if I can write software simply to suit my own needs, and then 
if others find it handy and want to contribute to it, so much the better. 
It's why in my opinion the most reliable Open Source business models are 
those where the business has nothing to do directly with the Open Source 
code underneath; but that the code underneath *enables* the business to 
pursue its goals faster, cheaper, or better.

Unfortunately, that kind of development is rarely going to result in 
dramatic functionality improvements in that underlying code, and runs the 
risk of the company going "hey, I want to make sure this tech only helps 
_me_ run 'faster, cheaper, better,' not my competitors".  I think every 
project needs a mix of people who are there for sustaining reasons, those 
who are there for core development of new features (and whose salaries 
have to be paid somehow), and those who are there for fun, for the 
intellectual stimulation of it all.  In fact it's the latter who are 
typically the most innovative, as they have nothing to lose if their wacky 
idea falls apart once implementation is attempted.  All three groups help 
buffer the worst aspects of the others - the sustainers keep quality and 
stability up, the core devs add new features that keep the system relevant 
in the long term, and the hobbyists show what's possible with an open 

> I don't want to just 
> avoid the *appearance* of all the "evil" things that stubbornly appear to be 
> indispensable to turning a profit in the software "publishing" industry, I 
> want to really *do* good work, leave the world a bit better than I found it, 
> *and* get paid well.  Too much to ask?

Certainly not; I think developers of both frames of mind would agree with 
you.  I bet Microsoft saw themselves as great liberators when compared to 
IBM; Google likewise says "don't be evil."  Easier said than done, lots of 
shades of grey depending on your vantage point, and it's tough to realize 
that customers don't *think* the way you and I do.  Most of us initially 
assume that we should be building a solution that we ourselves would pay 
money for, at the price we're asking.  That couldn't be more wrong.  We 
all know how to build something, but only understanding the customer will 
tell us why we should be building it and how much to charge.

> The concept of selling free software seems, to most people at first 
> blush, to be an oxymoron.  And some of my friends, family and colleagues 
> (most of who are outside the free software community, but even some 
> inside or on the fringe of it) tell me that I'd be a moron to try.  I 
> can usually counter that they are not qualified to make such a judgment, 
> standing there with a $1.50 bottle of Free Water in their hand.

Heh!  I think this proves my point.  They don't buy bottled water out of 
ignorance.  For all your admirable talk about treating customers like 
*people*, you risk treating their choices with dangerous contempt.

> But over time I've disappointedly succumbed to the disappointing 
> possibility that Doing Good, and Doing Good Work (i.e. Free Software 
> work), may just have to be mostly orthogonal to Getting Paid Well, or 
> having a successful business that scales well.  So disproving that 
> suspicion has a windmill at which I repeatedly tilt.  As do many of you, 
> right?  And everyone else that's "in" the software industry, but who 
> prefers to spend their time in the free software community?  Hence this 
> list?

I suspect that's why we're here and willing to help each other, yes. 
Though from my own perspective, it requires an awareness that there is no 
binary categorization of Good and Evil upon which you can put most 
actions; it's an N-dimensional space, and where solving for the best 
possible Good while being able to pay everyone's salaries is better than 
packing up and going home.

> So the Sourcefire/Snort situation has admittedly gotten me a little excited 
> that theirs may be a way to effectively straddle the free software and 
> profitable software worlds without compromising the freedoms and fairness we 
> enjoy in the former or settling for mere "lifestyle business", one-man-show, 
> small-potatoes or also-ran status in the latter.