[TriLUG] Can open source solutions be viable companies?

[Tarus Balog]

Vestal, Roy L. said:

> Question to me:
> "Do you have pointers to info on how open-source solutions can still be
> viable companies.  There is a big debate in the bioinformatics community
> about this, and I thought you might have some info."

I need you to clarify one thing - are you talking about using open-source
solutions in your company or are you talking about building a company to
supply open-source solutions?

I think the former is a definite yes. In fact, I just finished a project
with a bioinformatics company in New Jersey that was moving all of its
software from HP-UX to Linux (including Oracle).

As for running a company that supplies open-source solutions, I believe it
really depends on the business model. While I may be opening myself up to
flames here, I would never initially open-source, say, a computer game.
Why? Because when I buy a computer game, I expect to be able to install it
and run it with little outside help. Also, the cost of the game is pretty
minimal, so it isn't a real financial hardship. [Note that this applies to
the initial release - I can make a case for releasing the code to a game
after a certain amount of time has passed.]

But what about high-end software such as databases, enterprise resource
management (ERP) systems, and network management? Traditionally, these
products do not work out of the box and require a large amount of
customization. In these cases, the expertise that you buy with the
software provides much more direct value than the software itself.

The software becomes a tool, and as such benefits from the open-source
model. I have never seen a master carpenter complain that I can buy the
same lathe at Home Depot that he can, because while he can make fine
heirloom furniture with his lathe, I am lucky to keep my fingers. Those
companies that can best use the tools will be successful.

Not to belabor the topic, but I work at a company where open-source tools
provide a direct benefit. We do network management consulting with an
emphasis on open-source solutions. In the past, one would take a variety
of proprietary products and build a solution out of them. Often, many
products were required because a required feature or two would be missing
from the main product. This software was glued together by a variety of
home-grown code.

Now, when the engagement was over, rarely was this "glue" documented, and
with every additional software upgrade the client was faced with calling
in the consultants once again, and they had to remember what they did the
first time, and try to reverse engineer the proprietary products enough to
rewrite the "glue" code.

Enter in open-source. Now the consultants show up on site with a lathe and
not a black box. They have the tools to design a solution that fits the
customer's needs, and then they can place that code back into CVS so that
it can grow (as well as be tested by a much larger audience).

For my business model, there is also an added benefit. When you run a
consulting shop, what do you do when your folks are not billable? In my
case, they work on support and on improving the open-source code. Thus I
can level out their activity over time.

I have also started to see a blacklash against expensive software. How do
some of these companies come up with their prices? I believe they simply
make them up and then see what the market is willing to pay. Most offer
support at some percentage of the initial price on a yearly basis, and it
is required to get bug fixes. Think about it: you pay a large some of
money for software to perform a task, due to a bug it doesn't perform that
task,
and you have to *pay* more just to get what you thought you were buying in
the first place?

Then there is the ROI argument. Suppose I have a network problem that
results in $60K a year in lost revenue.  I can prevent that problem by
buying a $300K management solution. Should I do it? Well, on the surface,
you divide the savings into the cost and find a five year ROI. Not great,
but not bad. But wait, suppose the software vendor charges you 10% a year
for support? Then your yearly savings is cut in half and ROI doubles to
ten years. What if it is closer to the industry average of 20%? Then ROI
is infinite and you are better of living with the problem.

Whew - sorry - I got carried away.

The idea of open-source solutions makes sense in many markets. However,
there is still a lot of resistance to the idea, and it is something you
have to deal with if you decide to base your business on open-source. Some
companies are willing to shell out large sums of money for proprietary
software and services, but if it is open-source they expect everything to
work perfectly for free. But I have seen enough people coming around to
the idea that I can put food on the table and keep a roof over my head.

I honestly think the days of expensive, proprietary software are numbered
(with a few exceptions). It's not going to happen tomorrow, but probably
in the next decade. Think about the number of children who have been
exposed to and are comfortable with computers. As they grow up, they can
have a great impact on both the market for software and its production. As
businesses struggle to justify software price tags in the hundreds of
thousands to the millions of dollars, this new work force can create
viable alternatives.

Sorry for the long answer to your short question. I am no expert in
bioinformatics, but if you build your business model correctly, I think
that open source solutions can be viable companies.

-T

-- 
Tarus Balog
Consultant
Sortova Consulting Group, http://www.sortova.com
+1-919-696-7625
tarus at sortova.com