[TriLUG] Can open source solutions be viable companies?
Fri, 5 Jul 2002 13:16:43 +0500
On Friday 05 July 2002 07:56 pm, Mike Mueller wrote:
> On Thursday 04 July 2002 23:46, Tom Bryan reputedly wrote:
> > 3) fear that some competent programmer at the customer site will see the
> > code and say, "This code sucks!"
> I learned a lesson from a colleage that had a 2 year tech degree who was
> the most prolific hardware designer that I've known. His designs were
> criticized as sucking by the more educated and experienced....I learned to
> appreciate sucky code that works if it's an early revision.
That reminds me of the whole "worse is better" thing.
> > If I ran a business and didn't understand much about software,...
> Mind if I re-use your words?
No, not at all. Unless, of course, you use my words to write a book and
become rich and famous. :-) It's all been said before (probably more
coherently) by others. There really are good business reasons to use open
source software. Most of those reasons involve some sort of risk management
(cost, obsolescence, platform support, etc.). And if the customer has a good
tech staff, they may be surprised how well techies can leverage open code to
make everything go better. Of course, if their tech staff sucks, they'll
probably just screw everything up.
> > If you haven't done so, you should probably read Raymond's "The Magic
> > Cauldron" at http://tuxedo.org/~esr/writings/magic-cauldron/. Whatever
> > your opinions of ESR and his work, it may add some fresh ideas to your
> > license conversations.
> I've read some of ESR's work but not this one.
That's the paper that's probably the most relevant to thinking about business
plans. I thought that he had some links to competing views, but I don't see
them. I know that he took a lot of flak from MBAs and other business types.
Just because the paper says so, doesn't mean it's right. :-) If you give a
customer a pro-open pitch, and they don't go for it, listen to their
concerns. It's possible that they're just confused about something they
don't understand, but it's also possible that they have some legitimate
business concerns about open software.
And it's always possible that open sourcing your code will be so good for your
customers that they won't need to give you another dime ever. That would
fall into the "good for them, bad for you" category. ;-)
> I once wrote to ESR for help on the OSS question as it applied to commercial
> applications but there was no reply.
Not surprising. I get the impression that he's a busy fellow. There are
others in the community with whom you could discuss such things. If nothing
else, find yourself a good legal/business person with experience in the
independent software vendor industry. That type of person may be able to
help you formulate a good business plan and corresponding license.
> I think OSS in its most well-known form applies when you can
> find a community that will benefit equally from the effort required to make
> the product. I wonder how OSS works when there is a distinct and separate
> group of producers and consumers which is the situation I find for ss7box.
> The rational behind the direction I am taking is that I trust that the
> consumer is disincented from becoming my competitor and incented into
> keeping me alive.
Cool. I hope it works well for you. Unless you're talking about huge
customers (thinking about Fortune 500s here), it's unlikely that they'll want
to take on the risk of becoming an independent software vendor of your code.
They'd be more likely to "partner" with your company so that they'd get your
software, and you'd get to sell under their brand name with their business
> Your story of your customer making changes and attempting to get you to
> support them is educational.
I'm not sure whether it applies in your case. We're not even sure that's what
happened. Support for our customers sometimes goes like this:
Customer: "Your software is broken. It's giving me the wrong answer."
Us, after finding out what they're doing: "Odd. That should never happen.
Run this SQL statement, and paste the output in an e-mail to us."
We receive an e-mail that says: "I ran the command. Everything looked fine."
Us: "Thanks. That's exactly what we wanted." :-/
Anyway, when the customer explains results and sends you logs, and it just
doesn't seem to match the code you sent...you become suspicious. Once we
started obfuscating the code, those problems went away.
Imagine a medium-sized company with hundreds of employees. Some critical
software supports billing, and a small team it assigned to operate that
software system. The (semi-)technical lead of that team decides to play with
the code a bit because she imagines herself to be something of a developer.
Suddenly, billing appears stuck. Your manager yells at you. Now, would
*you* tell your boss and the vendor, "I was making changes to the code. Now
it doesn't work. I can't seem to find/restore the original. I need a new
copy."? Or would you simply tell your boss that you'll call the vendor with
a bug report?
Customers, what fun!