Ironica's Journal: A new business model

An idea popped into my head today, and I wanted to put it down somewhere before it got too heavy and fell out. It comes out of thinking, "What would it take to get someone to develop an OS version of GIS (Geographical Information Systems)?" Some may know, and others may not, that there's pretty much one company who makes GIS software (ESRI). It's hideously expensive, and not that great either. And while version 3.x is available for a variety of platforms (including Unix), the newest line, 8.x, is only available for Windows-NT based systems. (Apparently Microsoft threatened to create a competing product [which is MS-speak for steal their code] if they didn't introduce that limitation.)

So here's my thought: start up a non-profit company that is geared towards creating applications for government agencies (which tend to have the most specialized needs) and other large companies. The process goes like this:

• Agency or company comes to non-profit OS software developer with specs for a program they need. It may be an existing program that they want ported to an Open Source model, or something entirely new.
• NPOSSD bids the job based on how much it will cost to build it (remember, no profit margin).
• Here's the fun part: now the two companies, in drawing up the contract, establish a licensing fee that seems reasonable. Subsequent people who want a license to the same software will pay this fee.
• The entity that originally funded the project will get 90% of any collected licensing fees as a dividend, until their entire cost is repaid (accounting for present value and market interest rates). The remaining 10% will go into unfunded R&D work at the NPOSSD.
• In the event that multiple entities want to split the costs of production, the dividend will be similarly split in proportion to what each paid.

It's crazy, but it just might work. Certainly it would be something of a safety net for governments that want to adopt a preferential policy towards Open Source. It's basically a high-risk low-yield investment, since there's no guarantee you'll get any of it back... but you might get it all back with interest. Now to write up a proposal, I guess ;-)

One flaw

[DarkMan]

I can see one small flaw in your idea. Namely, it would appear that you cannot get the code unless you pay.

I submit that this is a significant difference from the normal open source model. Firstly, it prevents casual effort. Such things as trying to compile the system with obscure compilers, or on differing platfroms (read library support, rather than hardware) can often point out problems waiting to happen.

Also, it means that the perfect people to help with such an effort, namely programmers with GIS

Re:One flaw

[Ironica]

Very constructive and informative, if negative. I appreciate the comments.

I hadn't thought that far ahead, really. ;-) Perhaps they could have a developer program, where you sign up, get the code for free for personal development use, and can contribute. The licensing thing would be more or less an honor system, but if the fees are reasonable (I'm thinking in the $100 or less range, and only one license needed per entity) then people will probably be on the whole willing to pay for it. But this is a we