Because one of my extracurricular projects depends on the software, I've had to follow the progression of an open-source package which shall remain unnamed. The package is a set of code that can--with a few tactical edits to a configuration file and some minor knowledge of how hosted databases work--be installed on nearly any rent-a-domain website in the world.
Until recently (in my project's history, at any rate), the developers have been relying on donations to defray the costs of improving the package and making it available to the world. But apparently they've decided to also "eat their own dog-food" (so to speak) by also providing it as a service-for-fee. In other words, they will, for a price, take care of the set-up and maintenance for you.
Not a bad idea, in principle. You can write your own web software and host it for the price of your time plus less than $100 a year, which is a tasty-sounding business model. If you can charge your customers more than the cost of signing them, it's profitable. On the surface, this beats the stuffing out of worrying about folks pirating your handiwork.
But there's one big difference between selling software off the shelf and hosting it. That difference is that the needs & quirks of each customer will pull the software in a different direction. If each customer pays the same price for hosting, then they're all created equal--maybe the pushes will cancel out the pulls and so forth.
But the minute that one customer (by size, reputation, etc.) become more equal than others, the whole game changes. That's a contigency that should be written into the business-model from the get-go. Because as difficult as it is to make complexity simple in software terms, doing the same in business terms is, in my experience at least, much more difficult.
