"Which does lead to the question of exactly why a company would want to get into the business of guessing whether "the community" will continue to support a project (the article has a nine-point checklist for assessing the odds) as opposed to relying on the good old profit motive."

First point: A commercial product is no less risky. Offerings come and go, products are deliberately obsoleted (think any flavor of Windows prior to XP), licenses force upgrades and/or time-expire. In either case, doing your homework is necessary.

However if the product is open source and the community fails to support it, you can always support it yourself (in-house) or pay someone else to do it, since you have the code. You as the user have control over the product's lifecycle instead of another company. If the product is commercial and the vendor decides to stop supporting it, you are out of luck.

Lewis is right - like many overbroad criticisms of open source, this post accuses OSS of problems that are common to all software projects, be they OSS or commercial. Every commercial product in the world "exists at the whim of its funders," and every IT professional must predict "winners and losers" before using or supporting any software. OSS projects are less risky in this way because a new funder can step in without negotiating for rights.

James implies that to commercialize an OSS project is to be dependent on the whims of volunteers. This just isn't true, and the example he cites proves him wrong: if a project has commercial potential, then companies like IBM will fund it.

I guess OSS can't win - if it's a "community" project then it's unreliable, and if it has corporate funding then it deserves criticism for not being a true "community" project.

Can anyone provide a comparable ratio of proprietary software projects that made "lasting contributions" to ones that did not, or never saw the light of day? Hard to come by, of course, but I would guess the ratio is similar.