A number of people have asked that Yahoo release Pipes as open source, as if it would just be a matter of taking the source and uploading it to GitHub and that's that.

I've released a number of products as open source this year, MyWord, PagePark, River4, NodeRunner, NodeStorage, and a bunch of others, and it's a lot more work to get something ready for a source release than it is to just deploy it to your server and let it run.

Suppose Yahoo Pipes includes libraries that are proprietary to Yahoo, that they don't want to release open source? What then?

And what if they use licensed code to create Pipes, code that they don't have the rights to release? Or even worse, they have licensed code, and the company they licensed it from is gone, or bankrupt? Then you're in a gray area. Not the kind of risk a company would want to take, esp when they're trying to get out of a business, as opposed to building one.

And then there's the question of what kind of shape the software is in. Try doing a careful code review of a real-world piece of software that's been around for eight years, and maintained and extended by a dozen different people. Do you really want people to see it in that kind of shape?

Done right, an open source release isn't anything like putting out the trash, it's more like planting a seed for the future. In any case the chances that someone would be able to deploy a product like Pipes with source from Yahoo more easily than starting over, from scratch, seem pretty slim.

However, one thing that does make sense is to develop an app like Pipes, originally, from the start, as open source, and to encourage other deployments. That way people could invest with confidence, knowing that if you ever shut down your service, they would be able to continue. You should also look for a redirection guarantee before you invest in such a free platform.