I'm just catching up with the proposal by Lawrence Lessig and all of the resulting discussion. As a software professional, the post that most hit home for me was Sandy Wilbourn's post which touched on the underlying complexity of executing the idea. I think there's some especially tricky questions around the issue of what kind of software is covered. If I write a piece of software for a company under contract to them, is that covered? How about if design and sell a product specfically for a internal use of a company, or the government? These seem problematic if they are covered, but if they aren't I'm sure we'd see Microsoft's "Personalized Software Products" initiative very quickly. Other than the complexity, I actually can't find much wrong with the proposal. The problem is that I actually can't find much right with it either, mainly because I don't understand how anyone's behavior changes. The main thing is that the halflife of software is so short — who would benefit today from the source code to the versions of Windows pre '95? On the flip side, I also don't buy Dave Winer's argument that Symantec would have been reluctant to buy MORE with such a law in place. The only interesting concept I've heard is around 'orphaned' software, but the overall benefit of even that seems like a stretch. This thing gets more interesting if you say 2 years instead of 10, but for every good thing I can think of bad things. I think that really does dis-incentivize alot of positive activity. I'm glad to see some creative ideas thrown out, but I think we'll need some more before we get somewhere.