>Licenses like the GPL exist for a reason. And that reason is an attempt to maximize the amount of software that is free.
Does that work well out in practice though?
I think companies are much more likely to work with and contribute to open source programs if they are MIT/BSD/Apache.
>The result was that people who used generous licenses got frustrated that others would come along, use their software, criticize the original author for "not protecting freedom", then create a modified version that the original author was not free to use as they wish!
In 2015 most of the internet is running on free applications running on top of free operating systems and commercial companies are spending billions to maintain this situation. If you had predicted that circa 1990, you would have been dismissed as a lunatic. So something resembling success has happened for the free software movement. And no matter how you ascribe credit, it is hard to say that it would have happened without RMS.
How many cases has that happened?
A lot, but you don't hear about it if you're not swimming upstream trying to maintain code under a generous license. For a random example where it boiled out in public, see http://lwn.net/Articles/247872/.
So, OpenWRT (https://openwrt.org/) exists only because of the FSF enforcing the GPL. A non-complient company tried to unlawfully withhold it, and only because the GPL has some teeth did it actually end up with the source available.
I'd rephrase parent's statement to "that reason is an attempt to maximize the relative amount of software that is free."
So, if with BSD/MIT you'd end up with a few more contributions, but you'd also help a few companies build proprietary software that restricts their users, it's better to use GPL since fewer users end up being restricted by proprietary software.
And yes, I know that the users have the choice to not use proprietary software, but regardless of that discussion, that still doesn't mean you should help that happen.
Does that work well out in practice though?
I think companies are much more likely to work with and contribute to open source programs if they are MIT/BSD/Apache.
>The result was that people who used generous licenses got frustrated that others would come along, use their software, criticize the original author for "not protecting freedom", then create a modified version that the original author was not free to use as they wish!
How many cases has that happened?