Sun had a clear business model with Java:
- Make JavaSE as free, open, and ubiquitous as possible.
- Make money off of JavaEE tools and support
- Make money off of JavaME licensing
The part where they're getting paid is what funded the development of all three. If software patents did not exist, then they would not have open sourced anything, and would have stuck to their previous "open source" strategy of allowing read-only access to the source. But, because software patents do exist, they opened up their copyright protection, allowing projects like Harmony. The idea being that, it would make JavaSE more ubiquitous and open, but wouldn't cut into JavaME licensing. Apparently they didn't realise that smart phones would become as powerful as desktop computers, and some company, with the resources of Google, would come along, and abuse the openness of JavaSE to create a mobile Java-esque environment, to avoid paying the costs of JavaME. Yes, blame Sun for being stupid and not grasping the ramifications of Moore's Law, and not adding even more value to JavaME. But in the end, Google has abused Sun's openness to circumvent their business model.
There are potentially huge ramifications from this abuse, for the whole software industry. Many software companies have an opensource / commercial business model, where some base product is open and free, and some pro variant of it costs money. That means that the big players pay for pro, and everyone else gets the free ride of the open variant, so everyone benefits. The free users help bug report the product, and spread its mind-share, so they're still contributing necessary aspects, but they're not paying the bills, so companies still need the big customers to actually buy the pro version. Now, if Google gets away with taking the open version, straight up copying it, and not paying for anything, damn the patents, then that will create a huge disincentive for having open software. We'll be back in the closed source days, which really sucked for everyone.
And don't tell me that only charging support is a solution. When your software works well enough, and is simple and straightforward to use, then no one needs support. What they will buy is a pro product that adds value. Plus, product companies get better share valuation than support companies, which affect VC funding. Either way, if the market believes that large companies will find ways to avoid paying for your product, funding will dry up, and the software will not get made.