Other changes that I had been working on weren't going to be merged upstream (already rejected), and though I wasn't going to make any attempts to prevent SSL/T

Other changes that I had been working on weren't going to be merged upstream (already rejected), and though I wasn't going to make any attempts to prevent SSL/TLS from being merge-able (I was actually starting from 4.0.9 for ease of current users, and had moved between 4.0.9 and unstable a couple different times), my code was derived from MIT code, and I can't really re-license someone else's code. This fork announcement was a, "Hey, you can probably use this stuff, but it might not be usable for everyone now. In the future, it might be a materially better solution for some of your problems (transactions). I'd appreciate some sort of vote of confidence if you all think it's cool, or even an 'eff off' if I've somehow pissed in your coffee." With the SSL work in PR #4855, some of the announcement has no real use now, and I've redirected my own efforts in its wake.