That being said, I see the major things lacking being:
a) Getting the license to be PHP compliant, working on getting it into
the core (will be a nice political issue. I will try to talk to Rasmus
today, actually, as he is in town giving a speech!) - I am not sure if
we can change the license easily or not, without asking Andrei, or
perhaps by simply changing the code and enhancing it, we can grant it
whatever license we want.
b) Getting the adaptive process spawning piece to work
c) Making sure that the patchwork continues to happen for PHP versions
until it can be baked into the core
d) Possibly changing the config file. I think he had mentioned liking
nginx's syntax. I don't mind which way it goes, as long as PHP
internals would accept that format :)
We'll need someone who can hack the code, and some funding to pay for
the development time.
Also any other bugfixes could be baked into this of course.
So, can anyone pledge their coding skills or money? I am willing to
put in some cash. I -may- be able to get my company to put in some
money as well.
I want to try to bridge the final steps here, and accelerate it, since
I don't know what Andrei is doing, and PHP-FPM is awesome but a bit
hard to sell when people don't see much of a website or official
development in a while. By getting it into the PHP core (if it can
happen that way) it will essentially negate the need for PHP-FPM as a
standalone project too. Of course, it will also require PHP to now
support a different style of configuration file.
Feel free to reply to the list or off-list to me. Especially if you
have dollar amounts you'd be willing to pledge. I can keep a running
count offline and try to organize the development.