[watwatwatwat]

Would you pay for it? If yes build a list of competent developers, ask them if they would build it and lunch a funding campaign. Give all the money to them and donate yourself. Building software is hard, building what you asked for is even harder.

[chme]

I'd rather do it the other way around. Find people with money to spend and let them pay me (and other developers) for implementing it.

[mikekchar]

Just implement it. It's not hard. All the libraries you need already exist and are relatively high quality. I have done it myself (non-free system) getting on to 10 years ago. It took me less than a year by myself and I even wrote my own SIP implementation. These days you could certainly do it faster.

There are a few problems, though. One is that NAT traversal will still probably mean that it will be unreliable unless you also set up a service to carry data. The other main problem is that nobody will pay you for it. There are literally hundreds of failed startups who realised how easy it is to build these things. They couldn't find a way to make money.

It should be easy because if people are forced to pay for a service to carry data in order to make it reliable, you should be able to make money. Right? The problem is that people won't pay for reliability because they think that the software is buggy (see comments above re: "I tried Hello, but it was buggy. No audio, etc." Classic).

Skype famously solved this problem by being evil and routing it's traffic through un-NATed users. Google solves this problem by having more bandwidth than god (and hangouts still sucks half the time).

But if you ignore the "there's no way to make money at this" problem, it's a pretty fun project to write.