[notJim]

> But not much effort has gone into making these solutions user friendly and giving them the marketing push of something like this venture.

This is the key thing. The desktop app I'm imaging is user friendly, and probably at least somewhat pretty. I think the main technical issue is probably the fact that most people are NATed by their routers, so peer to peer is tricky, but I think with uPnP, you could get around that.

I was originally thinking of this as an open source app, but I suppose one could go the paid app route.

[barbazfoo12]

"I think the main technical issue is probably the fact that most people are NATed."

It's really not much of an issue as long as at least one peer has a reachable IP. I sometimes wonder how many people are under the impression that NAT's are a showstopper. This is simply not true. The showstopper is probably the RIAA and MPAA.

Skype slipped under the radar because they branded themselves as VOIP not file sharing. But it's really no different. It's peer to peer data exchange.

Only if both peers are behind the same NAT does the NAT pose a problem, in which case an external "supernode" is needed. But that's easy to set up. And it does not need access to packet payloads.

You could do a paid app. But the code to accomplish the job is very simple and has been made public in various forms multiple times.

File sharing copyright concerns, monitoring communications to catch bad guys and all that stuff is what's holding this back, not lack of a solution for connecting through NAT's.

[notJim]

> File sharing copyright concerns, monitoring communications to catch bad guys and all that stuff is what's holding this back, not lack of a solution for connecting through NAT's.

Why? It's software transferring files between two parties, just like you can do with email and things like Skype and even AIM, and hell, even the service we're talking about. Is there precedent for the people who write such software having legal problems? I'm aware of cases like rapidshare and torrent sites, but for example, are the authors of bittorrent clients also targeted?

> I sometimes wonder how many people are under the impression that NAT's are a showstopper. This is simply not true.

Hmm, I'll admit I don't fully understand this, but back in my torrenting days, you always had to forward a port to be reachable. How do you get around that?

[barbazfoo12]

If you want some interesting precedent, do some research on "Internet2" and the testimony of the RIAA to legislators.

Do you know what a LAN is? It is an evil invention to share copyrighted works. It must be stopped.

If you want a better understanding, read everything you can find on UDP, Ethernet, firewalls, NAT and encapsulation, in that order. I would suggest not to waste time trying to figure out "pre-packaged" peer to peer software solutions (i.e. all the different approaches people have taken, e.g., aeroFS, Kicksend or whatever). They often include far more complexity than you need to accomplish peer to peer. As such, they won't help you much to understand the basics: how connections are made.