I hope any site that is going to use it, will ask its visitors for permission first.
PeerCDN website completely avoids this matter, making them seem like fancy squatters.
Using the resources of a third party without its explicit consent, is infringement afaik.
I think this kind of things (granting permissions) should be done by the Browser, not the application. If some technology can be abused (like using the microphone or the web-cam, and apparently, WebRTC) then the browser should ask if this is OK, not the web-application.
Look like a great but dangerous tool.
Stealing users bandwidth and publishing users IPs is a no go,
but it could be really handy for online games and other web applications with lots of assets.
"To improve server performance game data may be transfered from your browser to other players. Other players might see your IP."
Right now, it's not anonymous but we're working on adding anonymity by relaying traffic through intermediate peers (like Tor).
Re: stealing bandwidth, I feel like web apps are being held to a ridiculous standard compared to Desktop apps. If you download an app, does it explicitly ask you for permission to use your upload bandwidth? I've never had any app to do that. But, a lot of apps do use upload bandwidth. In fact, Spotify does P2P to serve a LOT of its content. Check out the stats from Spotify: http://torrentfreak.com/spotify-a-massive-p2p-network-blesse...
Ultimately, I think it's better for the web if the traffic is more decentralized because it renders sites hard to DDOS and censor. You can read more about this aspect of peerCDN in our Mozilla Ignite App here: https://mozillaignite.org/apps/479/
I don't mind sharing my bandwidth for this (I don't have mine capped "literally speaking", but I'm sure there's some invisible limit), but I hope it asks for permission first.
I didn't see anything about that in the FAQ (I'm still browsing). Or maybe it's up to the sites where it's implemented? If so, that may be giving them too much leeway.
(I work on peerCDN). Thanks for the question. Right now, it's not anonymous but we're working on adding anonymity by relaying traffic through intermediate peers (like Tor).
And won't this potentially slow down the loading of your site (well, everything you're doing on the internet, really) as well, if your user's are using bandwidth to simultaneously seeding to others?
WebRTC seems to become a great piece of technology. Hopefully the NSA hasn't made its encryption useless, too. It would be great if it was still secure, since it's also P2P, so harder to do any of that "lawful" mass interception.
How does it detect me being on (or moving to while still open) a mobile or otherwise metered connection so that it can disable itself? If I keep the site open in a tab for an extended time but not currently browsing the site, will I become a dedicated server?
I don't think you can do the first reliably and if you can't, you shouldn't do this at all. It's evil and need to be stopped at the browser level.
I've never really understood why the browser makers don't build this in at the browser cache level. Lots of people work in offices and end up sharing links, it could be a massive saving for everybody.
Not a great deal of privacy problems, SSL content is excluded by default and you are just asking local peers if they have a binary matching a SHA hash.
The privacy leak might be that you are surfing the net when you should be working, however a firewall could give you that same amount of information anyway.
Lol it's P2P right? So why isn't it just a JS file? Isn't it overcomplicated just to fit the business model?
Why not take donations and sell the analytics as an aded service?
Even though most of the work happens on the client side, there's a tracker server that we run that's crucial to the whole system working (tracks who has what, serves as an authoritative source for data hashes, fights against malicious peers, etc.).
We have and will always have a free tier for everyone to use, but for bigger clients who want more support, we'll have paid plans. We'll also be open sourcing peerCDN soon.
That's one of the points. The other being offloading the burden of serving common components off-site. This, I think, specializes in just off-site availability.
We continue downloading from a different client (generally, files are downloaded from more than one client to improve speed) or if we estimate that that's still going to be too slow, we just load the image from the origin server.
PeerCDN website completely avoids this matter, making them seem like fancy squatters. Using the resources of a third party without its explicit consent, is infringement afaik.