The End Of P2P Blocking ?

[GhostShip]

This is an interesting story, I had not bothered to read it however to find that out, sorry for the delay in posting it folks. 

http://www.slyck.com/news.php?story=1185

Rather then being scorned and lose customers, many ISPs are choosing to cache P2P traffic. Caching works by keeping P2P traffic off a ISPs network by retaining popular query results on a server farm. For example, let’s say the BitTorrent protocol is taking up 70% of Verizon’s network. It can implement a P2P cache server that will retain the most popular of search query results, thereby lessening the amount of traffic that must transpire on their network. Instead of downloading a popular song or movie from a customer on AT&T’s network, the information will download from Verizon’s cache server.

Several large ISPs are experimenting with the idea. NTL, a British broadband firm, has partnered with CacheLogic and BitTorrent in a bid to alleviate their bandwidth consumption.
Thailand’s True Internet, the country’s largest ISP, is also claiming success with PeerApp. PeerApp provides a technology very similar in nature to CacheLogic, where P2P traffic is routed to server farms rather than throughout the network. This has been an apparent success, as PeerApp claims that 60% of the ISPs traffic has been reduced.

This will of course only help rather specific types of P2P apps as many transactions are one offs on a real network and therefore its counter productive to cache such activity.
While this may help in the short term the future is more likely to be one involving fiber optics to a customers home and with it the end of the bandwidth and speed problems for the foreseeable future, you cannot after all expect to obtain a gallon from a pint pot.

[SamSeeSam]

cache P2P traffic
that will retain the most popular of search query results

What exactly does this mean?

[GhostShip]

Caching is simply making a copy of the information required and delivering that data instead of the data from original location (in this case a web site).

[SamSeeSam]

So how does that help them? Do they then restrict the download speed?

[KM]

it means that you are downloading it from their servers instead of the original user, your ISp doesn't care about you using bandwidth to download from their servers because their internal network has plenty of capacity and upgrades are relatively cheap (a 1 off cost in most cases to lay cables or to upgrade the equipment to use existing fiber at higher speeds or whatever)

the only real use of caches so far are for websites (for example AOL is a major ISP that caches all web traffic), when you request a page your request instead goes to your ISPs cache, that then gets the web page and sends it to you - no benefit there, however another user then connects to the ISPs cache and requests the page, it can send it from memory instead of getting it again - this reduces the bandwidth usage on their connections to other ISPs (the expensive ones to maintain and upgrade etc)

the general idea of caching for p2p is the same, the first download will be just like normal, but the next user who downloads it gets it from the ISPs server instead - this is good for the ISP because it reduces their bandwidth usage, and good for the user downloading because they get it from a faster cache, and good for other users because there is now an extra slot available on the person who was going to upload it...

web caches have some flaws (for example if a page changes and the cache has an old one) but in the case of p2p that is not an issue, all of the benefits and i can't see any problems with it - however I'm not sure how well it would work for p2p, because of the huge variety of files (anything popular has hundreds or thousands of files for the same thing... so 2 users on the same ISP picking the same one...)

if it works then great news, however it is much harder for a p2p network than for websites so I'm not sure if it will work... for torrents it could work to a certain extent, unfortunately i don't see them bothering with winmx

[SamSeeSam]

web caches have some flaws (for example if a page changes and the cache has an old one) but in the case of p2p that is not an issue, all of the benefits and i can't see any problems with it - however I'm not sure how well it would work for p2p, because of the huge variety of files (anything popular has hundreds or thousands of files for the same thing... so 2 users on the same ISP picking the same one...)

if it works then great news, however it is much harder for a p2p network than for websites so I'm not sure if it will work... for torrents it could work to a certain extent, unfortunately i don't see them bothering with winmx

Web pages are relatively smaller is size than a thousand videos. How can anybody cache all the files in a network like P2P? There are no size estimates for the software precisely for the same reason. It's too big.

For popular songs, yoour theory will (not 'might' ) work. But that to when you have a very large custemor base. Each person will download a different thing and who is going to maintain so much data. I think the biggest hardisk avaiable is 1TB (Terabyte) The will be needing 1 or 2 such hardisks per day initially and then mabye per week (After a long long long time). And lets not forget manitenance costs.

If it does work, then well and good. that's all is that can be said.