Specs for "Diligent sharing", anti-freeloader tools

[PhlemLoagie]

A simple solution to the freeloader problem:
Give diligent sharers the tools to identify freeloaders and subsequently block access by them. Once implemented this will 'encourage' honesty and improve p2p network health.
....
So we have newbies, who need a break, freeloaders, who only want to take, and sharers who want to improve the p2p experience. There is little doubt that the significant number of freeloaders are adversely affecting the p2p experience. Because freeloaders do not make their downloaded files accessible, sharers must wait longer and longer to get access to the same files from other shares. This, obviously affects newbies as well. The problem is huge. Looking at the fine post by Jorge at http://www.zeropaid.com/news/article.../03222002e.php I know this subject has been well explored. My "new" suggestion basically is to encourage sharing and discourage freeloading by placing the control in the hands of the "diligent sharer".

If I am able to use my p2p program to block freeloaders, that means that legitimate sharers will have greater access to my files. If other sharers start doing the same thing then it WILL start to make downloading a big headache for freeloaders... nowhere to go! So we need 2 concepts fully implemented:

1. The diligent sharer needs to have the tools to determine whether an uploader is a newbie,freeloader, or sharer. The determination will be somewhat subjective based upon variable factors. (But there will be an appeals route where a blocked address can send a message explaining why he/she has no files, etc..)

2. Having determined that an upload was originated by a freeloader, the diligent sharer then needs to be able to block that ip address from further access to his/her files. The blocking can be temporary or permanent.

Once implemented I believe these 2 concepts will begin to have a dramatic effect on the health of p2p networks. Freeloaders will begin to realize that many of the desireable files are becoming off-limits to them. Their activities will not be degrading access for diligent sharers. Performance of the network will improve.

We dont need complicated ranking systems, forced sharing, etc..
Just give sharers the simple tools above and changes will come.

[MamiyaOtaru]

I agree with this almost wholeheartedly. I wish LimeWire hadn't taken out the browse host option, it was a good way to see whether someone was a freeloaded, and their IP could then be blocked.

But, that reminds me of the problem with the proposal. Some people don't have static IPs. They would not be effectively blocked, while otehr people might find themselves unnecessarily temporarily blocked.

IMHO it would be ideal to have a service like Napster where people have a unique user name, and then that this username could be blocked when a diligent file sharer detirmined that it was a leech.

Of course, this may conflict with privacy issues..

For me, the ideal approach is eDonkey's partial downloads. If you're downloading, you're sharing, nothing you can do about it. There may be a lot of people who only share incomplete files, but it is a lot better than a bunch of people sharing nothing.

[PhlemLoagie]

Regarding the issue with DHCP addresses. Yes, many folks get switched around on the last node of their ip address. One day they are 207.199.103.44, the next day they are at 207.199.103.100. So, if you block the freeloader at .44 then he/she could be back the next day at .100. Point well taken.

How about enabled blocking of the class C node? So the first day we block 207.199.103.* You may have blocked possible innocents from accessing your files on this node but so what? I think the good outweighs the bad. It is only access to YOUR files, not the entire network. Over time the freeloaders will see fewer and fewer results showing up on their file queries, or, better yet, let them start the download the autmatically get a message "Download IP blocked by file owner". This way the message TO freeloaders will sink in: "SHARE YOUR (quality) FILES OR BE CUT OFF".

This cannot be hard to do. All P2P programs utilize the 2 peoples ip addresses to facilitate the file transfer. Simply check the requesters ip against the senders 'blocking list'.

[quintessence]

What would you do if the freeloaders started to IP masquerade, or go though a series of proxies as in something like peek a booty?

The problem I see in blocking at the c class node is that there will be many freeloaders initially, blocking at this level will cut of thousands of Sharing members, not just a few. This also opens the possibility of more abuse as people might try to mount denial of service attacks not by overload, but by "computer paranoia" (i don't. know of a more adequate term. see sit.1)

I would instead suggest possibly using the MAC address instead, or something that has a unique id attached to a computer. To protect privacy, you could do one way hash to make it difficult to identify whom the user is. It seems plausible that you could get the hash value in a handshake process when a user requests a transfer (sit 2)

*Situation 1

Malevolent user triggers lots of block orders b/c they look like a freeloader

Users on that node find it difficult to download

Malevolent users go to new node and repeats (via IP masquerade, proxies, etc)


*******

*Situation 2

User A requests download from user B
A sends MAC address of 123abc (hashed as ¥sdegj)
B checks A to see if they are sharing
A doesn’t' share
A gets logged as bad

*******

I know MAC cloning exists, and lots of people don’t have network cards, but there must be some other unique number attached to an individual computer. That could be used in lieu of a MAC addresses.

[PhlemLoagie]

Good reply. Thanx.

It's been a while since anyone replied to this thread. I'm pretty sure that whatever we may decide, nothing will happen. We'd have to get P2P programmers interested and involved. Seems like there is not enough interest in this kind of discussion. I've been a business systems analyst for many years, working to make manual and automoated system more efficient, user friendly and successful.

The main problem with P2P programs, from what I've seen, is that the programmers tend to create their programs to conform to their own particular file sharing philosophy. While this may give them a feeling of accomplishment in the short run it will doom the programs longevity. Software needs broad intellectual heterogenity in its DNA to be successful for the long run. Software users rule! Programmers wishing to have a successful product must be ready to abandon personal opinions regarding program feature sets... and listen to the users!

I agree with what your prediction that a freeloader could be able to thrwart 'freeloader control' features in P2P programs. However I dont think many freeloaders have the know-how to find and implement an effective ip-based crack, much less create one themselves. I dont think a 'Malevolent freeloader' could do much harm. They'll find it's easier to share files than to play hacking games. Using a computers MAC as a persons P2P ID is a great idea. I'm sure it's easy for a P2P prog to get it's MAC. But sending it over the internet might upset too many people so they would not want to use the prog? What are the ramifications of having your MAC sprayed all over the internet?
Would a MAC hash mitigate the political incorrectedness?

Regarding blocking of an ip 'C' class. This would block only up to 255 addresses. But I say, so what? Thats a very small number in the scheme of things. And it's only blocked at the descretion of the individual 'deligent file sharer', not at a super-hub level. So I dont see many innocent sharers getting affected. At any rate, the block could be set with an expiration date, say, 1 month later. So if the freeloader moves off the node it will become unblocked.

Thanx for the comment.