5 ms·

We need more decentralised P2P systems. Decentralisation has lots of problems that make it crap for most people (e.g. verifying that you get what you think you'

by rmc 14y ago

We need more decentralised P2P systems. Decentralisation has lots of problems that make it crap for most people (e.g. verifying that you get what you think you're getting, and slow speeds). These problems are Hard Problems™. This is chance for a real hacker to solve them. We can beat them with technology.

kenthorvath 14y ago
Junk files can be catalogued by hashes and hashes can be indexed by wiki services with an API. So, downloaded junk? The app has a report button that flags the wiki. About to download something? The app can check the wiki for you. A wiki of file hashes would be great for cataloging things as malicious, describing the content, and cross referencing files with other versions (e.g. same content different format). Thus, we can make something really useful with a simple DB. I've been working on something like this, and registered hashpipe.net as a potential domain for this service. Anyone interesting in helping out is welcome to contact me.
- rmc 14y ago
  But then the wiki is the central source that can be shut down? And how do you prevent people changing the wiki to say that their junk file is legit?
  - kenthorvath 14y ago
    I don't believe there would be a legal basis for taking down a site cataloging metadata about files. The information would be just as useful to copyright holders as it would be to copyright violators (e.g. tracking ownership, issuing takedowns on the basis of hash collisions, etc...). To be very careful, we might decide that no URLs to a hash be cataloged in the 'wiki'. As far as quality control - take a look at wikipedia - it's a beautiful mess. People make edits to entries all the time, some malicious, some not - there is a change tracker, and a community of editors. The consensus and the truth win out over individual acts of malice. If shut downs were truly a concern, the database entries could be stored locally by the user making the entry and retrieval of entries would be like doing a query on a decentralized p2p network (gnutella). But then, that's messier and less scalable. I think we should just make the wiki, and see how it goes as a first step. Establish its usefulness, build a community, and then support it.
    - Dylan16807 14y ago
      I don't believe there would be a legal basis for taking down a site cataloging metadata about files. Remind me what Demonoid did.
    - rmc 14y ago
      I don't believe there would be a legal basis for taking down a site cataloging metadata about files Laws can be changed. The copyright cartels have the money and political connections to change the law in their favour. We have seen them do it before and they're going to do it again. Torrent sites (like The Pirate Bay) are essentially just that (sites cataloguing metadata about files). They have not be immune to legal troubles, and they are certainly not "beneficial to copyright holders". You're essentially describing current torrent sites. We have them, they are very valuable and useful. They work well enough. However they can (and are) shutdown. We need better.
tomjen3 14y ago
Not really. To verify what you get you just need to have the groups sign the content. Doesn't have to be from anywhere official (that would be counterproductive), but given time you would learn to trust a specific signature, based on past actions/releases. Then it wouldn't matter where you got them from.
cookiecaper 14y ago
When you're against conglomerates like Comcast, Time Warner, etc., whose interest heavily leans toward draconian copyright enforcement, and who not only control the pipes but also have purchased most of Congress, any filesharing protocol is going to be an arms race just as BitTorrent has been. BitTorrent has easily won and been very successful, and a "harder-to-kill" protocol might end up that way too and give a bit more resilience, but if we want things like this to really stop, we have to fix it at a social level.
- rmc 14y ago
  Yes laws need to change, but it's also good to have a backup and redundant system. After all if the laws can change for the better, they could change for the worse later.

by rmc 14y ago

kenthorvath 14y ago

Junk files can be catalogued by hashes and hashes can be indexed by wiki services with an API. So, downloaded junk? The app has a report button that flags the wiki. About to download something? The app can check the wiki for you. A wiki of file hashes would be great for cataloging things as malicious, describing the content, and cross referencing files with other versions (e.g. same content different format). Thus, we can make something really useful with a simple DB. I've been working on something like this, and registered hashpipe.net as a potential domain for this service. Anyone interesting in helping out is welcome to contact me.

rmc 14y ago
But then the wiki is the central source that can be shut down? And how do you prevent people changing the wiki to say that their junk file is legit?
- kenthorvath 14y ago
  I don't believe there would be a legal basis for taking down a site cataloging metadata about files. The information would be just as useful to copyright holders as it would be to copyright violators (e.g. tracking ownership, issuing takedowns on the basis of hash collisions, etc...). To be very careful, we might decide that no URLs to a hash be cataloged in the 'wiki'. As far as quality control - take a look at wikipedia - it's a beautiful mess. People make edits to entries all the time, some malicious, some not - there is a change tracker, and a community of editors. The consensus and the truth win out over individual acts of malice. If shut downs were truly a concern, the database entries could be stored locally by the user making the entry and retrieval of entries would be like doing a query on a decentralized p2p network (gnutella). But then, that's messier and less scalable. I think we should just make the wiki, and see how it goes as a first step. Establish its usefulness, build a community, and then support it.
  - Dylan16807 14y ago
    I don't believe there would be a legal basis for taking down a site cataloging metadata about files. Remind me what Demonoid did.
  - rmc 14y ago
    I don't believe there would be a legal basis for taking down a site cataloging metadata about files Laws can be changed. The copyright cartels have the money and political connections to change the law in their favour. We have seen them do it before and they're going to do it again. Torrent sites (like The Pirate Bay) are essentially just that (sites cataloguing metadata about files). They have not be immune to legal troubles, and they are certainly not "beneficial to copyright holders". You're essentially describing current torrent sites. We have them, they are very valuable and useful. They work well enough. However they can (and are) shutdown. We need better.

tomjen3 14y ago

Not really. To verify what you get you just need to have the groups sign the content. Doesn't have to be from anywhere official (that would be counterproductive), but given time you would learn to trust a specific signature, based on past actions/releases. Then it wouldn't matter where you got them from.

cookiecaper 14y ago

When you're against conglomerates like Comcast, Time Warner, etc., whose interest heavily leans toward draconian copyright enforcement, and who not only control the pipes but also have purchased most of Congress, any filesharing protocol is going to be an arms race just as BitTorrent has been. BitTorrent has easily won and been very successful, and a "harder-to-kill" protocol might end up that way too and give a bit more resilience, but if we want things like this to really stop, we have to fix it at a social level.

rmc 14y ago
Yes laws need to change, but it's also good to have a backup and redundant system. After all if the laws can change for the better, they could change for the worse later.