The Pirate Bay as Blockchain?

[r_chen]

Hello everyone:

I have watched with some amazement the development of the Bitcoin project, especially admiring its technical properties and achievements. A rather off-topic side-project has recently popped into my head, which I would now like to publicly propose. It seems that, all cryptographic currency properties aside, the most interesting development of the entire project is the blockchain itself, and its distributed (and distributed-verified) database system. What has been considered about its application in other areas? The strong censorship resistance of the distributed blockchain model seems ideal in some ways for distribution of "forbidden" sites and data. As the subject implies, The Pirate Bay seems like a viable candidate for such a mechanism, if only as a proof of concept.

So to the Bitcoin technical team, and anyone interested, I ask the following questions:

1: Is it first of all conceivable for a presently dynamic web site like The Pirate Bay, to be created as a desktop app, and to have its entire database distributed and verified across thousands of computers, blockchain style?

2: What metadata could and should be reasonably and sensibly added to the blockchain, keeping in mind that with each torrent on The Pirate Bay you have: magnet link, description, comments, possibly an image, some user id, a time stamp and possibly more.

3: Could dynamic updating of content work in the same way? For example if someone wanted to add a torrent to the pool, how, in the blockchain style, might it be able to be flagged as good or bad?

4: As an addition to the above question, what could be done to dilute pollution to the blockchain (and its subsequent bloating in size), by parties interested in destroying the project?

5: What components that are currently a part of the main Bitcoin client and protocol that would absolutely not be necessary for this? I presume that various things about the transactions are unnecessary, but perhaps for the security of the network the need only be re-purposed?

Thank you very much for your time!

[1714167982]

1714167982

[Isokivi]

Now thats an interesting tought ... sadly one I have no answers to

[memvola]

This idea resembles Namecoin. In fact, maybe Namecoin could be used for this purpose as well.

[matthewh3]

Hello everyone:

I have watched with some amazement the development of the Bitcoin project, especially admiring its technical properties and achievements. A rather off-topic side-project has recently popped into my head, which I would now like to publicly propose. It seems that, all cryptographic currency properties aside, the most interesting development of the entire project is the blockchain itself, and its distributed (and distributed-verified) database system. What has been considered about its application in other areas? The strong censorship resistance of the distributed blockchain model seems ideal in some ways for distribution of "forbidden" sites and data. As the subject implies, The Pirate Bay seems like a viable candidate for such a mechanism, if only as a proof of concept.

So to the Bitcoin technical team, and anyone interested, I ask the following questions:

1: Is it first of all conceivable for a presently dynamic web site like The Pirate Bay, to be created as a desktop app, and to have its entire database distributed and verified across thousands of computers, blockchain style?

2: What metadata could and should be reasonably and sensibly added to the blockchain, keeping in mind that with each torrent on The Pirate Bay you have: magnet link, description, comments, possibly an image, some user id, a time stamp and possibly more.

3: Could dynamic updating of content work in the same way? For example if someone wanted to add a torrent to the pool, how, in the blockchain style, might it be able to be flagged as good or bad?

4: As an addition to the above question, what could be done to dilute pollution to the blockchain (and its subsequent bloating in size), by parties interested in destroying the project?

5: What components that are currently a part of the main Bitcoin client and protocol that would absolutely not be necessary for this? I presume that various things about the transactions are unnecessary, but perhaps for the security of the network the need only be re-purposed?

Thank you very much for your time!

I think what you may be asking for is Freenet - http://en.wikipedia.org/wiki/Freenet - with its distributed data store.

[paulie_w]

i don't think that solves any of the problems he's asking about matthew.

this is a really cool idea! mods, can we move this thread somewhere else, so that it can be seen by relevant people?

[Raoul Duke]

i don't think that solves any of the problems he's asking about matthew.

this is a really cool idea! mods, can we move this thread somewhere else, so that it can be seen by relevant people?

Any idea where it should be moved to?

[matthewh3]

Hello everyone:

I have watched with some amazement the development of the Bitcoin project, especially admiring its technical properties and achievements. A rather off-topic side-project has recently popped into my head, which I would now like to publicly propose. It seems that, all cryptographic currency properties aside, the most interesting development of the entire project is the blockchain itself, and its distributed (and distributed-verified) database system. What has been considered about its application in other areas? The strong censorship resistance of the distributed blockchain model seems ideal in some ways for distribution of "forbidden" sites and data. As the subject implies, The Pirate Bay seems like a viable candidate for such a mechanism, if only as a proof of concept.

So to the Bitcoin technical team, and anyone interested, I ask the following questions:

1: Is it first of all conceivable for a presently dynamic web site like The Pirate Bay, to be created as a desktop app, and to have its entire database distributed and verified across thousands of computers, blockchain style?

2: What metadata could and should be reasonably and sensibly added to the blockchain, keeping in mind that with each torrent on The Pirate Bay you have: magnet link, description, comments, possibly an image, some user id, a time stamp and possibly more.

3: Could dynamic updating of content work in the same way? For example if someone wanted to add a torrent to the pool, how, in the blockchain style, might it be able to be flagged as good or bad?

4: As an addition to the above question, what could be done to dilute pollution to the blockchain (and its subsequent bloating in size), by parties interested in destroying the project?

5: What components that are currently a part of the main Bitcoin client and protocol that would absolutely not be necessary for this? I presume that various things about the transactions are unnecessary, but perhaps for the security of the network the need only be re-purposed?

Thank you very much for your time!

I think what you may be asking for is Freenet - http://en.wikipedia.org/wiki/Freenet - with its distributed encrypted data store.

i don't think that solves any of the problems he's asking about matthew.

this is a really cool idea! mods, can we move this thread somewhere else, so that it can be seen by relevant people?

[andrew12]

This sounds like you are trying to create another peer-to-peer network separate from BitTorrent. There already exists a system where you can get many files simply from a hash, DHT and magnet links. You have an interesting idea, but it doesn't seem necessary.

[memvola]

TBP is just an index, so if the idea is to store the data so that users can recreate the index from scratch, a distributed datastore like Freenet or Tahoe-LAFS would do. People can run their own indexing site as a hidden service for convenience of the casual user.

A blockchain based solution is useful if ownership and authenticity is a concern, but it doesn't seem to be in this case.

[paulie_w]

i don't think that solves any of the problems he's asking about matthew.

this is a really cool idea! mods, can we move this thread somewhere else, so that it can be seen by relevant people?

Any idea where it should be moved to?

wherever the discussions of namecoin go, i suppose.

[paulie_w]

i don't think that solves any of the problems he's asking about matthew.

It solves this problem exactly so I don't know what your talking about. Plus someone would have to rewrite the block chain so no one can change it, probably too much work for a non-profit site like them.

no, because he's not asking you to "solve it" in another way. he asked some very specific questions because it sounds like what _he_ wants to do is attempt something like he describes. everything else is off-topic.

can anyone answer those questions?

[paulie_w]

TBP is just an index, so if the idea is to store the data so that users can recreate the index from scratch, a distributed datastore like Freenet or Tahoe-LAFS would do. People can run their own indexing site as a hidden service for convenience of the casual user.

A blockchain based solution is useful if ownership and authenticity is a concern, but it doesn't seem to be in this case.

he wrote:

2: What metadata could and should be reasonably and sensibly added to the blockchain, keeping in mind that with each torrent on The Pirate Bay you have: magnet link, description, comments, possibly an image, some user id, a time stamp and possibly more.

3: Could dynamic updating of content work in the same way? For example if someone wanted to add a torrent to the pool, how, in the blockchain style, might it be able to be flagged as good or bad?

...i wonder, really, if anyone even read what he wrote before responding?

[Jouke]

TBP is just an index, so if the idea is to store the data so that users can recreate the index from scratch, a distributed datastore like Freenet or Tahoe-LAFS would do. People can run their own indexing site as a hidden service for convenience of the casual user.

A blockchain based solution is useful if ownership and authenticity is a concern, but it doesn't seem to be in this case.

he wrote:

2: What metadata could and should be reasonably and sensibly added to the blockchain, keeping in mind that with each torrent on The Pirate Bay you have: magnet link, description, comments, possibly an image, some user id, a time stamp and possibly more.

3: Could dynamic updating of content work in the same way? For example if someone wanted to add a torrent to the pool, how, in the blockchain style, might it be able to be flagged as good or bad?

...i wonder, really, if anyone even read what he wrote before responding?

I think he'd better ask the people at freenet, they know how it is done, because they already did it.

[memvola]

2: What metadata could and should be reasonably and sensibly added to the blockchain, keeping in mind that with each torrent on The Pirate Bay you have: magnet link, description, comments, possibly an image, some user id, a time stamp and possibly more.

3: Could dynamic updating of content work in the same way? For example if someone wanted to add a torrent to the pool, how, in the blockchain style, might it be able to be flagged as good or bad?

...i wonder, really, if anyone even read what he wrote before responding?

Actually yes, but the viability of the blockchain approach affects the answers. I'm not a member of any team, but my answer to the first question would be "conceivable, but a more generic distributed datastore is more suitable than a blockchain".

Block chain is a solution invented to specifically solve the problem of double spending. It maybe isn't even ideal as a DNS system, though namecoin is doing fine because it handles very limited amount of data, and makes use of the extreme data redundancy. If namecoin allowed contractual exchanges through the block chain, it would be the ideal use of it, but I digress. At least let me add that you don't need the extreme data redundancy either.

Since we also don't need encryption at the store level, and instead need relaying of data (encryption goes on here), neither git nor Tahoe-LAFS seem to be ideal candidates either. Both would need to be used under a pseudonymous network layer, such as I2P's Tahoe-LAFS store.

Another important point is, how the filtering would be handled. This also determines how a potential voting system would work.

It's easier to implement the real world user hierarchy and maintain it as part of the data itself. That would mean that the final index would be consistent and from the perspective of the almighty Pirate Bay. The keys would be managed externally by a group of people, top-to-bottom, with real world liabilities attached. This is also the best solution if you really really want to avoid the data you don't want there.

Otherwise you could implement the system from the perspective of the index-generator, in which case you would require a web of trust mechanism.

With Freenet, you could use the already implemented WoT system and implement a filter on top of it. Persistence of data would need to be actively handled by this system. All updates, including votes, are new inserts. Spam or otherwise unwanted data will not be healed and die off, though the WoT will probably tend to compartmentalize the indexes depending on the unwantedness of certain types of data (if you get my drift). If you can generalize magnet links, this could serve as an index for Freenet as well, though I guess your intended use case is (or, in fact, should be) web-facing index sites using Freenet as a datastore.

Tahoe-LAFS supports user permissions and healing data is pretty straighforward, but I don't know of any hierarchical trust mechanisms implemented for it. If you intended to manage it top-down though, it's the easiest way to increase resilience, and it is much lighter than Freenet.

ETA: By the way, namecoin supports name hierarchies and all kinds of URI's, so you could easily generate an index using the namecoin database as it stands. A trust/permission mechanism would not be too hard to implement either, and it can even make use of the name hierarchy without any further improvement. The problem is that the additional data you will need for images/descriptions and updates in the case for comments/votes will cost a lot. Maybe it could be used in combination with a DHT. I'm not following namecoin development closely, a DHT was being discussed. In effect, you would get the crucial metadata from namecoin, and download the extra data from the bittorrent network. Maybe this answers the question better than the rest of my comment.

[dego]

By the way, namecoin supports name hierarchies and all kinds of URI's, so you could easily generate an index using the namecoin database as it stands. A trust/permission mechanism would not be too hard to implement either, and it can even make use of the name hierarchy without any further improvement. The problem is that the additional data you will need for images/descriptions and updates in the case for comments/votes will cost a lot. Maybe it could be used in combination with a DHT. I'm not following namecoin development closely, a DHT was being discussed. In effect, you would get the crucial metadata from namecoin, and download the extra data from the bittorrent network. Maybe this answers the question better than the rest of my comment.

You got it. Thats exactly how it could work. There could be a new BT/NMC Client which charges a small amount of MNC of the downloading person and sends it towards the one who created the metadata update on the namecoin blockchain (which costs some NMCs).

There's a discussion about the possible uses of Namecoins, where khal, the main dev is involved: http://dot-bit.org/forum/viewtopic.php?p=2618#p2618 Using it as a torrent tracker is also mentioned.

[lophie]

I think what you may be asking for is Freenet - http://en.wikipedia.org/wiki/Freenet - with its distributed data store.

+1, Already made (well in a similar manner). But I think a whole blockchain for 1 website is an overkill. But again thinking about a universal torrent tracker....... <3

[dego]

Its about taking the (namecoin) blockchain to distribute the tracker database and store the magnet links...

[inaltoasinistra]

Seems too heavy to keep all contents into the blockchain :O
What about to use a DHT for all contents: links, rates, comments, images; every content is signed by a namecoin id/ . Every desktop-app-node can hide/not-memorize content by malicius isd (spam, not liked contents)

I don't understand the advantage to have all content into blockchain

[maxsolnc]

Datacoin allows such possibility today

[phelix]

Datacoin allows such possibility today

Namecoin has been allowing it for two and a half years.