I asked the same question of TorrentFreak a few years ago:
Ernesto ernesto@torrentfreak.com
2/16/08
to Christopher
Hi Christopher,
No, I never heard of something like that, but I like the idea
There is http://www.coralcdn.org/, but that's not a complete solution
Cheers, Ernesto
On Feb 16, 2008 9:56 PM, Christopher Froehlich <somecallmechief@the-shades.net> wrote:
Christopher Froehlich <somecallmechief@the-shades.net> wrote:
Question: do you know of anyone developing distributed website platforms? For
example, if I'm seeding and leeching a raw text file–as long as one other
person in the world is doing the same, that file can never be erased. If
someone were to share a PHP/MySQL website, with Apache perhaps integrated into
the BT client–the website could never be removed.</p>
I've been looking into places like Freenet, services like OpenAFS, and the
potential for extensible BT clients like uTorrent or Azureus; but nothing seems
to be available. A website cluster, shared P2P would be indestructible.</p>
Have you seen anything of the like?
Sadly, I haven't seen any substantive leaps since. Even TPB's recent cloud move is entirely predicated on moving a single, virtual instance about.
Edit: Personally, I see this as the most important problem to solve for a free Internet in the next 20 years. Distributed DNS against distributed content is going to be critical.
I see distributed DNS to be the more difficult problem (because when you have that, the content will follow fairly easily).
The current DNS providers are mostly for profit (? - i m talking about the registras), and they ahve no incentive to move or provide the leeway to move, to a system which is harder to profit off.
BitTorrent is OK for files whose content remains static over time, like a video. But if you had a web page that needed to be updated over time, wouldn't having multiple, distributed copies of it make it difficult to ensure that everyone had the most recent version? Who would have the right to make updates? And what if there were conflicting updates?
Imagine a web where all links were magnet links. Each site could also have a magnet link to another site containing a diff in case of updates which the browser would automatically check.
The content hash in the magnet link would have to be based on something other than content (because content is not known at the time of page creation). The magnet link could contain a cryptographic hash of a secret string that only the owner possessed.
haha yep, i thought of this like 6-8 years ago. from what I've sorta thought of in conjunction with your statements, I'd like to add that there's numerous processing and hacking issues and whatever else. It's extremely insecure. The closest we will get to this is the "cloud".
I love the idea for one particular reason. It has 'curation' natively built in. Worthless stuff (which isn't popular, accessed, shared, commented) would be automatically filtered out. Sure there's risk that Internet would change in homogenic pop-pulp but it can be avoided.
9 comments
[ 3.1 ms ] story [ 31.3 ms ] threadErnesto ernesto@torrentfreak.com 2/16/08
to Christopher Hi Christopher,
No, I never heard of something like that, but I like the idea There is http://www.coralcdn.org/, but that's not a complete solution
Cheers, Ernesto
On Feb 16, 2008 9:56 PM, Christopher Froehlich <somecallmechief@the-shades.net> wrote: Christopher Froehlich <somecallmechief@the-shades.net> wrote: Question: do you know of anyone developing distributed website platforms? For example, if I'm seeding and leeching a raw text file–as long as one other person in the world is doing the same, that file can never be erased. If someone were to share a PHP/MySQL website, with Apache perhaps integrated into the BT client–the website could never be removed.</p>
I've been looking into places like Freenet, services like OpenAFS, and the potential for extensible BT clients like uTorrent or Azureus; but nothing seems to be available. A website cluster, shared P2P would be indestructible.</p>
Have you seen anything of the like?
Sadly, I haven't seen any substantive leaps since. Even TPB's recent cloud move is entirely predicated on moving a single, virtual instance about.
Edit: Personally, I see this as the most important problem to solve for a free Internet in the next 20 years. Distributed DNS against distributed content is going to be critical.
The current DNS providers are mostly for profit (? - i m talking about the registras), and they ahve no incentive to move or provide the leeway to move, to a system which is harder to profit off.
Imagine a web where all links were magnet links. Each site could also have a magnet link to another site containing a diff in case of updates which the browser would automatically check.
The content hash in the magnet link would have to be based on something other than content (because content is not known at the time of page creation). The magnet link could contain a cryptographic hash of a secret string that only the owner possessed.
https://freenetproject.org/