4 ms·

I really like that IPFS is trying to change the way we think about the internet and HTTP. That being said, I'm very skeptical of a lot of the design choices. It

I really like that IPFS is trying to change the way we think about the internet and HTTP. That being said, I'm very skeptical of a lot of the design choices. It seems like it's just trying to incorporate a lot of the latest buzzword technologies without any real consideration why. I get that blockchain, Git, BitTorrent are all powerful but that doesn't mean that mixing them all together into IPFS is going to be useful. Most likely it will end in a sort of internet Frankenstein's monster: overly complicated and lacking real benefits over traditional HTTP, FTP, and the rest. My biggest concern is that in the end IPFS isn't even really "permanent" in the way I understand it. Objects added to IPFS still need someone to in a sense "seed" them for that content to be available. What advantages does that give over just hosting the internet over static torrents?

_prometheus 11y ago
Take a look at this: "Content Model, or Replication on IPFS" https://github.com/ipfs/faq/issues/47 https://github.com/ipfs/faq/issues/47 -- it's a discussion on WHY we had to separate out the replication part. Check out also: - ipfs-cluster - discussion on a tool we'll build to replicate archives to many nodes and have a RAID-like cluster configuration - https://github.com/ipfs/notes/issues/58 https://github.com/ipfs/notes/issues/58 - ipfs-persistence-consortium - https://github.com/pipermerriam/ipfs-persistence-consortium https://github.com/pipermerriam/ipfs-persistence-consortium - pincoop - https://github.com/victorbjelkholm/pincoop https://github.com/victorbjelkholm/pincoop - http://filecoin.io http://filecoin.io (long term solution) - more at https://news.ycombinator.com/item?id=11137313 https://news.ycombinator.com/item?id=11137313
jokoon 11y ago
> overly complicated and lacking real benefits over traditional HTTP, FTP, and the rest. Authenticity over P2P is complex, that's the cost of having no server. But if such a cost can save you hardware and server hosting, it's worth it. > What advantages does that give over just hosting the internet over static torrents? No need for DNS. That is pretty huge.
- geebro 11y ago
  >But if such a cost can save you hardware and server hosting, it's worth it. The vast majority of this discussion is miles above my head, but this made my Spidey Sense tingle. Someone somewhere has to host the data, and seemingly in more than one place. There is no such thing as a free lunch.
shazow 11y ago
> It seems like it's just trying to incorporate a lot of the latest buzzword technologies without any real consideration why. I get that blockchain, Git, BitTorrent are all powerful but that doesn't mean that mixing them all together into IPFS is going to be useful. While the IPFS whitepaper and spec does utilize a lot of modern technology terms, it does not mean they were assembled without serious consideration and purpose. What gives you that impression? All of those pieces have important properties which create the essence of what IPFS is. When you say Blockchain, Git, BitTorrent, I hear: Directed acyclic graphs (like git) with hashed hierarchical checkpoints (like a blockchain or merkle trees) distributed peer to peer (like bit torrent). This is literally what IPFS is, and using those terms is one way to describe it.

_prometheus 11y ago

Take a look at this: "Content Model, or Replication on IPFS" https://github.com/ipfs/faq/issues/47 https://github.com/ipfs/faq/issues/47 -- it's a discussion on WHY we had to separate out the replication part. Check out also: - ipfs-cluster - discussion on a tool we'll build to replicate archives to many nodes and have a RAID-like cluster configuration - https://github.com/ipfs/notes/issues/58 https://github.com/ipfs/notes/issues/58 - ipfs-persistence-consortium - https://github.com/pipermerriam/ipfs-persistence-consortium https://github.com/pipermerriam/ipfs-persistence-consortium - pincoop - https://github.com/victorbjelkholm/pincoop https://github.com/victorbjelkholm/pincoop - http://filecoin.io http://filecoin.io (long term solution) - more at https://news.ycombinator.com/item?id=11137313 https://news.ycombinator.com/item?id=11137313

jokoon 11y ago

> overly complicated and lacking real benefits over traditional HTTP, FTP, and the rest. Authenticity over P2P is complex, that's the cost of having no server. But if such a cost can save you hardware and server hosting, it's worth it. > What advantages does that give over just hosting the internet over static torrents? No need for DNS. That is pretty huge.

geebro 11y ago
>But if such a cost can save you hardware and server hosting, it's worth it. The vast majority of this discussion is miles above my head, but this made my Spidey Sense tingle. Someone somewhere has to host the data, and seemingly in more than one place. There is no such thing as a free lunch.

shazow 11y ago

> It seems like it's just trying to incorporate a lot of the latest buzzword technologies without any real consideration why. I get that blockchain, Git, BitTorrent are all powerful but that doesn't mean that mixing them all together into IPFS is going to be useful. While the IPFS whitepaper and spec does utilize a lot of modern technology terms, it does not mean they were assembled without serious consideration and purpose. What gives you that impression? All of those pieces have important properties which create the essence of what IPFS is. When you say Blockchain, Git, BitTorrent, I hear: Directed acyclic graphs (like git) with hashed hierarchical checkpoints (like a blockchain or merkle trees) distributed peer to peer (like bit torrent). This is literally what IPFS is, and using those terms is one way to describe it.