4 ms·

Is there any compelling reason to use hypercore (Dat) or IPFS over the other? To me, it seems they both have a similar feature set.

Ericson2314 6y ago
IPLD, and its ability to embed existing references (e.g. git hashes) I think is a good "big tent" approach to get people on board. Relatedly, I think files / encouraging application layer to marshall/unmarshall from flat bytes was always a terrible idea. IPLD encourages not doing that. The most important thing is betting people to agree on how we content-address data (see [1]). At that point, we can just let the network protocols duke it out, using bridges to ensure any fragmentation doesn't hurt network affects so bad. Ultimately, shared addressing scheme -> commodification, something this industry desperately needs, and indeed something humanity needs to make better and fairer use of technology. It's a lot like the identity problem with human-oriented distributed systems, just way easier to solve prior to coordinating better since the actual content addressing (as opposed to agreeing on the scheme) is trustless and coordination free. [1]: https://www.softwareheritage.org/2020/07/09/intrinsic-vs-extrinsic-identifiers/ https://www.softwareheritage.org/2020/07/09/intrinsic-vs-ext...
- hardwaresofton 6y ago
  For those who didn't immediately know what IPLD is: https://docs.ipld.io/ https://docs.ipld.io/ So similar to JSONLD (JSON Linked Data) -- if you know what the schema of a blob it can help with pre-processing... this seems solvable as an add-on but being included is a difference.
anchpop 6y ago
Dat makes it easy to update your dats (just be careful not to lose your private key). That's not possible with IPFS - you have to use IPNS. IPNS used to be slow - I'm not sure if it still is, but I'd be surprised if it's as fast as Dat.

4 ms·

Is there any compelling reason to use hypercore (Dat) or IPFS over the other? To me, it seems they both have a similar feature set.

by MayeulC 6y ago

Ericson2314 6y ago
IPLD, and its ability to embed existing references (e.g. git hashes) I think is a good "big tent" approach to get people on board. Relatedly, I think files / encouraging application layer to marshall/unmarshall from flat bytes was always a terrible idea. IPLD encourages not doing that. The most important thing is betting people to agree on how we content-address data (see [1]). At that point, we can just let the network protocols duke it out, using bridges to ensure any fragmentation doesn't hurt network affects so bad. Ultimately, shared addressing scheme -> commodification, something this industry desperately needs, and indeed something humanity needs to make better and fairer use of technology. It's a lot like the identity problem with human-oriented distributed systems, just way easier to solve prior to coordinating better since the actual content addressing (as opposed to agreeing on the scheme) is trustless and coordination free. [1]: https://www.softwareheritage.org/2020/07/09/intrinsic-vs-extrinsic-identifiers/ https://www.softwareheritage.org/2020/07/09/intrinsic-vs-ext...
- hardwaresofton 6y ago
  For those who didn't immediately know what IPLD is: https://docs.ipld.io/ https://docs.ipld.io/ So similar to JSONLD (JSON Linked Data) -- if you know what the schema of a blob it can help with pre-processing... this seems solvable as an add-on but being included is a difference.
anchpop 6y ago
Dat makes it easy to update your dats (just be careful not to lose your private key). That's not possible with IPFS - you have to use IPNS. IPNS used to be slow - I'm not sure if it still is, but I'd be surprised if it's as fast as Dat.