Re:An introduction to IPFS
by weavejester on Friday February 19, 2016 @03:20PM
(#11136327)
Real-world applications require backend systems with access control, mutable data, certain information being kept secret, and so on - something that seems fundamentally at odds with the design of IPFS.
I don't think that's at odds with the design.
If you want to keep something secret, you encrypt it. If you want to manage access control, you sign it. A client application can discard information that cannot be decrypted, or isn't signed by an authorised key.
You don't get atomicity guarantees of course, but you can often get away without them. I don't see a system like IPFS as a replacement for all of the web, just a sizeable percentage of it.
I don't think that's at odds with the design.
If you want to keep something secret, you encrypt it. If you want to manage access control, you sign it. A client application can discard information that cannot be decrypted, or isn't signed by an authorised key.
You don't get atomicity guarantees of course, but you can often get away without them. I don't see a system like IPFS as a replacement for all of the web, just a sizeable percentage of it.