- From: Melvin Carvalho <melvincarvalho@gmail.com>
- Date: Tue, 20 Jan 2015 23:20:42 +0100
- To: Kjetil Kjernsmo <kjetil@kjernsmo.net>
- Cc: public-rww <public-rww@w3.org>, public-webid <public-webid@w3.org>
Received on Tuesday, 20 January 2015 22:21:11 UTC
On 20 January 2015 at 23:03, Kjetil Kjernsmo <kjetil@kjernsmo.net> wrote: > On Friday 16. January 2015 09.47.03 Melvin Carvalho wrote: > > 1. The RDF Dataset Normalization spec > > 2. The Linked Data Signatures spec > > 3. The Security vocabulary > > Very cool! > > I've been looking for a simple way to create a digest of RDF for the > purpose of creating HTTP ETag validators. This appears to be overkill for > the purpose as it doesn't need to be cryptographically strong, it just > needs to be pretty good so that if you have two messages that have the > semantics, they will probably have the same digest. > > But I suppose that since you already wrote the code, is this something > that could be reused for the purpose? > I do this alot. In fact I also use this to turn bnodes into UUID's, in particular, sha-256 of the normalized RDF. You can use the jsonld normalize() function in node for example. I think have ni:///sha-256;<hash> URI's which are normally dereferenced at (according to the named instance spec) /.well-known/ni/sha-256/<hash> <-- notice the decentralization here the origin is independent of the data! > > Cheers, > > Kjetil >
Received on Tuesday, 20 January 2015 22:21:11 UTC