HTTP Discovery Taking Resource URI and finding metadata
- Should be able to get the metadata without accessing the resource.
- You don't know what it is yet, so you shouldn't be interact with it yet.
- The the resource can point at its metadata.
- Must be accessible to different levels of developers.
- Can't *just* require headers. Must also support an HTML solution.
- Must work with web architecture.
- Must be friendly to little and big players.
- The same solution for your blog should work for a large company.
Background for the following table is at Eran Hammer-Lahav's blog post.
|Resource Declaration||Direct Metadata Access||Web Compliant||Scale Agnostic||Extendable|
|1. HTTP Response Header (Link, X-XRDS-Location)||+||+|
|2. HTTP Response Header over HEAD||+||+||+|
|3. HTTP Header Negotiation||+||+/-|
|4. HTML <Link> Element (<Meta>)||+||+||+|
|5. HTTP Content Negotiation (Accept)||+|
|6. HTTP OPTIONS Method||+||+||+|
|7. WebDAV PROPFIND Method (MGET, ARK)||+/-||+/-||+/-|
|8. Custom HTTP Method||+||+|
|9. Static Resource Mapping (Prefix, Suffix)||+||+/-||+/-|
|10.Dynamic Resource Mapping (Templates, /site-meta)||+/-||+||+||+/-||+/-|
A combination of 1, 4 and 10 seem promising.
For dynamic mapping:
b. Known Location (like /robots.txt)
b is the /site-meta proposal.
This session was also written up on Drummond Reed's blog post XRD Begins.