[rfc-i] "canonical" URI for RFCs, BCPs

[Joe Touch]

John R Levine wrote:
>> No need - I think we're all in agreement that the URI should look like:
>>
>>     http:..../rfcnnnn
>>
>> AND that the URI should return the authoritative version (currently text
>> or postscript) as currently known.
> 
> No, we're not, but since you seem to feel so strongly that it's
> important to return unpredictable data, I'm not going to keep arguing.

OK, so now we need it to be "predictable"?

I'm confused. If I put in the suffix, it's definitely predictable. But
you don't want that.

So maybe you want something that can change? That may be the foundation
of our disagreement. I do not want mutability.

> Note that type negotiation doesn't do what you want; all of the
> Postscript RFCs have text stub versions, and if a web browser says it
> prefers text to postscript, it'll get the text.  You really need a
> container that points to the various versions (no doubt with a flag of
> some sort in case it's not obvious which version is authoritative) so
> people can find what they need.

What I want is for http:.../rfcnnnn
to return the canonical version. I don't care what the browser asks for
unless there are truly alternate authoritative versions (there aren't,
AFAICT - even for the postscript, the derived PDF isn't considered
authoritative).

>> There can be other URIs that point to metadata, e.g.:
>>
>>     http:.../info/rfcnnnn
>> or
>>     http:.../rfcnnnn/info
> 
> OK, so info/rfcnnnn is the useful canonical URL which returns a
> container with pointers to the various stuff.  Close enough.

As I said, the info/ URI is fine if that's what I want to point to.

When I want to point to an RFC, that is NOT what is canonical.

Joe

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: OpenPGP digital signature
URL: <http://mailman.rfc-editor.org/pipermail/rfc-interest/attachments/20100201/8efa0e90/attachment.sig>