[rfc-i] The alternateURI element in v3

Julian Reschke julian.reschke at gmx.de
Thu May 8 09:39:11 PDT 2014


On 2014-05-08 18:25, Paul Hoffman wrote:
> On May 8, 2014, at 9:04 AM, Julian Reschke <julian.reschke at gmx.de> wrote:
>
>> On 2014-05-08 16:50, Paul Hoffman wrote:
>>> ...
>>
>> Summarizing, here's my alternative (:-) proposal:
>>
>> 1) Add a <link> element as child of <rfc> (that can occur multiple times.
>>
>> 2) The link's target attribute is a URI.
>>
>> For the ISSN, use the ISSN URN namespace.
>>
>> For the DOI, use the DOI's URI.
>>
>> Both can be identified by inspecting the URI scheme or URI scheme + URN namespace.
>>
>> If typing is needed, use link relations as per RFC 5988 (in an additional "rel" attribute).
>>
>> Thus,
>>
>>   <link target="urn:issn:2070-1721"/>
>>   <link target="doi:whatever"/>
>
> This is clean for ISSN and DOI. But, I thought you wanted individuals to also be able to say in drafts "here is a URL for my own representation of this draft". If that is still true, the value of "target" would be that URL. We could say that we only define semantics for urn:issn and doi:, not other URI. Is that reasonable?

At some point, we need to decide whether we just want to allow 
"alternate" versions of the document (and which), or just links in general.

If we do more than "alternate" versions, we would indeed need a link 
type ("relation type"  in RFC5988-speak).

WRT to "my own representation of this draft" -- this is likely something 
that wouldn't be allowed in the canonical format, right?

So yes, we could say that *if* the target URI matches a specific syntax, 
it has a special role. For "urn:issn", it would end up in the 
boilerplate, replacing the current default value of "2070-1721".

For DOIs I believe we currently do not have a place to put it, right?

Best regards, Julian




More information about the rfc-interest mailing list