1) If the objective is to have a URN for RFCs this has already been done:
RFC 2648: http://www.ietf.org/rfc/rfc2648.txt "A URN Namespace for IETF
These identifiers must be the canonical identifiers for the RFC series. But
they need not be the only identifiers.
2) Schemes which rely on paying registrars to sell people numbers are probably
unsustainable in the long run unless there is a business reason to use that
This is certainly the case for IP numbers. I don't see the business reason for
this particular application. Hence I don't see a value in purchasing a DOI
identifier at the reported $1500/annum or for accepting one for free use. I
would consider that to be an endorsement and I don't think that the IETF or
ISOC should get any further into that game than it alrady has.
3) Whether the documents are paper or digital is now irrelevant. Dead tree
publication technology will certainly disappear at some point. My book sells in
both paper and Kindle editions. The killer application of Kindle appears to be
sale of periodicals and newspapers rather than just books.
The industry has a clear business need and so they will apply ISSNs to this new
field regardless of what the rules might say on the matter.
4) ISSNs are used in the library system. They are used in the Z39.50 protocol
which is the principal protocol used to support that infrastructure today. I
think we should get one.
5) This topic is a very interesting one and thus one on which a large number of
people may have an opinion. The problems raised in the ESDS BOF are very
Because it is an area where many people may have an opinion it appears to me
that the decisive technical breakthrough we might need in this area might well
be to develop a technology that allows people to have separate opinions in this
area and not attempt to impose more homogeneity than is actually required.
IETF mailing list