Summary

  • RFC 3044 wrapped an existing ISSN in the URN framework so a serial publication could be named independently of the current web location of its record or electronic edition.
  • The RFC explicitly warned that stored URLs were not persistent. Its demonstrator therefore included a “URL Watcher” to find broken or redirected addresses; the proposed global discovery system was still future work.

A magazine can change hosts without changing its title. A journal can stop publishing yet remain part of the bibliographic record. RFC 3044, published in January 2001, tried to separate that durable identity from the address at which a reader might currently find it. Its example was compact—urn:ISSN:1560-1560—because the existing International Standard Serial Number already identified a serial publication.

That reuse mattered. The ISSN system already linked a number to a constructed “key title”, maintained records for ceased publications and never reassigned an ISSN. RFC 3044 treated that persistence as a good fit for a Uniform Resource Name. Its syntax rules made comparison mechanical: an optional hyphen could be dropped, and a check-digit x normalized to uppercase. A reader or application could carry the serial’s identifier without treating a publisher’s present URL as its identity.

But a URN that resolves to an obsolete URL can still leave a reader stranded. RFC 3044 was unusually direct about the gap: the URLs stored in the resolution mapping were not guaranteed to persist. They had to be checked and updated. The proposal’s “URL Watcher” detected addresses that were broken or redirected and recorded their state so resolution could use a more relevant mapping. The watcher did not make an address permanent; it made maintenance visible as part of the service.

The resolver was described as a system already developed at the ISSN International Centre, but the abstract called it a demonstrator intended to test what operational deployment would require. One central server mapped an ISSN to one or several URLs: one could lead to bibliographic metadata, another to the online serial itself. The next step—global resolution through a Resolution Discovery System such as NAPTR—was still an evolution to consider, and the RFC called global resolution a major issue. A demonstrator and a future architecture were not the same deployment claim.

Identifier, description and destination therefore ran on different clocks. The serial could retain its ISSN after it ceased; the record could preserve its title and relationships; a URL could break, redirect or be replaced. The number did not have to be reassigned when an address changed. But persistence at the identity layer did not itself supply retrieval at the access layer.

There is a useful present-day comparison, with an important date boundary. The ISSN International Centre now describes ISSN-L as a linking identifier that groups medium-specific ISSNs for print and online editions. That later/current relation helps databases collocate editions; it should not be projected backward as part of RFC 3044’s 2001 demonstrator. Separately, today’s IANA formal URN list contains an issn entry, while the IETF Datatracker records RFC 3044’s move to Historic in 2017. These answer different questions. Neither proves that the demonstrator or its proposed global resolver is operating today.

Read through Heng Lu’s Note 64 only as a design lens—not as a source about the ISSN project—the story sharpens: a coordination artifact may define a useful reference, but publication alone is not deployed reality. RFC 3044 itself separated what its team had built from what it still expected to integrate into global discovery. Its enduring lesson is practical rather than triumphant: a persistent name can make a record easier to carry, but access remains a maintained relationship among the record, the resolver and locations that change.

Sources