[rfc-i] In defence of the ISSN number

John C Klensin john+rfc at jck.com
Sat Aug 29 03:08:06 PDT 2009

--On Sat, 29 Aug 2009 13:15:47 +1200, Brian E Carpenter
<brian.e.carpenter at gmail.com> wrote:

>> Before we encourage people to use the ISSN number with the
>> RFC Series, it would be important to know whether or not it
>> is acceptable to associate an ISSN number with something that
>> was published before the ISSN number was issued (which in our
>> case is everything more than about a year old). Yes, this is
>> pedantic, but the ISSN folks may be more pendantic than we
>> are.
> Bob Braden can confirm this, but my understanding is that it
> can be used retroactively. It's only a label, not really an
> indicator of status, but it's a label that is understood by
> the library community in particular.

I'm not Bob, but I am one of the people who pushed for ISSN

It identifies the series, not a particular issue, and is not
tied to when the issue appeared.

As to the comments about references, the ISSN is much more often
used in classifying the document for archival (e.g., library)
purposes and for the avoidance of ambiguity.  It is less often
used in particular citations, although some fields and some
journals are fussier than others.

The bottom line is that there were some requests to register the
series and have the number available.  The people making those
requests will presumably use the number.  Using it within the
RFC series, i.e., to point from one RFC to another (or, by
extension, from an I-D to an RFC) would be fairly silly.  Beyond
that, if a publication for which someone is writing requires or
requests it, use it.  If not, ignore the thing -- I don't think
we are either encouraging or discouraging use of the thing, just
making it available for those who want it.

On 29 Aug 2009 03:48:27 -0000 John Levine <johnl at taugh.com>

> I've published an actual quarterly paper journal with an ISSN
> (1042-5721), and they don't care.  ISSNs are assigned to
> serials, not to individual issues, and it's quite normal to
> get an ISSN for an existing serial.  The point of an ISSN is
> similar to that of an ISBN, so that someone who wants to
> subscribe can find the publisher and buy a subscription.

Please stop after "find the publisher", because "buy a
subscription" is only one reason, albeit an important one.  And
the other point --and really the main one-- is the avoidance of
ambiguity.  Remember that we do not own an
internationally-recognized, "well-known", trademark on the term
"RFC" or even "Request for Comment".

> Considering that you can't subscribe to the RFC series, and
> they're given away for free, there's precious little practical
> use for an ISSN.

You can certainly subscribe to announcements of when new RFCs
are available, with pointers to how to download each one.  That
is a pretty good approximation to a subscription to the series
in this day and age.

> For info on ISSNs in the United States, see
> http://www.loc.gov/issn/
> By the way, that web page says that serials with ISSNs are all
> cataloged in Worldcat.  My ISSN shows up, but the RFC ISSN
> does not, presumably because no RFC is cataloged in any
> library that Worldcat covers.

Or because the libraries that do catalog RFCs (there used to be
several and I'd be surprised if any of them have stopped) have
yet to add the (fairly new) ISSN number to their cataloging
records.  I imagine we could get that fixed if anyone thought it
was important, but the ISSN number does identify the publisher
(more or less) and provide that uniqueness identifier without it.

> I'm wondering, when a library writes to whoever the contact is
> for the RFC series ISSN and asks to buy a subscription, then
> what?

I hope they would be pointed to subscription information for the
RFC announcement list.  If the IAOC were so inclined, the party
involved might reasonably be told about the possibility of
making voluntary contributions to some RFC Editor fund at ISOC.


More information about the rfc-interest mailing list