Can we really label acronyms and abbreviations "well-known" in an archival series?

[ajeanmahoney]

Section 3.6 states:

Abbreviations should be expanded in document titles and upon first
use in the document.  The full expansion of the text should be    
followed by the abbreviation itself in parentheses.  The exception is
an abbreviation that is so common that the readership of RFCs can be
expected to recognize it immediately; examples include (but are not
limited to) TCP, IP, SNMP, and HTTP.  

But can we predict what will be well known and understandable in the future?

Perhaps the guidance should be simpler: expand all abbreviations and acronyms on first use.

[jrlevine]

I think that it makes more sense to keep maintaining a glossary of well known acronyms and we can figure out how to ensure that the glossary gets archived with the documents. Spelling out HyperText Transfer Protocol and Transmission Control Protocol and Simple Network Management Protocol in every Request For Comments document would be quite annoying.

[huitema]

The question is where to set the "well known" bar. HTTP, TCP, even SNMP certainly meet that bar, but there are other examples in the list that probably don't. Take for example ATM:

ATM       *- Asynchronous Transfer Mode (ATM) 
          *- Automatic Teller Machine (ATM)

That's a great example of an acronym wiped out in the dustbin of history. In the early 90's, ATM networks were actively investigated by telecommunication providers, and yes, in a networking context nobody would mistake the reference for a cash dispenser. But ATM development was pretty much abandoned more than 20 years ago. Modern students reading that will most probably never have heard of that technology, and will wonder why we are discussing teller machines.

But it not just history, it is also about references that appear only in a very narrow field, like for example BDS:

BDS   *- Buchmann, Dahmen, Schneider (name of an algorithm)

If readers do a search for BDS, they can scroll through many pages of responses without finding a mention of "Buchmann, Dahmen, Schneider". The Wikipedia disambiguation page for BDS does not mention it either. I understand after searching for "Buchmann, Dahmen, Schneider" that the "BDS Algorithm" is a Merkle Tree traversal algorithm used when implementing some Post Quantum Cryptography algorithms. RFC 8391 mentions it as "the BDS algorithm [BDS08]", and that's fine because the citation explains it. But those initials should never be used without a citation or an explanation. RSA it is not, or at least not yet. (And arguably RFC 8391 should have added the citation on the first reference, not on the second one.)

I only have an amateurish idea of possible solutions. There are clear-cut cases like:

FIFO      *- First In, First Out 
                [This is listed in Merriam Websters Dictionary.]

If the acronym is listed in a mainstream dictionary, we are indeed probably safe. We are also safe with acronyms of established IETF standard. But what do I know?