Re: [tcpm] I-D Action: draft-ietf-tcpm-rfc793bis-16.txt

["Scharf, Michael" <[e-mail]>]

Hi Joe,

Regarding question 3): For header bits, at least as far as I know, there is no unauthorized use as of today. My comment was about authorized use.

Bit 7 is reserved now but was officially allocated to NS before. So, question 3) is about whether and where to record that *official* use of bit 7.

Right now, https://www.iana.org/assignments/tcp-header-flags/tcp-header-flags.xhtml records the past use in the table column “Name”. That works as of today, but if bit 7 is reallocated, the question is whether and where to record a past and completed assignment. Even more tricky is the question how to record the use of bits allocated to RFC 3168 by newer RFCs, e.g., as foreseen by AccECN. All of this is about official IETF allocations by future RFCs.

As I have now looked at https://www.iana.org/assignments/service-names-port-numbers/service-names-port-numbers.xhtml, I see that this registry has a column “Assignment Notes”. This would also be a reasonable name for a new column in the header flag registry. If we decided to go down that road, the IANA table after publication of 793bis could look like more or less as follows …

Bit          Name                                                                   Reference          Assignment Notes
4             Reserved                                                           [793bis]
5             Reserved                                                           [793bis]
6             Reserved                                                           {793bis]
7             Reserved                                                            [RFC8311]           Previously used by Historic [RFC3540] as NS (Nonce Sum)
8             CWR (Congestion Window Reduced)      [RFC3168]
9             ECE (ECN-Echo)                                                [RFC3168]

As I wrote, I think that change could be done by 793bis, as it is purely editorial, but I don’t have a very strong feeling whether this is really MUST HAVE.

Unfortunately, it is possible that there could be future unauthorized use of header bits, e.g., if the IETF decided not to standardize one solution or several solutions that claim header bits. I agree that we could use then a term such as “known unauthorized use”. But right now I think how to deal with that problem would be outside the scope of 793bis.

Michael




From: Joseph Touch <[e-mail]>
Sent: Wednesday, March 25, 2020 10:31 PM
To: Scharf, Michael <[e-mail]>
Cc: Wes Eddy <[e-mail]>; [e-mail]
Subject: Re: [tcpm] I-D Action: draft-ietf-tcpm-rfc793bis-16.txt


Hi, all,

I didn’t see these proposed ways forward before, so I’ll add my preferences to them in particular below.

Joe


On Mar 25, 2020, at 1:23 PM, Scharf, Michael <[e-mail]<mailto:[e-mail]>> wrote:

Hi Wes,

Regarding 1), I think that has been some (supportive) list discussion on cleaning up the IANA registry.

To provide some data point, here is how I would answer the first two questions in my summary – as usually as individual contributor:


Question 1: Should an IANA registry be created for bits 4, 5, and 6 with status "Reserved", which are missing on https://www.iana.org/assignments/tcp-header-flags/tcp-header-flags.xhtml?

=> Michael’s opinion: Yes, I think this should be done in 793bis

+1

Question 2: Should the IANA registry move to a sub-registry on the Transmission Control Protocol (TCP) Parameters page (https://www.iana.org/assignments/tcp-parameters/tcp-parameters.xhtml).

=> Michael’s opinion: Yes, I think this should be done in 793bis

+1

A little bit more tricky is the third question, which mostly matters for the entry created by RFC 8311…

Question 3: Should past (and current) use of bits be documented in the IANA registry, and, if so, how?

=> Michael’s opinion: We could discuss whether to change the table structure (see https://www.iana.org/assignments/tcp-header-flags/tcp-header-flags.xhtml#tcp-header-flags-1) to record past use, e.g., by adding another column and moving the statement “previously used by Historic [RFC3540] as NS (Nonce Sum)” to that new column. The new column could have a title such as “Comments”, “Past Use”, etc. That would be a purely editorial change, as it would just reorganize the table, but not change any entry. I don’t have a particularly strong opinion on that, though. Even if we decided that this was useful, we could also postpone that change to the first standards-track document that actually needs this, e.g., by reallocation bit 7 to a new use. Until this happens, the current table format still works. So my suggestion would be to do nothing unless others from the community strongly support a change by 793bis.

I favor the approach we use for IANA ports here.

Document official assignments officially.

Document unofficial uses vaguely - enough to say “someone uses it” but NOT specifically who or how. The latter ends up effectively endorsing squatting.

IANA does this by saying “known unauthorized use”.


Joe