Re: [tcpm] draft-eggert-tcpm-historicize-00

[Lars Eggert <[e-mail]>]

Hi,

On 2010-6-9, at 20:35, Hagen Paul Pfeifer wrote:
> Lars, I thought about RFC 1145 "TCP Alternate Checksum Options", there are no
> real shortcomings in the RFC. It is not _widely_ deployed but there is no real
> security concern like say T/TCP.

true. But nobody has bothered to ship it in twenty years, AFAIK.

> I thought only superseded or defective RFCs can be declared historic?

See Section 4.2.4 of RFC2026: 

   A specification that has been superseded by a more recent
   specification or is for any other reason considered to be obsolete is
   assigned to the "Historic" level.

RFC4614 has classified this extension as Historic for "lack of interest" in 2006. This ID just nails the coffin shut.

> I mean
> TCP Alternate Checksum Options _can eventually_ useful in the future for
> example Interplanetary TCP.

Sure, but it hasn't been in use in twenty years, and the Interplanetary TCP is also not immediately around the corner.

> Maybe some military sites already employ RFC 1145.

If you know of deployments, I'm all ears. But even if we do declare it Historic, that does *not* mean that folks that are running it shouldn't run it anymore. The code points, for example, do remain assigned.

Lars