[ih] STD 7, RFC 9293 on Transmission Control Protocol (TCP)

[touch at strayalpha.com]

Notes below..

—
Dr. Joe Touch, temporal epistemologist
www.strayalpha.com

> On Aug 19, 2022, at 1:42 PM, John Day <jeanjour at comcast.net> wrote:
> 
> 
>> On Aug 19, 2022, at 15:16, touch--- via Internet-history <internet-history at elists.isoc.org> wrote:
>> 
>> It was a lot of work and a lot of the credit goes to Wes, but FWIW the task was scoped to NOT change the spec or TCP, but rather just roll in changes that have accumulated.
> 
> That is abundantly clear!  At least to those of us who know the spec inside out.
>> 
>> It wasn’t a “hey, let’s fix what we think needs fixing”… - that work has been happening through various other RFCs developed separately over the past decades.
> 
> O, no. that wouldn’t have been the intent.  But as long as you brought it up, which sort of RFCs did you have in mind that weren’t included?

They’re cited in 9273, but not rolled in. And they keep on coming. But that’s not what I’m referring to - I’m noting that we specifically did NOT try to open “new cans of worms” and solve problems that had not already been documented when the process began 7 years ago.

>> This helps developers start with a better ‘checkpoint’ and not need to keep tracking all the various corrections and errata.
> 
> O, absolutely. It should have been done a long time ago. Developers wanting to do a TCP implementation shouldn’t have to buy one of the books telling you how or search through all of the RFCs to find which ones are relevant.

Note that this issue was partly addressed in 2006 by RFC 4164, which was revised again in 2015. But even a roadmap didn’t help roll in the errata or make the original text more clear, something we only learned over the past decades.

Joe