RFC 209

Host/IMP interface documentation

πŸ“… published August 1971, from ietf

HOST/IMP Interface Documentation πŸŽ€

okay bestie, this one is from B. Cosell at Bolt Beranek and Newman (NIC 7187) and it's dated 13 August 1971 πŸ¦‹ It's RFC 209, and the header doesn't list a category or say it obsoletes or updates anything, so it's just a standalone doc πŸ’…

On 25 June 1971, BBN put a new version of the IMP system up in the network, Version 2500 🌸 Among other changes, this version fixed the problem where Hosts kept getting spurious "Destination Dead" [Type 7] messages πŸ™ˆ The fix was that the IMP refused to accept any messages from its Host for about 30 seconds after the Host said it was about to come alive. Ngl, that caused drama for a bunch of NCPs. They had put an upper limit on how long an IMP could take to accept a message, so when the IMP went past it, the IMP got declared inoperative and the NCP shut itself down πŸ’€ site personnel were annoyed and confused, and honestly, same.

BBN was like "we're sorry we caused so much trouble 😭" but they really thought the new version didn't change any advertised property of the system πŸ₯Ί So to stop this kind of misunderstanding from happening again, the IMP/Host interface should be well documented, well understood, and invariant ✨ The only way to make and keep that kind of interface is to assume that BBN Report No. 1822 is complete and correct. BBN knows it isn't, but treating it as correct is the best way to push 1822 to get better πŸ”₯

So here's the vibe for Hosts: anything that 1822 doesn't specifically guarantee does not exist and should not be used 🫢 If you run into problems with this, don't "code around" them, and don't treat your IMP as a "black box" and guess how it works from a bunch of experiments πŸ™…β€β™€οΈ Instead, send your comments and problems to the NCC at BBN. They'll fix the IMP system or 1822, whichever one actually needs it πŸ‘‘

Also, BBN was working on an update to 1822 that will reflect the removal of the "cease" mechanism [as per RFC #107] 🧚 They were also thinking about adding a mechanism to the IMP/Host protocol so a Host can detect a change in the IMP system's "version" and get status info about who is alive in the net 🌈 main character energy for the network, fr fr ✨


[ This RFC was put into machine readable form for entry ] [ into the online RFC archives by Ryan Kato 6/01]
πŸ“¦ where this page comes from