eyedeekay
I've got the basic design of all the first three handshake phases in what I'm calling "MessageProcessors" an NTCP session uses these as handlers for the handshake phases. I've factored out KDF and AEAD operations into interfaces I can centralize on/use and re-use as well. I've got some disorganization between the noise and ntcp interfaces as a by-product of development which is leading to some slightly
eyedeekay
confusing bugs, but my next step is to go back and refactor the noise library to fit with what I know about the needs now, which is going to be about half subtractive, half moving stuff around to better facilitate the structure of the NTCP2 library. Once I've done that I'll have a very clear idea as to what the rest of the implementation bugs are.
eyedeekay
All of which is very straightforward Go development stuff. Basically, I'm cleaning up, and when I'm done cleaning up, it's going to be pretty much ready.
zzz
nice eyedeekay so it sounds like you're still at the unit test stage?
zzz
skimmed your changes and have a bug report for you
eyedeekay
Fire away with the bug report, got a lot of changes here locally so can't guarantee it'll be the same
eyedeekay
Yeah unit tests still need to be worked on but they'll look sane
eyedeekay
When they're done
zzz
in outbound_handshake it appears you're obfuscating the session confirmed static key; that's not in the spec and won't work
zzz
noise encrypts it for you
eyedeekay
Yup, that's in the changes, copypasta thing on my part I think, sort of did a sprint on it
eyedeekay
Thanks for the report though
eyedeekay
I got excited now that I finally figured out how to build reusable noise parts
zzz
we caught two more very aggressie probers yesterday over in #saltr, once per minute for at least a week, I pushed a new newsxml blocklist, please update the news when you have time
zzz
and I assume this won't be the last time
eyedeekay
Can do