orignal
zzz, question for you. When do you send termination reason 13 in NTCP2?
orignal
I send if RI in SessionConfirmed has wrong time
zzz
orignal, yes, usually for old or future RI, but could also be signature verify fail. you could also validate NTCP2 address params there like the length of 'i'
zzz
eyedeekay, ping re: Monero CCS
orignal
then why don't we change the specs
orignal
saying it's not just ServerError3
orignal
but like "malformed RI"
zzz
happy to do it, send me a patch to ntcp2.rst
orignal
can you also doble check how you handle NTCP2 address without i
orignal
?
orignal
onon compains that he can't establish outgoing NTCP2 session
orignal
if NTCP2 only and not published
orignal
he gets code 13 back for SessionConformed
orignal
if NTCP2 is with address(and i) everything works fine
zzz
no i should be fine
orignal
thanks
zzz
if you can prove it's a java bug let me know ))
orignal
we are just investigating
zzz
ok
orignal
and it's about SessionConfirmed only
orignal
never mind. that's because he published 0.0.0.0
zzz
ok, thanks for the info.
eyedeekay
zzz currently afk for an unavoidable emergency, will be online around 2pm, I owe them a revision to the CCS proposal removing i2p-zero specific parts already
zzz
eyedeekay, let's discuss first, holler when available, I have some thoughts
eyedeekay
OK will do
orignal
but why don't you send code 15 if RI signature fails?
orignal
it would be nice to have NTCP2 termination reasons sorted out
zzz
that's just the way it worked out, sig check is off in other code, near where we check the RI time. And it's not like sig fail is normal anyway
eyedeekay
Sorry about that, cable was out when I got back to keyboard. I'll be here for the rest of the evening
zzz
ok
zzz
so I don't think you told me you put up the PR? just discovered it this morning
zzz
only two weeks now until monerocon
zzz
anyway
zzz
none of the big guns have commented yet
zzz
the ofrn guy that nacked it isn't one either, and from looking at the comments on his own CCS, people aren't so thrilled with him
zzz
I'm not sure there would be any pleasing him, so I don't suggest making any changes in response to his comments alone, at least not yet
zzz
we need some of the main people to comment first
zzz
the i2p=zero parts are important to leave in as that's their current recommendation. our CCS is definitely not just a "docs level change"
zzz
eot ^^ eyedeekay
eyedeekay
Ack I'll leave the zero stuff in
eyedeekay
re:the CCS I thought I announced it here but maybe not
eyedeekay
I have been preparing to litigate it being more complicated/involved than docs level changes, my way of tackling that has been to try and encourage them to align monero's ideas about transactions anonymity with the tooling SAMv3 provides
zzz
yeah no amount of rewording will convince the "everything is fine" crowd
zzz
I did one tweet this morning but will try some more to get the heavy hitters over
zzz
read the comments in that guy's CCS here: repo.getmonero.org/monero-project/ccs-proposals/-/merge_requests/457
eyedeekay
Hm. Yeah I see why it gives people pause
eyedeekay
Re: everything is fine people, I'm worried I'll have to prove them wrong, maybe I can, maybe I can't
eyedeekay
There's this thing in Monero called a black marble attack, I have a hunch that if you can actually *succeed* at a black marble attack, and somebody you de-anonymize at the Monero level are using a long-term destination, you may also gain information useful for future attacks, so potentially I can convince them they have something to gain from transient dests
eyedeekay
Actually I think they might have a *lot* to gain from that
zzz
the everything-is-fine people won't be a problem if it gets strong support from the serious people
zzz
maybe link the PR on reddit? ours or monero's or both?
eyedeekay
Sure I'll promote it a bit
zzz
my recent 'don't do i2p over tor' tweet got a fair amount of action, might be worth a x-post to reddit also
eyedeekay
Can do
zzz
thx. any i2p+bitcoin tweets I do get the most action of all, maybe not surprising
zzz
the vasild private transaction bitcoin PR would be a big step forward, seems like it's getting close to merge but hard to tell
zzz
I think we'd get some mileage out of it if it does go in