@eyedeekay
&zzz
+FreefallHeavens
+R4SAS
+RN_
+Romster
+cims
+hagen
+nilbog
+nyaa2pguy
+postman
+snex
Arch
Danny
Irc2PGuest17692
Irc2PGuest28384
Irc2PGuest78852
Irc2PGuest85111
Onn4l7h
Onn4|7h
Over1
SilentWave
Sleepy
T3s|4_
U1F642
Watson
Wikk_0
Zapek
aargh4
ac9f
acetone
ahiru
anontor
b3t4f4c3__
dr4wd3_
duanin2
eyedeekay_bnc
leopold
mahlay
makoto
marek
mareki2p_
n2
n2_
not_bob_afk2
orignal
poriori
profetik1
qend-irc2p_
r00tobo_BNC
rapidash
rednode
sahil
test7363673
uop23ip
urist_
vivid_reader56
wodencafe2
x74a6
zelgomer
orignal
i2pd uses DSA by default
orignal
for compatibility
orignal
well MLDSA is supported by both Java and i2pd but not officially
orignal
e.g. it's not a problem to connect to zzz.i2p with MLDSA
altonen
zzz i had a q about pq ssu2, specifically the HolePunch message: i2p-projekt.i2p/en/docs/specs/ssu2-hybrid/#long-header
altonen
that message is sent by the Bob but the PQ connection type is selected by Alice so which version number should bob put in HolePunch
zzz
yeah you're right altonen I don't think it matters
altonen
ok
zzz
I'll fixup the docs
orignal
what did I miss?
zzz
ssu2 version in holepunch header
orignal
I though it's always 2
orignal
why do you care?
zzz
spec says set it to 3 or 4 but as altonen says, it's the wrong direction
zzz
so I think we're all agreed, we don't care
orignal
my answer is "doens't matter"
zzz
any version 2-4 is ok
zzz
yup
zzz
just have to fix the docs
orignal
because you take this version either from RouterInfo or SessionRequest
orignal
specs should say "ignored"
zzz
I have some low-level code that enforces 2-4 for all messages, so not "ignored", but "any version 2-4"
orignal
also do you support MLDSA at zzz.i2p?
zzz
right now, yes I think, but not guaranteed forever, depends if I'm building with all my WIP or not
zzz
I have some low-level code that enforces 2-4 for all messages, so not "ignored", but "any version 2-4"
zzz
right now, yes I think, but not guaranteed forever, depends if I'm building with all my WIP or not
orignal
fine just 2-4
zzz
also orignal the IETF LAMPS group changed how hybrid signatures are generated so that has to change
zzz
I haven't changed my code yet for hybrid sigs
zzz
I did update the proposal though to match, for now, we'll see if they change it again
orignal
fine
orignal
but we don't use MLDSA
zzz
you asked about MLDSA....
orignal
does it currently work or I should remobve it for now?
zzz
It should work right now on my routers. But I never checked anything in, and all the WIP is clogging up my git workspace, so I'm thinking about ripping it out for now to simplify my life
zzz
it's probably going to be another year before IETF makes it all official
zzz
pure MLDSA probably won't change but the composite (hybrid) sig stuff is really up in the air
orignal
fine
zzz
and ofc MLDSA breaks SSU2 relay and peer test because the sigs won't fit in a packet, so that's another mess I don't want to deal with. but only for RI sigs
orignal
ofc we shoudln't use one for routers
orignal
rigth now we are talking about destinations only
zzz
yup
zzz
I shouldn't have put MLDSA in prop. 169 at all, it was way too early. Didn't realize it at the time
orignal
I will exclude it and reduce streaming packet siize from 8K to 3K ))