@eyedeekay
&eche|on
&zzz
+R4SAS
+RN
+T3s|4
+acetone
+dr|z3d
+hottuna
+orignal
+postman
+weko
An0nm0n
Arch
FreefallHeavens
Gid
Hidenet
Irc2PGuest19856
Irc2PGuest26747
Irc2PGuest2827392
Irc2PGuest39199
Leopold
Minogami
Onn4l7h
Onn4|7h
Palimpsest
ProRu
R4SAS_
Sleepy
SoniEx2
Teeed
aargh3
admin
anon2
b3t4f4c3__
cheddah
eyedeekay_bnc
itsjustme
j6
limak
profetikla
qend-irc2p
rapidash
tbqdrn
theglitch
uis
w8rabbit
x74a6
zer0bitz
itsjustme
lol RN
dr|z3d
missing a " zzz.
dr|z3d
FamilyKeyCrypto.java line 540.
zzz
?
dr|z3d
if (familyName.contains("/") || familyName.contains("\\") || > if (familyName.contains("/") || familyName.contains("\\"") ||
dr|z3d
I assume that's what you meant, escaping the "
dr|z3d
or maybe not.
dr|z3d
in any event, that line plays havoc with the syntax parser in my editor.
zzz
the code you're referring to was removed in the checkin
zzz
you're looking at the old version?
zzz
it was in there from 2015 until now
dr|z3d
oh, yeah, you're probably right. in the middle of a conflict merge / cherry-pick.
dr|z3d
yeah, sorry, my bad. note to self: don't fix reported errors on compile until after the cherry-pick has completed.
zzz
np. I got interrupted about 10 times yesterday and was left with this huge pile of unfinished stuff. I had to finish it all this morning and push it so I could think straight.
dr|z3d
so desk is cleared of backlog now then? good.
zzz
down to 38 modified files locally, which is manageable
zzz
dr|z3d, you get yourself unclustered?
dr|z3d
oh, you mean merge-wise? yeah, all good thanks, zzz.
zzz
ok, holler if you see any problems, that was kindof a lot
zzz
trying to get stormycloud out of the doghouse
zzz
will take a week due to sybil persistence
dr|z3d
thanks, will do. my fault, I shouldn't have trusted a compile in the middle of a cherry-pick. otherwise, no issue :)
dr|z3d
oh, and yeah, he surely is in the doghouse right now :)
dr|z3d
I've got visibility on about 40 stormycloud routers, all blocked right now.
zzz
spent two days working with them to straighten it all out
zzz
I believe you can get a clean slate if you stop, delete blocklist-sybil.txt, and restart
dr|z3d
would be handy to be able to add a familyname to the manual peer controls input and unban that way?
dr|z3d
perhaps with a prompt before you do to indicate all known routers in said family.
dr|z3d
(or just a count)
zzz
we don't remember the mapping of family to banned ips and hashes
dr|z3d
ok, so maybe plan b is button on the sybils overview page to "Clear existing bans" ?
zzz
even thats not easy because all the various ban sources get merged at startup
dr|z3d
ok, maybe not that critical, then. as you say, will self-fix after a while, and most families don't have 40+ members. notwithstanding, maybe a whitelist for hashes/families might be something to consider.
zzz
that's what you just cherry picked :)
dr|z3d
ah, whitelist. ok, very good, as you were :)
dr|z3d
never looked to hard, but I guess the whitelist is automatically generated from families with locally stored keys?
zzz
it just doesn't clear out the previous blocklist entries, for the reasons stated above
zzz
the whitelist (cough allow list) is hardcoded
dr|z3d
ah, I see you. :)
dr|z3d
_familyExemptPoints24.add("SUNYSB");
dr|z3d
_familyExemptPoints24.add("stormycloud");
zzz
yup
zzz
otherwise stormy would have 200 points
dr|z3d
fluffy toy territory, or a holiday in Seychelles.
zzz
wooooooooooo big news, first handshake messages exchanged between i2pd and java i2p
dr|z3d
\o/
dr|z3d
changing topic entirely, I was raising lockless ring buffers in the context of your post on the forum re context switching etc. dunno if you've looked into those.
zzz
it's madness, and I won't respect you anymore if you keep pushing them :)))
dr|z3d
lol
dr|z3d
this is a full stacktrace of the zzzot encoding bug: cake.i2p/view/g5QPzrs67T_cbc7bGyDtQwXZcRRJKyQqvB6fomo3w_IIZXbNuJKl/g5QPzrs67T.txt
zzz
jetty version?
dr|z3d
current
dr|z3d
9.3.30.v20211001
dr|z3d
stacktrace was from wrapper logs, the only thing that gets sent to the router logs is: WARN […t Jetty-382] …ty.server.Server: /: org.eclipse.jetty.http.BadMessageException: 400: Bad query encoding