orignal
u3fm with new padding
RN
orignal, would the public key as padding be particularly compressible?
orignal
If we use public key itself for padding it should be more compressive
zzz
sure, you could just repeat the key for the padding
orignal
what do you do for it?
orignal
also I send compressed SYN packets now
RN
:D
zzz
orignal, in prop. 161 and my MR, I put repeated 32 byte random padding in there
orignal
yes, but it's not clear
orignal
if you generate new 32 bytes or use public key
orignal
I want my RIs look the same as yours
zzz
I don't do anything right now, nothing's checked in, won't work on it until December
orignal
I do it already
dr|z3d
there's a not great sort toggler at the top of the forum, defaulting to hot iirc.
zzz
oh, ok, time to offer your design services
dr|z3d
haha, funny.
dr|z3d
I might as well paint a target on my forehead if I did that.
zzz
maybe he would pay you in assassination credits or something :)
dr|z3d
what have you been smoking? :)
dr|z3d
whatever it is, you've had enough already.
zzz
looool
dr|z3d
zzz: in router/java/src/net/i2p/router/transport/ntcp/NTCPConnection.java there's a hard limit for message size that limits bids to 16K with a comment suggesting it can go up to 32K. Is it time to have another look at that?
dr|z3d
@@ -153,7 +153,8 @@ public class NTCPConnection implements Closeable {
dr|z3d
* Needs to be fixed. But SSU can handle it?
dr|z3d
* In the meantime, don't let the transport bid on big messages.
dr|z3d
I2P+ has been running with that set to 32K for a while now. No obvious issues observed.
dr|z3d
If possible issues with SSU have kept it from being set to 32K, perhaps it's a good stress tester for SSU2?