IRCaBot 2.1.0
GPLv3 © acetone, 2021-2022
#ls2
/2023/02/20
orignal might be later today
zzz 0) HI
zzz lets keep it short today, and orig. said he might be late
zzz anything for the agenda today?
zzz I don't have anything
orignal no I'm here ))
eyedeekay Nothing from me, been busy with other project-related stuff
zzz anything for today orignal ?
orignal I think we have a lot to discuss
zzz ok then, give us the list, but let's keep it brief
orignal have you thought about SSU2 I2NP messages without session?
orignal no problem we can discuss later
zzz yes I like the idea a lot, but really a topic for a few weeks/months from now
zzz what else?
orignal I though you might have something to say
orignal RouterInfo encryption
orignal please enlight the scope
orignal I need to understand for my design changes
zzz for SSU2 without session?
orignal for what you do now
zzz oh I understand
orignal because it's something new for me that you do it
orignal and it changes behaiour dramatically
zzz if we send it through tunnel we do the same as with LS
zzz but my stats from the other day indicate it's not really a connection problem i2pd->ff
orignal but there are much more RI updates then LS
orignal and according to you you usually send though tunnels
zzz sure, but my theory that sending through tunnels would help may be wrong
zzz so I'm not asking you to do that at this time
orignal let me example
orignal *explain
zzz go ahead
orignal when I receive I2NP garlic message not from tunnel I assume they come to router
zzz right
orignal but I decrypt them in the same thread as transport I receive from
orignal my assumption was that it doesn't happen often
orignal but it seems I was wrong
zzz maybe... take a look at one of your ffs and get some stats
orignal and ordirnary FF should invoke sich decreption often
zzz do some research before doing design changes
orignal it means I need to move garlic decryption to a spearate thread
zzz take a look, might be a problem, might not
orignal and I will also do the same, e,g. encryopt RI when send through tunnels
orignal we must think abot future problems, not what we have now
zzz the number of encrypted RIs may be about the same as number of LSes. More routers but less often
zzz sure, just want to make clear I'm not asking you for design changes, my theory from the other day was wrong
orignal no just need to understand the scope
zzz got it :)
orignal another question is
orignal people are asking
orignal do you have any plan for DHT?
zzz "real dht" -> not at this time
zzz my stats show that we use very little CPU on ff peer selection now
zzz so just a future todo
orignal strange since we see a lot
zzz not sure why if we are o(n log n) and you're just O(n)
zzz what did I say? 650 uSec 12x/second on a slow ff
zzz basically nothing
zzz more research for you :)
orignal we will see
zzz anything else for today?
zzz ok then, thanks everybody
orignal 2.46.1 fixes some bugs