~dr|z3d
@RN
@RN_
@StormyCloud
@T3s|4
@eyedeekay
@orignal
@postman
@zzz
%Liorar
+FreefallHeavens
+Over
+Xeha
+acetone
+bak83
+cumlord
+hk
+poriori
+profetikla
+uop23ip
+weko
An0nm0n
Arch
Danny
DeltaOreo
Irc2PGuest21357
Irc2PGuest21881
Irc2PGuest5995
Irc2PGuest94373
Meow
Nausicaa
Onn4l7h
Onn4|7h
T3s|4_
anon2
anu
boonst
enoxa
mareki2pb
not_bob_afk
plap
shiver_
simprelay
solidx66
u5657
dr|z3d
ok, they've arrived, by a different route than I intended, but hey. svg identicons in your addressbook, available on the latest + dev build.
dr|z3d
banging my head against the wall trying to get java to handle the rendering is why I was a bit testy earlier, T3s|4. Sorry.
dr|z3d
Max 4 digits for in/out bandwidth in snark until we identify the root cause of the issue.
T3s|4_
sounds like a plan dr|z3d :)
zzz
dr|z3d, unable to reproduce your access filter issue, see gitlab for details
dr|z3d
10-4, zzz
dr|z3d
The issue was caused when performing an echo > denyfile.txt on a non-empty file while the tunnel was running.
dr|z3d
Not an unreasonable thing to do, or expect, if you look at a deny file and find several thousand entries.
dr|z3d
I like the idea of skipping bad lines that you suggest, either skipping them or removing them from the file. Setting an upper limit on the number of entries might also be worth considering.
zzz
that would create a one byte file with a newline, not an empty file. which was it?
dr|z3d
yeah, probably that.
dr|z3d
apologies for the not-100% accurate report.
zzz
np, so I have reproduced it, but this was completely self-inflicted ((
zzz
> denyfile.txt would have been fine
dr|z3d
I wasn't aware that echo > inserted a newline, my bad.
zzz
echo -n > denyfile.txt would have worked too...
zzz
anyway, I know enough to work on it now
dr|z3d
it's a powerful tool that's probably underutilized and could do with a buff up.
uop23ip
Hi, can i change my "Minimum threat points to block:" in a way to include the peers i blocked because they are slow. I got so many of them. Canon here.
zzz
it's a botnet, there's thousands and thousands of them. We don't block them permanently because you'd burn a ton of memory doing it
zzz
no you can't do it that way either
uop23ip
wow, than better ban them i guess. I got 60+ percent of known in numbers banned, that seemed too much for me, but with your explanation it make sense to block them. Is this a "friendly" botnet (meaning they are doing their thing and don't hurt the network) and the last attack is gone? New attack?
zzz
the sybil stuff is pretty much disabled because the netdb spam broke it
uop23ip
is it needed for the next release? Should i let it activated or does it hurt(cpu wise or other)?
zzz
there's no sybil changes so far in this cycle
zzz
no such thing as a friendly botnet; ofc we don't know the details, but they're generally poorly configured and harmful, even if not doing explicit attacks
zzz
working now on the one that appeared a couple weeks ago and is knocking several points off our expl. build success
uop23ip
ok thanks. Wish you much success. Didn't realize that there always something going on. I thought after build success had gone up and the chinas were gone, everything is normal again and my router was just a little bit picky :)
zzz
right. the 'slow' botnet still exists; not sure about the 2.5.1 botnet, it may be gone; and there's the new one I'm working on. And ofc china may come back
uop23ip
Is there a danger that a massive banned bot net can fill my network view (known) to that level, that my "real" known and used peers getting that low that it would affect my anonymity?
zzz
that's the kind of things we're thinking about and trying to mitigate, yes
uop23ip
can i make the calculation known-banned<100-> plug out it is getting dangerous unanonymous :)
zzz
??
uop23ip
i looked at my known numbers and my banned numbers and would do (known-banned). Completely wrong here?
zzz
oh
zzz
banned usually not included in known, so known-banned is meaningless
uop23ip
ah ok good to know. But i could say that there is a ratio to the of what percentage of my network view are bad apples. in my case 2k known and 1k banned. 1/3 of my network view are bad apples, right?
uop23ip
known 100 and banned 3k would be a reason to plug out the service, because 100 is to low to have anonymity. Is there some number were would say that or does have known nothing to do with anonymity??
zzz
we try hard to maintain a lot more than 100
uop23ip
that good that i don't have to expect that case, because i would plugout at 100/3k. that ratio doesn't feel right to me. Thanks for your time zzz and happy hunting!
T3s|4
o/ dr|z3d - your head banging on the *.svg addressbook icons was worth it - looking very snazzy on 19+. Noted on the hard coded snark down/up entry limits of four nines. Without surprise, given my recent hair pulling, new added torrents proceed as expected :D
dr|z3d
thanks T3s|4, happy you're enjoying the new identicons. I was originally trying to modify the existing java identicon library, but after much headbanging and the realization that the new identicons look better, I went with those instead (javascript).
dr|z3d
phase 2 is to cache all the svgs so they only need to be rendered once.
T3s|4
yeah, they really 'pop' on my 4K display :)
dr|z3d
yeah, good. I suspect the old bitmap icons were looking a little blurry @ 4K (resolution, not b/w) :)
T3s|4
dr|z3d: no critique intended toward you, but WTF is up with SVG? w3.org/Graphics/SVG - Wasn't 2016 8 years ago?
dr|z3d
it's a mature format.
T3s|4
dr|z3d: sure, but project maturity is not an excuse for 8 years of neglect
dr|z3d
well, it's not so much a project as a defacto standard, and you don't really want to be messing with those too much once they're implemented everywhere.
dr|z3d
it's a bit like png. set in stone. you want new things, you define a new file format, like webp.
T3s|4
would have been far better, even 8 years ago, to express thanks to all those who contributed to standard development and testing...then sign off
dr|z3d
I think you might be right about the site, if nothing else. It's not exactly current :)
T3s|4
nah, I prefer to roll with, the svg site is current, assuming you mean by 'current'; it's currently an embarrassment to all the folks who donated many hours of their lives agreeing on the parameters of the standard. They should be explicitly recognized and thanked :)
dr|z3d
:)