logs archiveIRC Archive / Oftc / #tor / 2015 / August / 19 / 1
kvm234
mrphs: what do you propose, that tor all any and all fonts?
mrphs
kvm234: to at least have all the system default fonts, included.
kvm234: and to make this a high-security setting, only
(why dragging this from tor-dev to here though? :)
kvm234
I don't think I can speak on tor-dev unless that was recently changed
it looks like I can heh
cacahuatl
isn't #tor-project specifically for commentary on #tor-dev stuff, according to it's topic at least?
kvm234
I thought so
tyil
when running tor as proxy in my browser, I get "General SOCKSv5 server failure" when trying to load and 192.168.*.* address
can someone point me as how to disable tor for private ip ranges?
cacahuatl
It does already?
velopian
didnt realise this channel was +M sorry
cacahuatl
do you mean as an exit policy for a relay or for a client making tor connections?
         

velopian
tyil in your browser or in the tbb?
also hiii tyil
tyil
velopian: in my browser, which is qutebrowser, if that means anything
I set it to use a SOCKS proxy on 127.0.0.1:9050 (which is tor)
velopian
it means you may have to find a way to except certain ip ranges in your proxy settings if thats even possible
cacahuatl
Tor as a client won't make connections to private IP ranges
velopian
tor doesnt manage it itself
if you access localhost through tor unadulterated, and you SOMEHOW get a misconfigured exit, you can access them through localhost
tyil
ah okay
serban
~/ww
tyil
thanks for the info, I'll look into that
maybe ask the browser dev if there's a way to disable the browser's SOCKS proxy for private ip ranges
velopian
that would be helpful since firefox does that too
amorphous
hello
is there anyone home ?
i really needed to talk about non-blocking sockets and Tor ?
tyil
you can talk to me about that, but I have no knowledge on it :')
but I'm a great listener
amorphous
ok
its not too bad - im just trying to open a sock4a connection to tor after setting the iomode to non-blocking....
fails every time
but when its a blocked connection the call to connect() works
but i thought Tor documentation suggests non-blocking sockets work
mrphs
... why socks4?
amorphous
i have a library that does the connection and i cant go socks5 with the library without Tor complaining - but its more than happy with blocking sockets with sock4a
i thouhgt it a good idea to capitolise on the socks4a success by adding non-blocking to that implementation
cacahuatl
Also your socket options only go as far as the local socket :P you're just setting the options of your local comms to the tor process, non-blocking shouldn't be an issue.
it's infact, entirely unrelated to Tor
amorphous
ok
thanks
but i wonder why blocking is good, and non-blocking is bad...
i see code online that simply sets the socket to non-block and then makes the call
cacahuatl
Consult stackoverflow ;)
amorphous
oh ...
ok
will do a search...
thanks for the help
sorry i dissapeared
on stack exchange now...
kernelcorn
are you working in C? I have some C++11 code for asynchronous IO via Boost Asio.
amorphous
wow
sounds awesome
im working in C and C++ here
the code base is a bit mixed..
old style c with some c++ objects....
mmm
any code would be good..
since im sure i can at least take the idea away from me i see in the code
sorry - that was crap english
i mean i can takeaway from the code ideas and cues i think
         

kernelcorn
are you looking for just SOCKS code, or do you want general asynchronous IO?
my SOCKS code: https://github.com/Jesse-V/OnioNS-common/tree/master/src/tcp It's heavily adapted from the Boost Asio example code.
you may be particularly interested in SocksClient::getCircuitTo
it is SOCKS4a, I have not upgraded to SOCKS5
amorphous
i think socks code is fine to be honest
socks4a more specifically...
thanks for the link
checking it out now - and apologies for the delay...
this is awesome code..
thanks very very much
this is hard to reconcile with what i have here in my code..
its quite hard to see how the socket is set up before connecting
it could of course be because i am a luddite also :)
kernelcorn
no problem. I had to look around for a bit to figure it out. Boost's documentation was helpful, I stole their example. :)
amorphous
thank you !!!
kernelcorn
just so you're aware, SocksClient::getCircuitTo does expect certain replies from the other end and SocksClient::sendReceive is blocking
amorphous
ok
thanks for this
can i post code here ?
kernelcorn
please don't if it's more than 4 lines or so
pastebin if it's longer, or just link to it
amorphous
damn its its 6 lines...
thats ok is it ?
kernelcorn
6 lines if you remove the brackets?
amorphous
with all brackets... :)
ok its 5 now..
kernelcorn
I think it'd be okay to paste. Usually we don't like long pastes because it's spam in the chat.
amorphous
i agree with you
int iResult = ioctlsocket(in_Socket, FIONBIO, &iMode);
if (iResult != NO_ERROR) {
printf("ioctlsocket failed with error: %ld\n", iResult);
}
retval = connect(in_Socket, in_Socks, socksLen); //Connecting to Socks Server
thats the socket being set up for non-block and then the call to connect
which fails to -1 i think
if i change to block - by changing the iMode it works fine
kernelcorn
wait, this is Windows development?
« prev 1 2 3 4 5 6 7 next »