[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #2355 [Tor Bridge]: change the meaning of UseBridges
#2355: change the meaning of UseBridges
---------------------------------+------------------------------------------
Reporter: anonym | Type: enhancement
Status: needs_review | Priority: minor
Milestone: Tor: unspecified | Component: Tor Bridge
Version: | Keywords:
Parent: | Points:
Actualpoints: |
---------------------------------+------------------------------------------
Comment(by anonym):
In my patch I made it so that Tor will abandon all previous circuits when
the bridge settings are changed. This makes sense since Tor then will
start/stop using bridges quickly. This is done just like it is done for
changed settings for !EntryGuards, Exclude*Nodes, etc, i.e.:
{{{
#!c
circuit_mark_all_unused_circs();
circuit_expire_all_dirty_circs();
}}}
However, using the above, circuits that handle streams are not closed
until their streams are closed, which is especially outstanding with long-
liveed streams like SSH/IRC. At least these left-over circuits will not be
used for any new streams, but they show up in Tor controllers such as
Vidalia and arm, and will most likely confuse users, and I believe it is
important to avoid that.
What is the right thing to do here? I guess it all boils down to what the
user expects. Does the user expect long-lived connections to continue
using old circuits that the user has instructed should not be used any
more (and will the general user understand the concepts relating to this)?
I personally do not expect that, and would like a more ruthless circuit
killer. Maybe something like this (in circlist.c):
{{{
#!c
/* Closes all circuits. */
void
circuit_kill_all_circs(void)
{
 circuit_t *circ;
Âfor (circ=global_circuitlist; circ; circ = circ->next) {
ÂÂ if (CIRCUIT_IS_ORIGIN(circ) && !circ->marked_for_close)
ÂÂÂÂ circuit_mark_for_close(circ, END_CIRC_REASON_FINISHED);
Â}
}
}}}
The question is what general users expect, though. Also, since this new
behaviour would kill all streams, any long-lived sessions would be lost.
Are users expecting that? Given the context, i.e. changing
bridge/EntryGuard/Exclude*Nodes/etc, perhaps it is expectable that Tor
will do drastic things?
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/2355#comment:13>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs