svn:r4810

svn:r4809

port we're testing, so people are more likely to notice that they're
using the wrong ip address.


svn:r4808

as the culprit


svn:r4807

svn:r4806

svn:r4805

svn:r4804

svn:r4803

Make set_options a little smarter: have options_act handle transitions on its own, and only dirty our descriptor when we really want to.


svn:r4802

svn:r4801

svn:r4800

svn:r4799

svn:r4798

svn:r4797

svn:r4796

svn:r4795

svn:r4794

svn:r4793

svn:r4792

svn:r4791

it if it's not.
(this also fixes some potential security problems with people providing
hostnames as their address and then preferentially resolving them and
partitioning users.)


svn:r4790

sections.


svn:r4789

svn:r4788

will affect them.


svn:r4787

This is actually a bit more subtle, because we need to rotate
the dnsworkers and cpuworkers if certain options (logs, orport, etc)
change. Maybe set_options() should detect this and notify the caller
that it should rotate them.

Expect some mysterious crashes.


svn:r4786

out as a server.


svn:r4785

svn:r4784

svn:r4783

make MonthlyAccountingStart truly obsolete now.


svn:r4782

svn:r4781

svn:r4780

and there's a running Tor server at that address which allows exit to
the destination, then extend the circuit to that exit first.
Also, if the user asks for a .exit node, cannibalize general circs for it.


svn:r4779

svn:r4778

fooled by a low bandwidthrate


svn:r4777

svn:r4776

nick, is this an ugly hack of your protocol?


svn:r4775

Add a 'getinfo unregistered-servers' for arma's internal use.


svn:r4774

try this band-aid to see if it gets better.


svn:r4773

hidden services faster on the service end.


svn:r4772

svn:r4771