On Mon, Dec 3, 2012 at 12:20 PM, David Fifield david@bamsoftware.com wrote:
I noticed a change in behavior in cb62a0b69a7d67b427224ca4c3075b49853a3a1f or thereabouts. tor opens a new SOCKS connection to a client transport plugin while bootstrapping at 50% (if descriptors are not cached) or at 85% (if descriptors are not cached). The upshot is that the flash proxy transport, for which new connections are not free, needs two connected browser proxies in order to bootstrap. Earlier revisions permit bootstrapping and circuit creation with just one connection to the proxy.
Hm. If you haven't figured this out so far, it doesn't seem likely to get solved on tor-dev. Open a ticket on the bugtracker, please?
You could also investigate (using extra log messages in Tor or a debugger or something) whether the reason the first channel isn't getting reused is because of the "is bad for new circs" test as you suspect. If so, that would help confirm the hypothesis, and help come up with a fix.
yrs,