bridge and bridge authority proposal

Roger Dingledine arma at mit.edu
Thu Nov 29 15:00:25 UTC 2007


On Thu, Nov 29, 2007 at 09:27:42AM -0500, Roger Dingledine wrote:
> But! There is a problem. See Section 1.4 of dir-spec.txt. There is a
[snip]
> We could also make it 3/4 and 1/4, for better fairness. But this still
> raises the issue that over time we can distinguish a Tor client that
> never asks in the last fraction of the interval. Does this mean that
> *all* clients should fetch in only the first fraction of the interval,
> to hide whether they're using this more aggressive schedule?

Just to move the discussion along, here's my preference: we should
make all ordinary clients fetch the consensus in the first 90% of the
available client interval. Then subclients should fetch it in the last
10% of the available client interval.

Thus we don't get into the issue of using no-longer-valid consensus
documents, and we avoid the all-bridge-users-download-at-once issue
that might come up if a bridge happens to attract many users and they
all download at exactly the cutoff.

Then Tor would decide if it's a "normal" client or a "sub" client based
on the UseBridges config option, and/or on some later config option we
add if we need one particularly for this purpose.

The values of "90" and "10" are fuzzy, of course.

--Roger



More information about the tor-dev mailing list