Lower MAX_SAMPLE_THRESHOLD by a lot

the current value of 0.3.0 in my prop271_030_v1 branch is ridiculously large.

changed milestone to %Tor: 0.3.0.x-final

added component::core tor/tor milestone::Tor: 0.3.0.x-final owner::nickm parent::19877 priority::high resolution::implemented severity::normal status::closed type::defect labels

Trac:
Status: new to accepted
Owner: N/A to nickm

In 45889fafbd7d7e747793f6e26074abf24b85352e, I've added a fix for this to my prop271_030_v1 branch.

Trac:
Cc: N/A to asn, chelseakomlo
Status: accepted to needs_review

a5a4640a9544d2 in my torspec prop271-changes branch specifies the change.

I wonder if the fractional restriction is any useful now that we have the absolute restriction. In most cases the absolute restriction will take priority over fractional restriction (20% of total guards is like 110 guards right now).

I feel like the absolute restriction is more accurate and powerful. Maybe we can kill the fractional one to reduce code?

Perhaps it would make sense to introduce the fractional restriction if it was based on guard bandwidth and not just number of guards.

Maybe. I'm thinking that there are network sizes where the fractional restriction matters more -- like on a 30-node network, for instance. At least, that's what I had in mind.

Hmm I can see how that makes sense. However, if we are trying to provide security for such tiny networks (30 nodes) we would probably have to change the rest of our guard picking strategy quite a bit as well.

Anyhow, no strong opinions here and the code seems reasonable.

Trac:
Status: needs_review to closed
Resolution: N/A to implemented

closed

moved to tpo/core/tor#20920 (closed)