[tor-dev] Draft Proposal: Random Number Generation During Tor Voting

[George Kadianakis]

teor <teor2345 at gmail.com> writes:

> Another implementation note on directory caching of the SR doc:
>
> I just noticed the following code in update_consensus_networkstatus_downloads():
>
> for (i=0; i < N_CONSENSUS_FLAVORS; ++i) {
>     /* XXXX need some way to download unknown flavors if we are caching. */
>
> This means that any new consensus flavour will only be cached by new versions of Tor, and they will automatically cache it.
>
> I don't think this is an issue for the SR doc - did we really want / need it cached at directory mirrors? (If not, we could modify this code to skip the SR doc.)
>

Hm, the way we've been thinking about this, the SR doc is only useful to dirauths. 

I don't think that directory mirrors or clients will ever need to download
it. So I think we will indeed need to mod that code to skip the SR doc if we are
not dirauths.

Of course, SR docs should be puclicly available so that CollecTor and
ConsensusHealth can fetch them and analyze them.


> (If the SR doc is cached at directory mirrors, this will make different relay versions distinguishable, but I don't think we worry too much about that, as relays report their versions.)
>
> Tim
>
> Tim Wilson-Brown (teor)
>
> teor2345 at gmail dot com
> pgp ABFED1AC
> https://gist.github.com/teor2345/d033b8ce0a99adbc89c5
>
> teor at blah dot im
> OTR D5BE4EC2 255D7585 F3874930 DB130265 7C9EBBC7