[tor-reports] Some interesting things that Nick did in May

[Nick Mathewson]

Hi, all!

In May we finally merged my Ed25519 patches (with reviews from Andrea,
David, and George Kadianakis).  This currently authenticates router
descriptors using ed25519 keys.  Coming up, it will need to
authenticate connections and circuit extension as well.  It also
represents our first serious use of Trunnel-generated code to handle
wire inputs.  (Let's hope there aren't any showstopper bugs in
Trunnel; please review the generated code if you have the time?)

I've solicited review for my patch on offline/encrypted master keys
(see #13642) ; it should make relays a little more robust to
compromise.  (Though you still should make sure that you don't get
your relay compromised if you can possibly avoid it.)

Stuff we've been designing:
    * We've done work on figuring out what our guard algorithm should
be; we wrote up some criteria and use-cases and tried to figure out
the correct behavior in each.
    * I started on a writeup of how to finally remove deprecated tor
versions from the network at
https://pad.riseup.net/p/deprecating-old-tors2 . Needs more work.
    * Wrote proposal for using rfc5705 instead of raw access to
master_key/client_random/server_random.

With help from yawning, I tracked down a bunch of compatibility issues
with openssl 1.1, submitted patches to openssl 1.1, and then wrote
patches from Tor to incorporate the necessary changes.

I've spent lots of May in the office, acting as the interim deputy
executive director.  (Or is it deputy interim ED? The point is, there
have been a lot of jobs to do around here, lots of accounts to make
sure I've got access to, lots of sites to make sure I can log into,
and so forth.)

Also, it's been the typical bug fixing and misc haxxing too.

cheers,
-- 
Nick