[tor-bugs] #33752 [Applications/Tor Browser]: Electrum-NMC tries to connect before Tor has connected

Tor Bug Tracker & Wiki blackhole at torproject.org
Sat Apr 4 04:25:18 UTC 2020


#33752: Electrum-NMC tries to connect before Tor has connected
--------------------------------------+--------------------------
 Reporter:  JeremyRand                |          Owner:  tbb-team
     Type:  defect                    |         Status:  new
 Priority:  Medium                    |      Milestone:
Component:  Applications/Tor Browser  |        Version:
 Severity:  Normal                    |     Resolution:
 Keywords:  namecoin                  |  Actual Points:
Parent ID:                            |         Points:
 Reviewer:                            |        Sponsor:
--------------------------------------+--------------------------

Comment (by JeremyRand):

 > I'm going to put this bug on hold

 Lies!  :)  I actually did spend a bit more time debugging this.  The truth
 (as I currently understand it) actually is somewhere between my original
 statement and my follow-up.

 So, it's true that `name_show` is affected while other RPC commands are
 not.  However, this isn't due to any of the `name_show` inefficiencies,
 it's because `name_show` uses stream isolation in a slightly different way
 than the other RPC methods do.  The actual bug is that, generally
 speaking, the first server connection succeeds, while all of the others
 get stalled; `name_show`'s stream isolation code relies on a server
 connection other than the first being active, while the other RPC commands
 work fine if the first connection is active.  'Twas fun to debug that.

 Anyway, I was able to prototype a patch for Electrum-NMC that keeps the
 network paused on initial launch until the Tor daemon has connected.  This
 decreases the startup delay between double-clicking the Tor Browser icon
 and Namecoin resolution being available from ~55 seconds to ~6 seconds.
 The prototype patch is not really in a state where it could be merged,
 it's solely to test that the approach works.  I'll send in a proper patch
 sometime soon.

 In terms of upstream Electrum, Ghost really wants to know why it's hanging
 to begin with, rather than just going with a workaround patch without
 understanding the causes.  So I'll probably spend some time tracing what's
 going on there, in the hopes that either we can fix the underlying
 breakage, or get a patch merged to upstream Electrum.  Not sure when I'll
 get to this though.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/33752#comment:3>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online


More information about the tor-bugs mailing list