Hi all, For a very long time, ssh agent forwarding has been the #1 feature request for Mosh.
For a slightly less long time, Timo J. Rinne's ssh agent forwarding code has been available on GitHub, the current merge is at <https://github.com/rinne/mosh/compare/agent-forwarding-merge-20151128>. We're about to release mosh 1.2.6. Agent forwarding will *not* be in 1.2.6. But after 1.2.6 comes 1.3, and then I think I would like to pull that code into Mosh for that release. So I'm writing this email to restart the discussion on agent forwarding. Obviously, many people want it. But less obviously, it is a significant change to Mosh's security story, and the user is trusting us for their SSH authentication if they use this feature. Also, though I think the code is generally good quality, I have some specific concerns-- it seems possible that Mosh could block with a particularly uncooperative agent or agent client, leaving the user with a stuck terminal session. How do people feel about doing agent forwarding in general, and this code in particular? I'd like to see some consensus on the security issues, and I'd like to hear other opinions on the code involved. regards, --jh _______________________________________________ mosh-devel mailing list mosh-devel@mit.edu http://mailman.mit.edu/mailman/listinfo/mosh-devel
