<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body dir="auto"><div dir="ltr"><span></span></div><div dir="ltr"><div dir="ltr"><div>Hi,</div><div><br></div></div><div dir="ltr">On 4 Jun 2019, at 08:11, Mike Perry <<a href="mailto:mikeperry@torproject.org">mikeperry@torproject.org</a>> wrote:<br><br></div><blockquote type="cite"><div dir="ltr"><span>But this is a good point for long term: LTS is really going to make it</span><br><span>impossible to substantially improve the Tor network with any relay-side</span><br><span>new development improvements prior to <span>Feb 1, 2022</span> (when the 0.3.5 LTS</span><br><span>expires).</span><br><span></span><br><span>Most funders are unlikely to be impressed with our inability to provide</span><br><span>big-win results sooner than 3 years from now, especially wrt our</span><br><span>worst-case performance/variance problem. That's probably longer than it</span><br><span>would take just to build a new Tor-like network, instead... We need to</span><br><span>find a better way to support Debian/stable relay operators, other than</span><br><span>simply letting them hold back the network and increase development time</span><br><span>and maintenance costs by running ancient software.</span></div></blockquote><br><div>Most packagers upgrade to the stable Tor version within a month or two of</div><div>a release:</div><div><a href="https://trac.torproject.org/projects/tor/wiki/doc/packages">https://trac.torproject.org/projects/tor/wiki/doc/packages</a></div><div><br></div><div>For Debian, we also run our own <a href="http://deb.torproject.org">deb.torproject.org</a> with the latest stable.</div><div><br></div><div>I'm not sure if there is much more we can do on the packaging side of things.</div><div>We could look at how Rust and Firefox do it: some kind of auto-update</div><div>mechanism.</div><div><br></div><div>But we still need an incentive for operators to configure updates correctly.</div><div><br></div><div>Here's are some potential mechanisms:</div><div><br></div><div>0. We encourage relay operators to upgrade by contacting them.</div><div><br></div><div>1. Older relays log a message that encourages operators to upgrade.</div><div><br></div><div><div><span style="background-color: rgba(255, 255, 255, 0);">2. Relays on old Tor versions are given a reduced consensus weight.</span></div><div><span style="background-color: rgba(255, 255, 255, 0);"> We should set the penalty so most clients use newer relays, but not</span></div><div><span style="background-color: rgba(255, 255, 255, 0);"> so high that newer relays are overloaded.</span></div></div><div><br></div><div>3. Very old relays are removed from the consensus.</div><div><br></div><div><div><span style="background-color: rgba(255, 255, 255, 0);">(I can't imagine a world where we successfully backport performance</span></div><div><span style="background-color: rgba(255, 255, 255, 0);">improvements to very old tor versions.)</span></div></div><div><br></div><div>T</div></div></body></html>