<div dir="ltr"><div><div><div><div>Hi George, <br><br></div>We were having a similar discussion about what to include in the "receive new consensus" action.<br><br><span>We currently have three actions to remove dead/unused guards. They are: <br></span><ol><li>Marking guards that are not listed in the latest consensus as "bad"</li><li>Remove guards that have been dead for 30 days<br></li><li>Remove guards that were added more than 30 days ago</li></ol></div><div>Specifically, our question was whether #2 and #3 should be part of the algorithm for new guard selection, as this seems to be more state file maintenance, or if these will always be a part of the "receive new consensus" action.<br><br>If #2 and #3 can be separated, we were wondering where these would go- if there are other similar events for state file maintenance. <br></div><div><br></div>We have an updated document here- <a href="https://github.com/twstrike/tor_guardsim/blob/develop/original_algorithm.md" target="_blank">https://github.com/twstrike/tor_guardsim/blob/develop/original_algorithm.md</a><br><br></div><div>Looking forward to seeing what you think! <br></div></div>Chelsea<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Feb 15, 2016 at 10:15 AM, George Kadianakis <span dir="ltr"><<a href="mailto:desnacked@riseup.net" target="_blank">desnacked@riseup.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">Ola Bini <<a href="mailto:obini@thoughtworks.com">obini@thoughtworks.com</a>> writes:<br>
<br>
> Hi again,<br>
><br>
> Here is the newest version of the algorithm:<br>
> <a href="https://gist.github.com/olabini/343da01de8e01491bf5c" rel="noreferrer" target="_blank">https://gist.github.com/olabini/343da01de8e01491bf5c</a><br>
><br>
<br>
</span>Thanks! Looks better!<br>
<br>
I think we are reaching the point that we need good simulations and actually<br>
running and stepping through the algorithm with actual data to find issues and<br>
suboptimalities. And maybe we also need to get a third person familiar with<br>
guards to review it.<br>
<br>
BTW, I noticed you removed the "we received a new consensus" part. That's fine<br>
for now, but I think it should be added back at some point maybe as a separate<br>
event. I like the work done by rjunior and chelsea here:<br>
    <a href="https://gist.github.com/chelseakomlo/2acbe15314b5a809c6f4" rel="noreferrer" target="_blank">https://gist.github.com/chelseakomlo/2acbe15314b5a809c6f4</a><br>
and I think the events described are quite close to the Tor networking API. So<br>
it might be worth having this algorithm mirror that event structure.<br>
<br>
</blockquote></div><br></div>