Closed okdistribute closed 4 years ago
Looks like the sync: missing data still ends
test is timing out with this change.
I wonder if it's because the side that has sofar===total
can't become stale anymore? Maybe for "complete" peers the stale timeout is just longer instead of disabled altogether? I'm just wondering if these tests are depending on it being able to eventually become stale (and didn't investigate very closely).
@noffle I figured it out -- and/or issue :P
This is failing on windows on travis, but locally on my windows machine it is not failing, so I'm going to merge this in the interest of getting a release candidate out -- but the tests /do/ need to be fixed
This fixes a bug for when 'missing' would be falsely reported