Current olusanya HEAD gives different results than v4.1.3

[johpiip]

link_lengths.txt is due to #494 and it is fine, but others I can not identify.

[zptro]

This is my best guess: 121ecc0#diff-a31f8dd24148efc83d5e06974a77a7a7d4faa374dd84d0884fb6c5c377639bd2

This changed speeds in the test network, so that should affect exactly those results if you used the test network.

[hsl-petrhaj]

This sounds like a good explanation, does it work? @johpiip

[johpiip]

Thank you both, I will check it later.

[hsl-petrhaj]

@zptro @johpiip
I have been inverstigating the differences in the results.
Two changes occurred, both around April. The second one does not even show on the test network, only on the full network. The differences are very small, for example the volume could be like 5000 (+2), or 221 (-1).
My theory is that the changes between float32 and float64 might have in theory caused that, but it does not prevent the new version 4.1.4 from being released.