Hey thanks! Sorry I haven't responded to this. I was gone on vacation. Thank you so much for tackling it! I really apprecate the help. I'll try to look it over this weekend or first thing next week.

One question: will I need to note this as a breaking change in the changelog? I'm assuming that individual actors who were enabled before, may or may not be following this change. I understand that the same percentage will be enabled, but would upgrading flipper with any % of actors enablements were not 100% cause the set of enabled actors to differ? Just want to make sure I note it correctly for people.

I'll probably test it with 100 or 1k actors or something and get the enabled ones before this change and after and then do a set diff. If you have time to tackle that and drop the work here in a comment, it would help getting this merged faster, but if not I'm happy to look at it as soon as I have time.

Hey @jnunemaker, so, it was my turn to be on vacation, sorry for the delayed answer.

As you mentioned, the percentage will probably be similar, but with a really small decimal difference. In few cases, it will be higher & other cases lower than the old implementation %, so it makes sense to mention on the changelog this possibility.

I can do the test batch, just give me some time, so I work on it and drop the result here. Not sure about this weekend, but I do my best to get it done asap. :-)

@rodrigopinto sounds great. If I get a chance I'll whip something together first, but if not I'd love the help. Thanks!

It has a really long time since I opened this PR and it got stale, so I will close it but keep the fork implementation in case I come back to work with this lib again.