Probably better to post in the github issue rather than replying here.
I support opening up vote logs to moderators in their own communities. Voting records add useful context to the nature of the exchanges happening, eg. if two people are having a back and forth, but neither is downvoting the other, it contextualizes the disagreement as less hostile.
I don’t think it’s a good idea to give every new user the burden of using that information responsibly. A minority would use it to retaliate, stalk, and harass, and there would be too many of them to reasonably hold them accountable.
I’m split, but I lean slightly towards no. On one hand, it could be good for discoverability, and it would help my efforts to make a client-side algorithm
On the other hand, it will make one of Lemmy’s problems worse - engagement. Some people will vote less, and it’s already feeling a little quieter around here as the numbers settled after the Reddit Exodus. I doubt it’ll be a massive change, but a .5% decrease in voting, permanently, could make a difference
Ultimately, you can see it on federated platforms, so shrug
deleted by creator
On a backend yes. But it’s not as easy to get to as just opening your profile and forming opinion on you by looking at your voting history.
deleted by creator
Hard no from me
I don’t want some nutjob with too much time stalking me because I upvoted something about climate change or downvoted some bigoted shit. We all know those fuckos are out there
Voting on Reddit-like platforms is soft moderation by a community, and if you disincentive that, the whole model kinda falls apart IMO
Votes are already public. This isn’t a change






