weixin_39854778
2021-01-06 10:38 阅读 1

Voting threshold

There should be a voting threshold (it's actually a threshold on the votes difference).

Set to 1% of repo followers (currently, that means the difference between 👍 and 👎 should be at least 12).

Prevents things like #138 and #48.

Maybe we can change this later to use a percentage of active contributors instead of (possibly passive) stargazers? Not sure how to calculate that atm though.

该提问来源于开源项目:Chaosthebot/Chaos

  • 点赞
  • 写回答
  • 关注问题
  • 收藏
  • 复制链接分享

11条回答 默认 最新

  • weixin_39843782 weixin_39843782 2021-01-06 10:38

    I'd maybe do it off a percentage of active participants. Would prevent any deadlocks from occurring

    点赞 评论 复制链接分享
  • weixin_39854778 weixin_39854778 2021-01-06 10:38

    I'm all for it. Any ideas how to retrieve that?

    点赞 评论 复制链接分享
  • weixin_39980002 weixin_39980002 2021-01-06 10:38

    With #140 we could check against the number of people who have X or more points.

    点赞 评论 复制链接分享
  • weixin_39854778 weixin_39854778 2021-01-06 10:38

    Oh shit I voted against that 😆 Maybe we can retrieve a stat on activity on forks in the past x days and use that? Although I think that's quite expensive..

    点赞 评论 复制链接分享
  • weixin_39843782 weixin_39843782 2021-01-06 10:38

    sure that's one solution. It would definitely pick up on anybody who has voted for PRs. In the long run even that could get stale if somebody hasn't voted in a long time. Maybe storing some kind of "last active" per user could be useful

    点赞 评论 复制链接分享
  • weixin_39980002 weixin_39980002 2021-01-06 10:38

    I was planning to suggest using a certain amount of point decay over time (for each PR) to avoid that.

    点赞 评论 复制链接分享
  • weixin_39843782 weixin_39843782 2021-01-06 10:38

    Sounds good!

    点赞 评论 复制链接分享
  • weixin_39843782 weixin_39843782 2021-01-06 10:38

    I would just like to point out this PR would not pass if this was currently implemented. Perhaps the threshold should be adjusted more?

    点赞 评论 复制链接分享
  • weixin_39854778 weixin_39854778 2021-01-06 10:38

    Good point, yeah I think so. Although I'm leaning towards a conservative threshold, of course that conflicts with creating chaos. Hmmm...

    Maybe 0.5% (=7 votes difference atm) is better?

    点赞 评论 复制链接分享
  • weixin_40001275 weixin_40001275 2021-01-06 10:38

    As already mentioned, time decay is a better solution.

    点赞 评论 复制链接分享
  • weixin_39782500 weixin_39782500 2021-01-06 10:38

    :ok_woman: PR passed with a vote of 18 for and 2 against, with a weighted total of 16.0 and a threshold of 1.0.

    See merge-commit efb724b6ead94b4379ade6ab27a55f04a49936a7 for more details.

    点赞 评论 复制链接分享

相关推荐