This is a starting point, culled from discussions elsewhere.
- Be generally permissive. Users should be able to do basic operations from day one.
- Ramp up on limits – start small and increase based on a successful history.
- Don’t remove privileges once granted (except by moderator intervention).
MVP proposals (with some hedging):
A user can ask N+1 questions per day, where N is the number of questions that user has already asked that have non-negative score and are not deleted.
- Implication: a new user can always ask a question (once per day).
- This does not impose a penalty for downvoted questions – just that those don’t count for the positive record.
- We might need to do something for bootstrapping; when we open for business we should expect people to come in with more than one question ready.
Upon reaching 5 questions per day, the user earns the “unlimited questions” privilege and no longer has a personal rate limit. (The system might have general rate limits.)
- Reasoning: this is meant to be an easy ramp-up and not something that is forever dynamically computed.
(Do we need a metric for answers, or do we allow any number?)
A user can initially post (picking a number here) 10 comments per day.
When a user has posted at least (picking a number) 20 comments and has had no more than 20% deleted by other users (e.g. through flags), the daily limit is removed.
- Again, ramp-up and then remove the dynamic limit.
Third-party edits initially must be approved. When a user has had (picking a number) 20 approvals with no more than 20% rejections, the user gains the “edit without review” privilege. (I am assuming that edits will bump.)
A user imported from SE, where the SE user qualifies for the association bonus, starts with no personal limits on posts, comments, flags, or votes. (Based on this discussion.)
We’ve talked broadly about question closure but not mechanics. I think for MVP it’s sufficient for moderators to be able to do so. Once the community can, there’ll be a privilege path there.
I have not included any limits on flagging.