3 ms·
(HN meta, unrelated to the article.) What's the rough timeline for the user-assisted unflagging plan?
by JoshTriplett 12y ago
(HN meta, unrelated to the article.) What's the rough timeline for the user-assisted unflagging plan?
- dang 12y agoSoon! We're working on it. There turn out to be some technical subtleties that we want to get right before deploying.
- JoshTriplett 12y agoAwesome. Any plans to write about those technical subtleties at some point?
- dang 12y agoNot really, but only because any time I have for writing is better spent writing code—i.e. better for the community, plus I like writing software more than writing about writing software. To assuage your curiosity, though, the subtleties have to do with trying to make sure the design is consistent with itself and with other parts of HN. There is more than one way to generalize flagging, and some of the ways we tried led either to unwanted behaviors or unwanted complexity. For example: if you vouch (that's the word we're using for anti-flagging, i.e. you vouch for a post if you think it shouldn't have been flagged) for an item and then un-vouch for it, it should go back to its previous state. Posts that were killed by software vs. by user flags have slightly different properties, and the vouch feature needs to preserve those. The way that flagging affects story rank is different from how it affects the life cycle (dead-or-aliveness) of a story, and vouching needs to be consistent with both. When we release this, we'll have a thread about it and can discuss it more if people want to.