1

Per my reading of the timeline, this question has got 3 answers in 24 hours from new users (i.e. users who earned less than 10 reputation on the site, excluding any association bonuses): https://softwareengineering.stackexchange.com/posts/216371/timeline

  • May 10 '14 16:28, 16:50 and May 11 '14 08:16.

As far as I understand, these answers should have triggered auto-protection.

How come that this did not happen?

2

What happened is that two of the people answering got fairly quick upvotes on their answers. Enough to go over the 10 rep requirements.

At that point, when a new answer comes in, those are excluded.

This is working as designed - may need tweaking though.

8
  • consider editing the feature description to account for this (or better yet, tweaking it to ignore quick upvotes) – gnat May 12 '14 at 17:13
  • for the record, another example here: single "welcome" upvote to an meh answer (possibly from question asker) trashes auto-protection – gnat May 23 '14 at 4:00
  • Oded, could you please take a look at why protection didn't kick in when this answer was posted (at TWP)? Answer isn't from 10-rep user, but I expected check to be triggered and discover 3 "eligible" posts: this, this and this – gnat May 29 '14 at 23:57
  • 1
    @gnat - two of those posts (the answer that was posted and one of the now deleted answers) were by people who had earned rep on the site in the past. Those posts were not eligible. – Oded May 30 '14 at 7:18
  • I see, thanks! Did I miss something in the feature description then? it says about new users but reads as if these are "users who earned less than 10" (my understanding of "earned" is in the sense of balance of gained and lost) – gnat May 30 '14 at 7:24
  • ...it would probably be more consistent for auto-protection to kick in after 3 posts from users who aren't eligible to post into protected questions – gnat May 30 '14 at 7:36
  • Oded, I just asked about some stats related to this: How many questions escape community protection because of reputation gained within checked question? Would appreciate if you take a look – gnat Aug 25 '14 at 20:46
  • Oded, sorry for bothering, would you mind taking a look at above ^^^? The more I observe the way how it works, the more it feels weird. Today's example at TWP seems to be particularly troublesome; details of what I am particularly concerned about are in my flag on that TWP question (it looks especially impressive now in the "active" view - with this, this and this "answers" on top) – gnat Oct 31 '14 at 11:36
1

I think it may have to do with the protection/unprotection by a diamond the first day; see the revision history. This may have blocked the auto protection from kicking in.

1
  • this is important observation. Wonder if past unprotect is indeed intended to block protection in cases like this, and if yes then why is that. Per my reading of auto-protection feature "specification", this doesn't look like an intended behavior – gnat May 12 '14 at 14:14

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .