4
$\begingroup$

Fixing a typo currently moves a question to the top of the list. Probably it has been discussed on meta.SO, but: if there were a way of doing this a few characters at a time without bumping the post to the top one could incrementally fix all the minor defects without any saltation.

[add: as in the comment, I really mean two thresholds: one a minimum number of characters for bumping based on a single edit, and the other a maximum cumulative "number of edited characters since last bump" count, that triggers a bump when reached. e.g., 4 and 30. There are other numbers that could also be tracked to avoid most exploits.]

$\endgroup$
  • 2
    $\begingroup$ See the recent question here at meta.math.SE: meta.math.stackexchange.com/questions/1385/…. There are several relevant meta.SO threads, including: meta.stackexchange.com/questions/26483/…. Although your request may not be identical to the previous ones, I believe the same answer will apply. $\endgroup$ – Jonas Meyer Dec 23 '10 at 4:34
  • $\begingroup$ Viewing it as a question of thresholds (minimum needed to bump plus a cumulative maximum edit count that forces a bump) overcomes the objections I saw in the meta.SO answers and allows all the current functionality. Those who want to bump could delete and reinstate the max-threshold number of characters (e.g., 30 or so). Those who want to avoid bumping can make a few small edits, but cannot do it for long without hitting the cumulative trigger. Of course the trigger could be set differently for the post-owner, other users, and moderators. $\endgroup$ – T.. Dec 23 '10 at 4:42
7
$\begingroup$

It seems to me that introducing such a threshold will be the more exploitable the more useful it will be. For example, if the threshold is very low, then it will bump questions even for minor edits and the number of cases in which it will apply might not even justify the added complexity of behaviour. If the threshold is higher, then there is the possibility of exploits as mentioned in that meta.SO thread. Already 4 or 5 characters are enough to include an offensive word at the end. 15 characters are easily enough to link to an unrelated external site for ad purposes.

What really surprises me is that there is a very simple solution to meet those objections: introduce a rep cap for the "minor edit" check box that has been proposed on meta.SO. I seriously don't understand why they don't do it. High rep users already have lots of ways of sabotaging the system, so if we drop the assumption that they won't, then there will have to be much more safety features in place than anyone would care to see. It would also prevent any exploits by bots. It would also have the advantage that even if I correct 5 spelling mistakes, I can still prevent the question from being bumped.

$\endgroup$
  • $\begingroup$ Most exploits add, complete, or uncomplete (and thus change the number of) strings of an identifiable type: HTTP links, (X).(Y).(Z) almost-links, email addresses, usernames, special words. Such events, as well as multiple edits by one user on different postings, or on one posting by multiple users, can be used as additional triggers for automatic bumping or flagging. (The 10K moderator tools take note of highly edited postings, for example). Is there any style of exploit this would not effectively prevent, but is blocked by the present system? $\endgroup$ – T.. Dec 23 '10 at 12:56

You must log in to answer this question.

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