6
\$\begingroup\$

It would be FANTASTIC if editors can self-flag edits as trivial so as to avoid impacting the activity list. Kudos to those who want to correct grammar/spelling/caps/etc for archival purposes when they have a few moments, but they should have a way to do this without floating minor corrections back to the top.

Related: Editing for quality: The "lipstick on a pig" problem

\$\endgroup\$
7
  • 2
    \$\begingroup\$ +1. CodeProject site has a feature like that. \$\endgroup\$
    – Nick Alexeev Mod
    Commented Oct 7, 2014 at 16:15
  • \$\begingroup\$ Excellent suggestion/topic. \$\endgroup\$ Commented Oct 8, 2014 at 4:02
  • 5
    \$\begingroup\$ Good suggestion. I do make a lot of "trivial" edits and do wish they wouldn't be confused with new posts or major edits. While we're on the topic, I really wish they'd implement some way to notify downvoters about edits; I'd happily review questions or answers I previously downvoted if they've been recently edited. \$\endgroup\$
    – JYelton
    Commented Oct 8, 2014 at 22:12
  • 2
    \$\begingroup\$ You could simply stop making unimportant edits... \$\endgroup\$ Commented Oct 10, 2014 at 16:57
  • \$\begingroup\$ @Chris If edited items did not bump to the top of the activity queue, would you feel differently? \$\endgroup\$
    – JYelton
    Commented Oct 15, 2014 at 18:24
  • \$\begingroup\$ Somewhat, though I'm still opposed to excessive, trivial meddling in other's work. \$\endgroup\$ Commented Oct 15, 2014 at 18:57
  • 1
    \$\begingroup\$ @ChrisStratton I do not think that minor edits - small changes are unimportant - not relevant. The sum of small changes on a post, accumulating over years, it part of what makes StackExchange special. The quality of posts that are read often can get to pretty high levels. \$\endgroup\$ Commented Oct 21, 2014 at 16:21

4 Answers 4

3
\$\begingroup\$

The reason why all edits bring something to the top of the active queue is so that other people can verify that the edits are not harmful.

The two issues that come up are that it can be difficult to (automatically) identify a trivial edit, and that you are basically giving someone write access to the entire site without oversight. I don't think these are necessarily insurmountable, but the current system is trading the effort you use scrolling further down the page with transparency.

\$\endgroup\$
5
  • 4
    \$\begingroup\$ I thought that's what the edit review queue was for. I agree about problems ID-ing a trivial edit, but I think this system should be self-identified trivial edits, with a default for real edits. If we're worried about vandalism, make it a priv for higher rep, but I think the edit review queue will take care of it. Also, if popping it to the top for review is to take care of harmful edits, that's not too helpful, as it takes some real effort to figure out that it's been popped because of an edit. \$\endgroup\$ Commented Oct 7, 2014 at 17:57
  • 3
    \$\begingroup\$ What about creating a review queue to accept edits to old posts? They wouldn't get bumped but would be screened by established users. \$\endgroup\$
    – clabacchio Mod
    Commented Oct 7, 2014 at 18:27
  • \$\begingroup\$ Wouldn't the edited posts get bumped up once the review is accepted? \$\endgroup\$ Commented Oct 7, 2014 at 21:40
  • \$\begingroup\$ Though perhaps 5k users would be able to accept in such a way that small edits don't float to the top. \$\endgroup\$ Commented Oct 8, 2014 at 10:44
  • 1
    \$\begingroup\$ No, the system could easily recognize trivial edits. If only one or two letters in a word change, if the same change is made in multiple words, Cap/Uncap only edits, punctuation only edits. These would target typo and punctuation which are trivial aka minor. \$\endgroup\$
    – Passerby
    Commented Oct 8, 2014 at 21:46
2
\$\begingroup\$

I agree with this concept. I have not edited posts in the past because I didn't want to bump the question to the top of the list.

I understand the risk is that maliciously tagged trivial edits can go unnoticed, but that can be addressed with the following mechanisms:

  1. Edits can only be flagged as trivial below some maximum number of changed characters. The system already tracks this, since it says something like "24 characters were changed" in certain places.

  2. Require some minimum rep to allow a unreviewed trivial edit. Others can still mark a edit as trivial, but the edit goes onto the review queue and the reviewers have to agree it is trivial. The bar for this should be higher than allowing unreviewed non-trivial edits. That is currently set at 2000 rep, so maybe 5000 rep to allow unreviewed trivial edits.

  3. Add a flag for inappropriate trivial edit. In the very rare case where one person makes a unreviewed trivial edit and another person notices and disagrees, a mod can step in and decide.

  4. The penalty for inappropriate trivial edit could be substantial, like -500 rep, loose ability to claim trivial edit for some time or until the penalty rep is gained back, etc.

The system has to be able to protect itself, but I really think that inappropriate trivial edits will be very very rare, especially since it would be done by someone around the site long enough to have accumulated enough rep to be allowed to make un-reviewed trivial edits. I just don't see this causing a problem.

\$\endgroup\$
5
  • \$\begingroup\$ I like the approach, but I'm not quite sure about a penalty. Also, I'd make the trivial flag an option -- it's not really an issue for me if an occasional edit gets bumped to the top of the activity queue, but those instances where somebody with free time makes six or seven edits (again, kudos!) can really change the flavor of works in progress. \$\endgroup\$ Commented Oct 7, 2014 at 18:37
  • \$\begingroup\$ @Scott: I wasn't saying the system would detect trivial edits, only detect when something isn't allowed to be trivial. You'd still have to manually mark a edit as trivial to be treated as such. But, that option would not be available if the edit appeared too major to the automated system. \$\endgroup\$ Commented Oct 7, 2014 at 21:54
  • \$\begingroup\$ Why would trivial edits require more rep than non-trivial edits? That's stupid. \$\endgroup\$
    – Passerby
    Commented Oct 8, 2014 at 21:43
  • \$\begingroup\$ @Passe: Read what I actually wrote and think about it a little. Anyone can do a trivial or non-trivial edit. It would require more rep to do a non-reviewed trivial edit because there is greater risk to the system. Since it doesn't get reviewed and doesn't get bumped to the top of the active list, it's quite possible nobody will notice a non-reviewed trivial edit. That's why we need to have more trust in someone allowed to perform such actions. \$\endgroup\$ Commented Oct 8, 2014 at 22:36
  • \$\begingroup\$ The altered characters limit won't work for those obsessed with altering capitalization and other useless trivialities though. I suppose it could ignore capitalization changes and formatting punctuation, but to a degree I'd rather simply see unproductive meddling discouraged. \$\endgroup\$ Commented Oct 10, 2014 at 16:58
2
\$\begingroup\$

Here are some new changes to the editing system and badges. These changes don't directly address the ideas above, but might change the milieu under which they are understood.

\$\endgroup\$
2
\$\begingroup\$

To compromise between this feature-request and W5VO's point that bumping the front page can be good to verify edits (although I agree with one of the commenters that this is done in the review queue already, I see why the front page can be useful still), maybe it would be possible to add this as an option for reviewers: accept, but don't bump the front page. The edit would not bump the front page if all the reviewers that accepted the suggested edit chose the option. If there's one who thinks it should bump the front page, it would do so.

Then eventually there could be a checkbox for this on the edit page for 2k+ users, whose edits don't come up in the queue.

\$\endgroup\$

You must log in to answer this question.

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