Welcome to Codidact Meta!
Codidact Meta is the meta-discussion site for the Codidact community network and the Codidact software. Whether you have bug reports or feature requests, support questions or rule discussions that touch the whole network – this is the site for you.
Could we have a way to edit without bumping posts?
Whenever a post is edited, the question is automatically bumped to the top of the category post list.
This is useful for major changes to posts because people can see that something has changed. However, it is a slight annoyance when the edit is extremely minor, such as changing the tags of a question or fixing a spelling mistake. In those touch-up cases, there isn't really a significant change that warrants the increased attention.
Also, I'm sure nobody needs to be subjected to ten posts being bumped to the top of the feed just because a new tag was created, and old posts were updated with it.
I personally have avoided editing old posts for exactly this reason; we don't have much activity as it is, so having already well-answered questions be bumped and push down more recent unanswered questions seems counterproductive.
Could we have a way to mark an edit as "minor" or something, or otherwise have an option to not bump the post? This mark would ideally be applied by the edit reviewers, who we trust to make these types of calls.
Note that other than not bumping the post, there would be no functional difference from normal edits. The edits would still have to be reviewed, would show the edited indicator that includes who edited, and would still appear in the edit history, as normal.
Possible concerns raised
It will allow malicious edits to go unnoticed
[H]ow do you get around the case where someone makes a major or malicious edit, then tries to hide that by claiming it was minor? — Olin Lathrop
The review system exists for a reason. If someone attempts to maliciously edit another's post, that should be caught by the review system.
If they are editing their own post, well, the only harm done is to themselves.
If they can edit others' posts without review, well... I'd be more concerned about how a malicious person managed to gain that ability.
Bumping gives unanswered questions attention
There are better ways of giving attention to unanswered or poorly answered questions. Since they mention SE, I will note they have a specific tab for unanswered questions. They also have a bounty system, though it wouldn't be easy to set up here due to Codidact being relatively less rep focused.
Besides, it's not like this suggestion is to remove the ability to bump; it is to add the ability to not bump. If someone wants to bump a post to give it more attention, they still have that ability[1]
Edits should notify the author regardless of how minor
Edits to other people's posts should always generate notifications to the author, major or minor. — Monica Cellio
I certainly want to know about all edit anyone makes to my posts. — Olin Lathrop
I agree.
How would the minor label be determined?
We shouldn't try to programmatically determine what a "minor" edit is; in the right context, one character is a major edit. Designation as minor needs to be human-powered. — Monica Cellio
I personally think that the label should be applied at the time of review; We already trust reviewers to be able to judge good and bad edits, so it doesn't seem much of a stretch to let them judge whether to bump due to an edit.
-
Whether they should be doing such is another issue. ↩︎
Speaking for myself and not the team (I haven't discussed this with anyone else), here are some things I'd like to consi …
3y ago
By default all edits should show Every time an edit is made, the post or its parent should move to the top of the que …
10mo ago
While it would be nice for minor edits to not bump posts, the problem is defining what "minor" is, and leaving it for th …
3y ago
I agree, this is a problem when the site activity is low. For example I wrote a very long answer at Software Dev the oth …
3y ago
As @Lundin explained why minor edits shouldn't bump post. I will say why it has to. For old posts : - There's a l …
3y ago
5 answers
You are accessing this answer with a direct link, so it's being shown above all other answers regardless of its score. You can return to the normal view.
By default all edits should show
Every time an edit is made, the post or its parent should move to the top of the question list. This reduces the chance of an edit being missed by the community.
Not every edit is reviewed. Some people have the Edit Posts ability and can edit without review, and the post author can always edit without review. A reviewer's decision is also not reviewed. Showing the post at the top of the question list allows the community to provide an element of informal review and raise any problems they see.
Users can choose their own sorts and filters
If a user does not wish to see the default view of the question list, they are free to change the sort order or apply filters. For example, they may choose to view only unanswered questions, in order of age rather than activity. This way they would not see edits.
Since this option is already available, I see no reason to change the default view of the question list. All users who have not chosen to prevent it will see edited posts, reducing the chance of malicious or accidentally detrimental edits going unnoticed.
Bad edits can be arbitrarily small
While the idea of treating small edits as insignificant is appealingly convenient, this would allow a variety of problems:
- Deliberately offensive statements
- Dangerously misleading statements or code
- Spam and phishing links
A word or even a single character change can be sufficient to introduce any of these, so there is no lower limit which could be set as a safe threshold.
For example, the idea of introducing a malicious link sounds like it would require at least enough characters for a valid markdown link, but if the malicious editor is also the author of the post, they could post an innocuous link in the original post and later change a single character - malicious links are commonly a single character different from an innocuous link in order to take advantage of people who make typos while entering a URL.
Neither the editor nor the reviewer is well placed to decide
The editor, the reviewer, or both could be malicious, so allowing either of them to decide which edits should be hidden from the community is not appropriate.
This also applies to a post author editing their own post. A spammer should not be able to decide that an edit to their own post should be hidden.
Long game spammers can work in pairs to build up trust and then approve each other's edits. Getting past a single review is easy if you are in contact with a malicious reviewer. Getting past all of the eyes in the community is more challenging. Every edit needs to be available for the community to inspect.
Speaking for myself and not the team (I haven't discussed this with anyone else), here are some things I'd like to consider:
-
We shouldn't try to programmatically determine what a "minor" edit is; in the right context, one character is a major edit. Designation as minor needs to be human-powered.
-
However, the system should have some sort of threshold (TBD) beyond which, no matter what the submitter says, we don't treat it as minor. Even though we trust the editor's intentions, if you're overhauling most of the answer, it's good to make that visible for others to look over. Besides, it might change some votes.
-
Edits to other people's posts should always generate notifications to the author, major or minor. (I believe this is the case now at Codidact, so I'm saying: don't change that for minor edits. People should always have a chance to see what words someone else is putting in their mouths. Somewhere Else they don't notify for "minor" edits, and that's sometimes been a problem.)
-
I think we generally want edits to your own posts to bump. Vandalism happens, long-game spam happens, and if it's your own post then no one else is going to be notified. I think we could mitigate this by setting some sort of time limit -- maybe we only bump if it's been 24 or 48 or (TBD) hours since the last bumping edit? So you can do that flurry of fixes over the course of a day and only bump it once, but if you come back three weeks later and edit again, it'll bump. Or...maybe we say that if you have the edit privilege then you can make non-bumping edits to your own posts, but if your edit to someone else's post would have had to go through review, your self-edits bump? Thinking out loud here...
These are ideas for discussion; they're not ready to be proposed rules yet. That last one, in particular, needs a lot more thought.
Opinions?
0 comment threads
While it would be nice for minor edits to not bump posts, the problem is defining what "minor" is, and leaving it for the person who did the edit to judge. Hopefully, most people won't be subjected to the edit queue eventually, so that's not the answer.
I certainly want to know about all edit anyone makes to my posts. Most edits Elsewhere to my posts were to correct spelling and the the like. I have no problem with that. However, there were also too many cases where I wrote something a certain way quite deliberately, and someone else then changed it. The worst kind of edit is where someone "helpfully" inserts a mention of something I purposely left off because I felt it was a detail that was mostly distracting in the large scheme of things.
These sites have such low volume that currently a few bumped posts aren't much of a problem. We should leave things as they are now, and revisit this if bumped posts become a problem.
0 comment threads
I agree, this is a problem when the site activity is low. For example I wrote a very long answer at Software Dev the other week, then re-visited it later and corrected some typos etc. Then someone else found a typo too and edited it again. This could easily be mistaken for purposely "bumping" the post.
I really don't see why edits need to bump a post in the first place. It's almost always nothing but disruptive. The most obvious example is the issue with status tags added to bug reports here on meta.
New answers should bump a post, but I don't quite see why edits need to do so.
As @Lundin explained why minor edits shouldn't bump post. I will say why it has to.
For old posts :
- There's a lot of questions haven't answered in SE sites (It will happen to Codidact when very professional people will ask question). Suppose, I am having
grub-loaderissue on internet. I found some unanswered question inPower Codidact(The site isn't available yet but, I am using Codidact instead of SE). If I edit title/tag only in that post. So, it will bump. For that reason new user can see that post and some of them may answer on that post. If someone answer on that post than it will be helpful for others(who are facing the issue). So, minor is good.
For new posts :
- I found a post in Software Codidact which I can answer and, that is very hard. But I am familiar with that bug. So, no one answered on that question. But, I was very busy that moment. In Codidact there's no way to bookmark post as SE has. And, If I use Codidact than history won't be saved. I became very busy with my works that's why I couldn't answer on that post. So, when I come back again to answer on that question. I notice that I was using this site (Software Codidact) in private (also known as incognito) mode. Although I was searching that post. But, I couldn't find out. So, if you do minor edit in that post. Then, it will bump So, I can find the post easily cause, it will show in top (But, if doesn't bump. Then, it will remain unanswered). That's why minor edit is good.
While I was writing the post I was searching posts like this in SE also. I found a beautiful answer.
Minor edits can be good
1. Codidact is intended to be a top-quality Q&A site, meant not just for the OP, but for posterity. Thanks to search engines, questions and answers become authoritative for the whole Internet.
2. Spelling and grammar mistakes, even small ones, make posts more difficult to read, and negatively reflect on their quality as a whole.
3. Codidact has a very large community, who read and re-read many questions multiple times a day.
2 comment threads