I edit my comments like mad, but I think that showing the last modification time, which is presently done, mostly solves the problem in a better way than just a binary “edited or not” flag with a delay.
I believe that the reason you need to show that a comment has been modified is to deal with two problems:
Putting a popular comment in place, letting it accrue a lot of upvotes, and later editing it to contain wildly different content. This gives the new content unwarranted credibility.
Writing a comment, waiting until someone responds, and then editing the parent comment so that their content is placed in a different light.
Having a comment last-modified time does a better job of solving #2 than a binary “edited or not” indicator (since one can see whose comment was edited most recently).
One can’t always infer enough information to address #1 from just a last-modified time, but generally, one that’s way out-of-whack from surrounding comments in modification time is a red flag.
I think all of the clients I’ve seen just show the last-modified time, and I don’t know if the API exposes both original and last-modified time.
So clients have all the information they need to know how long a comment stood between initial publishing and last modification. You could maybe just have clients show the delta between publishedAt and updatedAt. That’s a superset of what a boolean “edited after grace period” indicator shows.
Ninja edits. A 30-second grace period to edit a comment before it’s marked as edited. It’s for fixing typos.
I edit my comments like mad, but I think that showing the last modification time, which is presently done, mostly solves the problem in a better way than just a binary “edited or not” flag with a delay.
I believe that the reason you need to show that a comment has been modified is to deal with two problems:
Putting a popular comment in place, letting it accrue a lot of upvotes, and later editing it to contain wildly different content. This gives the new content unwarranted credibility.
Writing a comment, waiting until someone responds, and then editing the parent comment so that their content is placed in a different light.
Having a comment last-modified time does a better job of solving #2 than a binary “edited or not” indicator (since one can see whose comment was edited most recently).
One can’t always infer enough information to address #1 from just a last-modified time, but generally, one that’s way out-of-whack from surrounding comments in modification time is a red flag.
I think all of the clients I’ve seen just show the last-modified time, and I don’t know if the API exposes both original and last-modified time.
searches
Apparently yes.
https://www.socialfetch.dev/docs/api/v1/lemmy/post/comments/get
You have the “publishedAt” and “updatedAt” times.
So clients have all the information they need to know how long a comment stood between initial publishing and last modification. You could maybe just have clients show the delta between publishedAt and updatedAt. That’s a superset of what a boolean “edited after grace period” indicator shows.
Pretty sure lemmy has that.
edit: test
Shows as edited for me, at least.
Not for me.
Does seem like Lemmy, unlike Mbin, records that information. And then it’s up to each client instance to honor or not honor the ninja edit window.
I have fat thumbs and almost every comment I make is edited.
Pretty sure Mbin has that, let me test. Edit.
Magic!
Tried fetching the JSON-LD, don’t see the
updatedproperty on it like with remon’s Lemmy test which does have it.I see the edit marker on piefed.