6.1 Delete or edit your comments everywhere
The inability to delete your comments everywhere is put in place so users do not post inappropriate comments or spam comments on other users' profiles, projects, or studios, just to delete them later on. In contrast, you are not really likely to spam your own projects or profile. The only benefits to this are to remove comments that you did not mean to post and to fix spelling errors. The former could be solved by thinking carefully before posting any comments, and the latter could be worked around by typing slowly and/or stating your mistakes in a reply to the comment. This also applies to forum posts; that is, a feature for Scratchers to delete their own forum posts will not be implemented. This extends to the ability to delete your own comments even if nobody has seen them yet.

Editing comments could similarly lead to comment manipulation, such as a user commenting “I like this project,” waiting until someone replies with “I agree,” and then editing the original comment to say “I hate this project.” Allowing Scratch Team members to view the edit history of a comment does not make it easy to moderate, since then they have to look carefully at the context of each edit, which takes time.

Sometimes, because of typos, you accidentally type a bad word. The keyboard letters are put very close together and some people might type quickly. Then, they might have to stop commenting for a long time. My suggestion is to make a warning pop up if your text contains a bad word that the detector detects. It will ask ‘Are you sure you want to send?’. If you click ‘yes’, it will block you from commenting like normal. If you click ‘no’, it will show a ‘edit’ or ‘cancel’ button. If you click ‘edit’, you can edit your comment and resend. If you click cancel, it will help you delete the comment.

Pros and Cons

Pros

• People won't accidentally typed a bad word.
  

• People might not know that the word will trigger the detector. If the person knew, he / she probably won't send.
  

• There won't be so much as a wrongful comment ban. Lots of people complain about them getting comment banned but the detector didn't read the context. Adding this, it will warn the people to not send.
  

Cons
None l can think of

Edit : There'll be a limit. Probably 5 times per day. On the sixth time, there'll be no warnings and immediately mute them.

So users would never be blocked from commenting unless they get this warning and then click “yes”? It seems like this would work for a bit but word would spread around pretty quickly that the only want to get muted is to click “yes” and that if they don't do that, then they can infinitely revise their comment and never be muted, making it very easy to figure out how to bypass the filter. So this would make incorrect mutes less likely, but correct mutes would also be less likely.

Presumably the only way this suggestion would work is if instead of the only way to get muted being to say “Yes”, instead, if the revised version (or several revised versions) also contain errors then you no longer get the confirmation and it just mutes you. What you're describing then is actually very similar to how the filter worked prior to February of 2021, but presumably the ST changed it for a reason.

mrsrec wrote:

So users would never be blocked from commenting unless they get this warning and then click “yes”? It seems like this would work for a bit but word would spread around pretty quickly that the only want to get muted is to click “yes” and that if they don't do that, then they can infinitely revise their comment and never be muted, making it very easy to figure out how to bypass the filter. So this would make incorrect mutes less likely, but correct mutes would also be less likely.

Presumably the only way this suggestion would work is if instead of the only way to get muted being to say “Yes”, instead, if the revised version (or several revised versions) also contain errors then you no longer get the confirmation and it just mutes you. What you're describing then is actually very similar to how the filter worked prior to February of 2021, but presumably the ST changed it for a reason.

How bout this. There's a limit per day. Like if he tries more than 5 times a day, he'll be muted on the 6th time.

This is an amazing solution.
Support.

Support, but maybe less than 5 times a day instead.