Not really a bug or a question so ig suggestions is the best place for this.

I've been seeing people make comments in which they say a banned word, but use a Scratch emoji to replace one or more letters so the word isn't banned

Here is an example of what I mean (using appropriate words so I don't get in trouble ofc though):

ExampleGuy wrote:

Hllo! You are mzng!!

In this example, ExampleGuy used to be able to say the words “hello” and “amazing”, and using this method they were able to bypass the (nonexistent) filter preventing those words to be used.

I have seen this a lot in project comments, usually using the waffle emoji, and oftentimes as a way of hating the ST (something like I hte the ST!!)

Scratch should do something to make it so this can't happen, as people have been saying some nasty words through this workaround.
I can't come up with any solutions at the moment, so please come up with ideas on how to fix it.

Edit: I forgot to mention that people are also using this workaround not just replacing letters, but adding the scratch emoji in between letters so a word doesn’t trigger the filter but it still shows the full word. For instance, saying “I ha_waffle_te you”. Scratch filter reads this as “I ha waffle you” and not “I hate you”, but if you remove the waffle it’s clear that they’re saying “I hate you”. A possible solution is for emojis to be added through an emoji keyboard rather than using words, and then emojis are ignored in comments, but I’m unsure how easy or hard an emoji keyboard would be to implement.

That's only if the person already knows the implied swear word.

Until the scratch teams modify the filter, report it

+~1

Oh, I got it… Support, they just use emojis to replace letters to prevent muting, which's unwanted.
I was surprised, how did they…??

Support! Also,

vudeptrai090211 wrote:

Oh, I got it… Support, they just use emojis to replace letters to prevent muting, which's unwanted.
I was surprised, how did they…??

I bet someone was using the bot to see what would be a bypass, and what wouldn’t.

We need to do something about this, it's really bad if there is any way for children to see swearing on the website.

No support, I feel that if a filter was added to emojis, it would have alot of false positives like the number filter system.

Also the fact that you could just report the post and it would be taken down. Ive never seen this before on this platform before either.(also it's not a big deal if a kid sees a omitted curse word lol)

Semi support, because this is basically impossible. The number of swears is a lot and st managed to add all of them. More so, The combo of swears letter being replaced is over 100 thousand. There are so many combos in words like awesomeness, and it is too hard to implement. However, if this was implemented I would find this to cut back swear words. There wouldn't be a lot of false positives because this would directly censor the swears. If this was implemented, it would remove swear words which is always good on a kid's site.

XxhackerlolxX wrote:

No support, I feel that if a filter was added to emojis, it would have alot of false positives like the number filter system.

Also the fact that you could just report the post and it would be taken down. Ive never seen this before on this platform before either.(also it's not a big deal if a kid sees a omitted curse word lol)

The number filter system is needed!

Maybe have the filter ignore emojis completely? I think that would solve this problem.
stupid 60 second rule

Yellowsheep43 wrote:

Maybe have the filter ignore emojis completely? I think that would solve this problem.
stupid 60 second rule

I agree, this is most likely the best solution.

Yellowsheep43 wrote:

Maybe have the filter ignore emojis completely? I think that would solve this problem.
stupid 60 second rule

I support, this is a problem. Yellowsheep's solution would be the best way to go, i think

7salad3salad wrote:

Yellowsheep43 wrote:

Maybe have the filter ignore emojis completely? I think that would solve this problem.
stupid 60 second rule

I support, this is a problem. Yellowsheep's solution would be the best way to go, i think

Doesn't the filter already know when you swear and censor it (say fire was banned, they would then use f*re)? In that case, treat emojis as asterisks and censor accordingly.

This is like saying (if heck was a bad word) hec* but it would be hec which is basically the same thing

By the way, I’m not really sure how much of a suggestion this is, it’s more pointing things out. I’m not quite sure how they’d solve this issue, though ignoring emojis when blocking things could work.

Bump

7/8 support and 1/8 no support
support cuz - of reasons stated above lel

no- support cuz - say , if “yuck” was censored, would i get censored for saying d u c k ? seems the case with grapes.

Yellowsheep43 wrote:

Maybe have the filter ignore emojis completely? I think that would solve this problem.
stupid 60 second rule

I disagree. Let's say “tera” was a bad word. Someone wanted to comment it, so they said “tea”. If the comment was posted, the filter would read “tea” which is a common word, so the user would get by saying “tera”, a bad word (in this scenario).

A better solution would be for the filter to replace the emoji with every letter and see what it spells with each letter. So with the above, the filter would see “teaa”, “teba”, and so on until it reached “tera”, the bad word (again, in this scenario), and the commenter would get muted.

Bump

I don't see any easy way to fix this, but you can and should report any post where someone says that.
Making a filter see this would be very hard, and people might commenly get muted for issues with this so just report it.

ScratchCat1038 wrote:

Yellowsheep43 wrote:

Maybe have the filter ignore emojis completely? I think that would solve this problem.
stupid 60 second rule

I disagree. Let's say “tera” was a bad word. Someone wanted to comment it, so they said “tea”. If the comment was posted, the filter would read “tea” which is a common word, so the user would get by saying “tera”, a bad word (in this scenario).

A better solution would be for the filter to replace the emoji with every letter and see what it spells with each letter. So with the above, the filter would see “teaa”, “teba”, and so on until it reached “tera”, the bad word (again, in this scenario), and the commenter would get muted.

Then how about it filters through every letter as well as null (nothing, not even a space). That way, we can cover every scenario.