Duplicate Mongolian vowel separator can be used to avert swear filter...

Liam W

Well-known member
#1
The unicode character with code 180E can be used to 'avert' the swear filter.

Technically, as the word is different, I'm not sure if this can be classed as a bug. Now, don't ban me for this - but I do have to demonstrate somehow :p

****ty
shi᠎tty

The second one has this particular character after the 'i'...

Liam
 

Martok

Well-known member
#2
Isn't this just the same as using another character in its place? You can replace that character with a full stop, a comma, an underscore and they will all bypass the swear filter and of course the swear word will be quite clear to everyone. Of course, you can also use a '1' in place of the 'i'.

You'll never catch every variant of a swear word. The best I think you can do is filter out the main ones and if there's an offender who continues to bypass this, stick them on moderation.
 

Liam W

Well-known member
#4
Isn't this just the same as using another character in its place? You can replace that character with a full stop, a comma, an underscore and they will all bypass the swear filter and of course the swear word will be quite clear to everyone. Of course, you can also use a '1' in place of the 'i'.

You'll never catch every variant of a swear word. The best I think you can do is filter out the main ones and if there's an offender who continues to bypass this, stick them on moderation.
Not really, the main reason I flagged this up is because it looks like there isn't a swear filter when it's used - when people substitute letters, or add normal ones it's normally pretty obvious, this one isn't...

Liam
 

Liam W

Well-known member
#6
This example is obvious to me:

View attachment 56219

(That's a screenshot. ;))

But what @Jeremy linked to is roughly correct. And there are thousands of "confusables" so it's not really feasible to prevent (at least while maintaining performance. See: http://unicode.org/cldr/utility/confusables.jsp We always say that active moderation is what's needed.
What device were you on :/ I thought unicode worked almost everywhere :p

Anyhow, I see what you mean. Oh well, if I ever want to annoy an admin...
 
Top