Consider the following in a online chat project:
If a blacklist list contains bad words, all you would need would to do is open the list to see them.
But could you encode the bad words somehow? Like use Caesar cipher so it really becomes some gibberish. When you send a message, when deciphered it gives gibberish and you can match the gibberish and block it from going through, so all you need is to literally cipher bad words and put it in a list.

But if someone was able to decode the cipher system, then it would not be allowed. Unless you made it into a hash?
Edit: hashing like MD5 would be fool-proof.. but I don't know the algorithm

Most people on scratch don’t know about decrypting ciphers, however I would encrypt it in something a bit harder than ceaser cipher, maybe(if it’s not to much hassle) vignere

fphiladelphia wrote:

Most people on scratch don’t know about decrypting ciphers, however I would encrypt it in something a bit harder than ceaser cipher, maybe(if it’s not to much hassle) vignere

Maybe Enigma encryption methods… although you will need a lot of computing power to decrypt it I think. Nobody would be bothered to do that much work to see bad words.

I don't believe any encrypted blacklist is allowed. This question has been asked many times before, and from what I've seen, any form of encryption for blacklists wouldn't pass. It has to be a whitelist,

There are lists available that contain pretty much the whole English dictionary. Just look it up on Scratch or perhaps use @griffpatch's Scratcharia project for a whitelist list.

I'm sure the ST would hopefully allow encryption later when they further develop cloud data, or find a workaround.

No, I don't think blacklists in any shape or form are allowed. Why? Because people could easily work around them.

Let's just assume that “eaten” is a swearword. (It's not, but still.) Your blacklist would block the word “eaten”. However, people could still get around that by typing: “eat en”, “ea ten”, “e a t e n”, “3AT3N”, “E4TEN”, etc. You get the idea.

-ShadowOfTheFuture- wrote:

No, I don't think blacklists in any shape or form are allowed. Why? Because people could easily work around them.

Let's just assume that “eaten” is a swearword. (It's not, but still.) Your blacklist would block the word “eaten”. However, people could still get around that by typing: “eat en”, “ea ten”, “e a t e n”, “3AT3N”, “E4TEN”, etc. You get the idea.

But you could have a neural network AI that detects the probability of someone saying a banned word. It still wouldn't be allowed though.

(removed)

jromagnoli wrote:

-ShadowOfTheFuture- wrote:

No, I don't think blacklists in any shape or form are allowed. Why? Because people could easily work around them.

Let's just assume that “eaten” is a swearword. (It's not, but still.) Your blacklist would block the word “eaten”. However, people could still get around that by typing: “eat en”, “ea ten”, “e a t e n”, “3AT3N”, “E4TEN”, etc. You get the idea.

But you could have a neural network AI that detects the probability of someone saying a banned word. It still wouldn't be allowed though.

How would this neural network AL work?

-ShadowOfTheFuture- wrote:

No, I don't think blacklists in any shape or form are allowed. Why? Because people could easily work around them.

Let's just assume that “eaten” is a swearword. (It's not, but still.) Your blacklist would block the word “eaten”. However, people could still get around that by typing: “eat en”, “ea ten”, “e a t e n”, “3AT3N”, “E4TEN”, etc. You get the idea.

Why not add common spellings to work around the filter.. all you gotta do is encrypt it. (of which includes most ASCII)

Invisible_Factory wrote:

-ShadowOfTheFuture- wrote:

No, I don't think blacklists in any shape or form are allowed. Why? Because people could easily work around them.

Let's just assume that “eaten” is a swearword. (It's not, but still.) Your blacklist would block the word “eaten”. However, people could still get around that by typing: “eat en”, “ea ten”, “e a t e n”, “3AT3N”, “E4TEN”, etc. You get the idea.

Why not add common spellings to work around the filter.. all you gotta do is encrypt it. (of which includes most ASCII)

There's way to many workarounds.

If “eaten” was a bad word:

e aten
ea ten
eat en
eate n
e at en
ea te n
eat e n
e a t en
ea t e n
e a t e n
EAT3N
3ATEN
3AT3N
E4TEN
E4T3N
eat-en
eatan
eeten
eatten
EAT 3N
E AT 3N
E4T EN
ea-ten
e-a-t-e-n

Do I have to go on?

Charles12310 wrote:

jromagnoli wrote:

-ShadowOfTheFuture- wrote:

- Disapproval of OP's suggestion -

But you could have a neural network AI that detects the probability of someone saying a banned word. It still wouldn't be allowed though.

How would this neural network AI work?

Like any other neural network AI?

Or, just do a whitelist. It's painful the way I have mine set up, but it's so that I can add usernames to a cloud whitelist. I've abandoned that project because of its size. It won't upload.