Hi there!

We all know private messaging is rejected, and we all know the reason for it. However, I think I've discovered a way to private message others on Scratch — which would definitely be bad.

The new project page for Scratch has a feature where if you write a comment that the Bad Word Detector doesn't like, it will make the comment turn red and say “marked by filter.” The comment would not actually be posted online for everyone to see — however, this comment is still able to be viewed by the creator of the project in their notifications.

I tested this out by posting something similar to the following comment on a project made by @CrystalKeeper7:

Za-Chary wrote:

I am writing this comment on my laptop in real life. This is a test. Message me on my profile if you see this comment.

I think the phrase that the Bad Word Detector marks in this comment is the phrase “real life”. From there, @CrystalKeeper7 reported that he was able to see this comment in his notifications, but not on the project page itself.

I see this as potentially dangerous because this means that someone could send inappropriate messages to someone. Not only would the recipient be able to see the messages, but they wouldn't even be able to report or delete them. They could use “Contact Us”, but it's probably better that these messages did not get sent in the first place.

I would definitely try it with other comments that are even more inappropriate, but I'm not sure that I'd want to out of fear of getting alerted. I am suspicious that this would happen anyway.

This is definitely something that needs to be fixed. Let me know if I can explain anything else more clearly.

My browser / operating system: MacOS Macintosh X 10.14.2, Chrome 71.0.3578.98, No Flash version detected

It's a very good point. I had a notification of someone posting what they alleged to be their address in a comment in one of my projects. When I looked in the project the comment was not there so had presumably been reported and/or removed. Regardless of that, I could see the full address that had been posted in my notifications (though it was actually that of the White House).

I am seeing it to, Here are some test screen shots.
(Ignore that message)

To be honest, I don't really like how this system works in the first place. If the bad word detector detects an inappropriate word, it should just block the comment and show an error like it used to before.

dude341 wrote:

To be honest, I don't really like how this system works in the first place. If the bad word detector detects an inappropriate word, it should just block the comment and show an error like it used to before.

I think that that would be a good solution.

54329 wrote:

dude341 wrote:

To be honest, I don't really like how this system works in the first place. If the bad word detector detects an inappropriate word, it should just block the comment and show an error like it used to before.

I think that that would be a good solution.

It's always been like what I said, but recently, they've been testing this new “marked by filter” way instead on certain users, from what I can gather.

I don't really agree with the “picking random users to test a feature when they didn't even volunteer to test” way of bug-testing, this is what Google/Youtube does, and we all know how that's turning out for them right now.
Google did help with early development of 3.0, so I wouldn't be surprised if they gave the ST some “business tips” on how to run their website (by that I mean a contract forcing them to make certain changes otherwise they wouldn't help. Not saying that did happen, but just what I'm thinking from the top of my head)

dude341 wrote:

54329 wrote:

dude341 wrote:

To be honest, I don't really like how this system works in the first place. If the bad word detector detects an inappropriate word, it should just block the comment and show an error like it used to before.

I think that that would be a good solution.

Google did help with early development of 3.0, so I wouldn't be surprised if they gave the ST some “business tips” on how to run their website (by that I mean a contract forcing them to make certain changes otherwise they wouldn't help. Not saying that did happen, but just what I'm thinking from the top of my head)

This is unlikely considering there's no money involved so this just wouldn't have happened.

bybb wrote:

dude341 wrote:

54329 wrote:

dude341 wrote:

To be honest, I don't really like how this system works in the first place. If the bad word detector detects an inappropriate word, it should just block the comment and show an error like it used to before.

I think that that would be a good solution.

Google did help with early development of 3.0, so I wouldn't be surprised if they gave the ST some “business tips” on how to run their website (by that I mean a contract forcing them to make certain changes otherwise they wouldn't help. Not saying that did happen, but just what I'm thinking from the top of my head)

This is unlikely considering there's no money involved so this just wouldn't have happened.

Any proof of this?

Oh, I think I've gotten this. Somebody was messaging me on a project, and one of their comments accidentally got marked by the filter. I could see the comment in my messages, but I couldn't find the actual comment. I thought this was just a one time thing, I'm surprised to see others have been getting this as well.

I sent a link of this post to @Paddle2See and here's what he had to say about it.

Paddle2See wrote:

Thanks for the excellent write up. I do agree that it is a concern. However, I believe that all messages that are marked by the filter get sent to the human moderators for review - so if something inappropriate is going on, it will likely be seen. It would be better if it were not sent through notifications though.