A new censor

[thefluke]

We're looking to get some help getting a more powerful censor for the boards. We have some very.... creative users at our current boards, and they seem to find joy in going around the censor we have in place.

The default censor just isn't powerful enough for what we need, and seems to be case sensitive (not sure why that is :/ ).

Anyway, her'es sort of a shopping list:
A configurable filterwith the user's ability to turn the filter on and off with the following conditions:

1) words wich will always be censored, regardless of weather or not the censor is on or off for the user.
2) the ability to replace words using it
3) no limit on the number of charachters in the censor (currently 250 max)

A copy of our old censor is here. (http://w1.480.telia.com/~u48024396/censor.html) Some of the symbols in there were to prevent users from using them in thier names (it tended to break the search function). The websites are generaly links to keyloggers or files with virii in them, and all of those pictures are links to those pictures that seem to find their way onto every respectable forum at least once (tub girl and that one gentleman who is... well, I'm sure you've seen it.).

Anyways, can anyone give us a hand?

[Logician]

As a matter of fact, your best bet is to change your rules and make it clear that if someone intentionally passes around a censorword, he will get banned. G-force's User Warning hack may help you achieve this goal too. This is because:

1- There is no way to stop a creative user pass a censorword. You can ban word "Microsoft" and a few more similiar versions like "Micro Soft", "Micro$oft", "M.Soft". but there are literally hundreds of ways to mean it like:
"M i c ro s o f t", "MI CRO SOFT", "Miicrosoft", "M-i-c-r-o-s-o-f-t", "Micro+soft". I think you get the idea.

So if a user intents to use the word, this way or that way he will. However according to me, if a user notices that forum administration is banned the word when he first used it in the normal form, then edits his post to make it visible, by all means he deserves a good ban.

2- Filtering is quite server intensive and that is why it is restricted with 250 words and there is no "user specific filters" in vbulletin. Say, I'm using "show 40 post per thread" setting here in vb.org so vbulletin display me 40 posts per thread. Say each post has 150 words in it which makes 6000 words per page. Say admins set 250 filter/replacement words in options and I have my own of 250 which makes 500 words. Can you imagine the server load and slowness of process if vbulletin passes through a 500 words list to match them in 6000 words in every thread I read here. You'll soon get the server on to its knees.

So maybe my message did not help you but it may clear the ground why the help you are asking is not so easy to get in technical aspect. ;)

[thefluke]

Today at 10:16 PM Logician said this in Post #2 (https://vborg.vbsupport.ru/showthread.php?postid=391892#post391892)
As a matter of fact, your best bet is to change your rules and make it clear that if someone intentionally passes around a censorword, he will get banned. G-force's User Warning hack may help you achieve this goal too. This is because:
We've tried that. The only problem is with 1.5k new posts per day, you sometimes have to ignore the censor abuse and go for the bigger offences. I was looking for something that we could do to automate that.



2- Filtering is quite server intensive and that is why it is restricted with 250 words and there is no "user specific filters" in vbulletin. Say, I'm using "show 40 post per thread" setting here in vb.org so vbulletin display me 40 posts per thread. Say each post has 150 words in it which makes 6000 words per page. Say admins set 250 filter/replacement words in options and I have my own of 250 which makes 500 words. Can you imagine the server load and slowness of process if vbulletin passes through a 500 words list to match them in 6000 words in every thread I read here. You'll soon get the server on to its knees.

So maybe my message did not help you but it may clear the ground why the help you are asking is not so easy to get in technical aspect. ;)

My intent was to not have users add words to the censor, but to exempt certain users from seeing censored posts- that is if 2 people view the thread, one can see the posts, as written, the other will see the posts after they have been censored. Certain words will always have to be censored, due to agreements with the host.

I'm not sure if that is in anyway different from a technical aspect, though it sounds like that might make the load slightly less? I'm still new at this sort of thin, so I could be very wrong.

[Logician]

We've tried that. The only problem is with 1.5k new posts per day, you sometimes have to ignore the censor abuse and go for the bigger offences. I was looking for something that we could do to automate that
I could suggest such a hack idea then: Have a hack which will check the post before written into the DB (at the moment it is sent) and display a warning to the user if his post includes censor words and tell he is not permitted to use "similiar words" to bypass this ban and ignoring the rule will cause an instant ban. I believe that this will reduce use of such words dramatically and you can ban the rest of the crowd who insist not to comply more easily as they will few of them.

My intent was to not have users add words to the censor, but to exempt certain users from seeing censored posts- that is if 2 people view the thread, one can see the posts, as written, the other will see the posts after they have been censored. Certain words will always have to be censored, due to agreements with the host.

I'm not sure if that is in anyway different from a technical aspect, though it sounds like that might make the load slightly less? I'm still new at this sort of thin, so I could be very wrong.
VB checks censor words both at the moment post submitted AND at the moment it is displayed. So the words exist in the posts are already converted into *** and you can not revert them.

However if you intend to use the hack for the future posts, yes it maybe possible then.

My reservations regarding server load if censor list is large or passing them easily by typing similiar words still apply though..

[thefluke]

Today at 10:52 PM Logician said this in Post #4 (https://vborg.vbsupport.ru/showthread.php?postid=391909#post391909)
I could suggest such a hack idea then: Have a hack which will check the post before written into the DB (at the moment it is sent) and display a warning to the user if his post includes censor words and tell he is not permitted to use "similiar words" to bypass this ban and ignoring the rule will cause an instant ban. I believe that this will reduce use of such words dramatically and you can ban the rest of the crowd who insist not to comply more easily as they will few of them.


VB checks censor words both at the moment post submitted AND at the moment it is displayed. So the words exist in the posts are already converted into *** and you can not revert them.

However if you intend to use the hack for the future posts, yes it maybe possible then.

My reservations regarding server load if censor list is large or passing them easily by typing similiar words still apply though..

The only posts on the forum currently are those of the admins getting it raeady for the move from phpbb to vb, so old posts are not currently an issue. The idea of having it turn on or off was an attempt to prevent them from attempting to go around the censor, thus making the list smaller (my beief is that they swear just becasue they aren't allowed to, most of the offenders are teenagers. Making it so that they see themselves swearing would cut down on the amount of swearing they do and the number of combinations we would have to censor.)

I'll bring up your idea about the censor warning with the other admins though and see what they think about it. It may be a bit of a hassel if it came up every post, and would only work if the original post contained censord words, and not misspellings, but it could be a workable solution.

Thank you for your help.

[007]

Ok I just made this: https://vborg.vbsupport.ru/showthread.php?s=&postid=402437

:)

[thefluke]

Today at 07:12 PM 006 said this in Post #6 (https://vborg.vbsupport.ru/showthread.php?postid=402439#post402439)
Ok I just made this: https://vborg.vbsupport.ru/showthread.php?s=&postid=402437

:)


Thanks, I'll have the person on our board who handles the handwavy programing bits look at this. It looks good though. :D