Increase in Bots/Spam Accounts

[carbonize]

There are companies that offer to do captchas for you and charge by the thousand. Some does this by running free adult sites that require people to do captcha to get in and the captcha on the adult site is taken from one on a forum or wherever it's been taken from. That and OCR is getting very good these days. Captcha is rapidly becoming pointless and we need to move on to other things such as cognitive tests (click all the red squares etc) or saptcha which is the usual question and response. SMF actually lets you put in several questions and expected responses and then specify how many to use per registration etc.

[m^(2)]

saptcha which is the usual question and response. SMF actually lets you put in several questions and expected responses and then specify how many to use per registration etc.

I believe that for small forums (and other systems) it's a great option...do something custom and rely on not being worthy getting custom circumvent code from scums.
Sadly, for big ones it's just a waste of time.
There's ongoing research on tests different from captchas, but last time I checked (quarter ago?), all that were fairly good took human 30++ seconds to solve and got rather high human error rate. The problem here is that even if a computer can solve correctly 0.1% of tests, a 1 million strong botnet can make any number of accounts.

[guinness]

It's a shame that only a few are spoiling it for the rest of us! Maybe TPFC should be an invite only Forum?!

[m^(2)]

It's a shame that only a few are spoiling it for the rest of us! Maybe TPFC should be an invite only Forum?!

I'm strongly against it. It's too limiting. I'd argue there's too little life on TPFC forum and restricting access is going to make it significantly worse over time. I think that the problem is not worth using solutions as costly as this one...especially that I'm not convinced it would work.

[Hydaral]

Is there something wrong with the "Report this post" button?

[SYSTEM]

Is there something wrong with the "Report this post" button?

I think that button no longer limits spam enough.

http://www.portablefreeware.com/forums/ ... 815#p30815

In addition, usually no one notices if a spammer adds links to posts which have looked fine before.

[harrietjames]

I run a forum and (so far) found that having to approve new posters has left it spam free since adding that option in the admin side.

[carbonize]

Yes but having to approve new posts can get tiresome and can also result in seeing the same information repeated in multiple posts.

[guinness]

I'm strongly against it. It's too limiting.

Of course and even though I brought it up as a suggestion doesn't mean that I would be happy with it either

I'd argue there's too little life on TPFC forum

I agree and even more since Spam has increased (tenfold) at the start of the year.

[Andrew Lee]

Guys, thanks for your suggestions, but please understand that any solution has to be feasible within the framework of phpBB3.

I did some research on this and have decided to turn on the "new user post count" feature in phpBB3. Basically, all initial posts by a new user have to be approved by a mod up to a certain number (currently set to 1).

This means for a new user, he must have at least one approved post before the moderation restriction is lifted.

Admins at other forums have done this and have had good results with it. I think it is a good balance between inconvenience and freedom if it helps keep the spammers out. Let's try this and see how it goes...

[Hydaral]

This XKCD comic is apt:

http://xkcd.com/810/

[webfork]

This XKCD comic is apt

Ha! Nice.

[SYSTEM]

Thanks Andrew.

[m^(2)]

Just fount this:
http://highscalability.com/blog/2011/2/ ... 00-re.html
It may be worth trying.

[Hydaral]

Just fount this:
http://highscalability.com/blog/2011/2/ ... 00-re.html
It may be worth trying.

That sounds good in theory, but spam detection is not that hard if you don't have to worry about false positives, from that webpage:

Mollom does not allow you to present an example of a wrongly rejected chunk of content that should be accepted