How difficult is it for you guys to create a query that will delete user profiles that contain references to prescription drugs?
::sigh::...wiredesignz if that was all there was to it, this wouldn’t be an continuing problem, but thank you for once again bringing up the issue in such a respectful and professional manner.
I’ve tried kindly on numerous occasions to explain the whole picture to you and you just don’t get it. Even if you choose to spend your time and train your eyes on new registrants and browsing member profiles (which makes the impact of this type of spam self-inflicted), you do not have access to all of the metrics necessary to determine who is and isn’t okay to delete and ban. And even if you did, it’s very easy as a human to look at the information and say, “duh that’s a spammer”; it’s an entirely different matter to turn that judgement fully over to an automated solution.
This is not a fan site, we don’t have the luxury of being able to make mistakes with false positives, nor do we have the time or desire to pay someone to spend all day making those judgement calls.
I’ve always agreed with you that the spammers suck and that we hate them an equal amount, the difference in opinion is in what action to take, and what priority EllisLab should give it. Yesterday, a company holiday, was honestly the first day I’ve had in probably a year where I had the convenience to sit down and craft some solutions that should minimize the spam while fully meeting our needs.
First and foremost was getting rid of many known spammers from our membership, finding new patterns, new spam email domains, updating the ExpressionEngine.com Blacklist, and deleting thousands of spam accounts (follow the links in your Google search above to Error pages). We’ve been employing Bad Behavior since this summer, and it’s helped with bot traffic, but not much with spam since the majority of spam is posted by unfortunate humans hired on the cheap at terminals.
So I also went ahead yesterday and created an extension to integrate Akismet, fitting it to match our own needs, particularly in providing us with information about requests being flagged as spam, and in handling false positives. I’m happy to report that in less than 24 hours, over 80 registrations / profile edits have been successfully caught, with no false positives. I do notice that there is a particularly prolific brand of spam that our forums get that Akismet oddly does not recognize as spam, but perhaps it will learn. I’ve Blacklisted those patterns though, so new attempts to post those should still not make it through.