Since we seem to be getting a spate of these folks lately, perhaps it's time to start looking at moderating new members and/or their first 1-5 posts. Spam doesn't seem to be a real problem as yet, but moderating might be something to have waiting in the wings if this forum becomes a bigger target. I suggest this because another forum I'm on absolutely refuses to mod new members and there are many days when we get hit with 20-30 new posts/comments which are nothing but spam and all from new "members" - and it's a small forum. I know other forums do this, and a couple sites I was a mod/owner of did as well, and there was like no spam. Just a thought.
I don't know how feasable your suggestion is, @shadowwalker, but if it's feasable for the mods it's probably a good idea. Mind you, our mods do seem to catch the spammers fairly quickly. I always report them if I see them. I don't know if reporting spammers actually helps or if it just irritates the mods further?
I've never really cared about irritating mods But, like today when I came online, there were 20 posts, 3 of which were pure spam. On the other forum, I quit reporting them because I was spending more time reporting them than I was reading the legit posts. As to feasibility, there's another extremely large writing forum where new members have to wait until they're "checked out" before they can post anything. I just think there's a point where a site ceases to be an occasional victim and becomes an active target - and there should be steps in place so members aren't spending a significant amount of time reporting these.
It's a perfectly feasible thing to do, but it's a Daniel-only parameter to set. No one else has the access. The function is the same as what keeps the Articles Subforum permanently under manual moderation. That subforum is the #1 target subform for spambots if anyone ever wondered why posts in that area always have to be approved. The forum actually already has an admittedly only partially successful automated spam catcher. It looks for certain parameters in posts (newbieness being one of the parameters) and shunts the threads to a Moderation Queue that the mods go through daily to filter out the real members from the not. Having such an initial "buffer zone" for new members wouldn't bother me in the least. It would just fall into a function we are already performing. I would worry, and I can predict this will be Daniel's worry, that it would prove frustrating for new members.
Well, I know that other forum that does this has no problem with frustrated members. They are told when they sign up that they have to wait for approval, which typically only takes 1, maybe 2 days. Same with the sites I was mod/owner. If someone is going to get frustrated over that, they probably aren't good writer material anyway
So ...do I take it that if a spammer is a new member who just joined 'today' that we don't need to report the spam? That it will automatically be caught?
No, if you see the thread then it got by the filter, go ahead and please report it. All the threads that get caught and shunted to the Moderation Queue are never seen by the members, unless they turn out to be valid threads that we then approve.
Geez. A lot of them got past the filter the other morning! All promoting the same thing ...some neuro whatsis. You guys must spend a lot of time pulling your hair out. It's like cold-calling salespeople? Why? Why do they do it? Surely NOBODY actually buys their stuff? Surely .....
Dealing with bots is a tedious task, to be sure, and I always feel an inner FAIL when a member has to report it before I catch it myself, but unlike cold-calling where there is a salesperson's salary to pay and offices and overhead to pay as well, spambottery is just a program sitting in a sever somewhere spitting out literally millions of "members" to countless forums and other venues across the whole of the web. If they get just a tiny fraction of hits, it's worth their while.
Any particular reason you know of that spam nearly always surfaces early in the morning? I wonder if that's a clue as to where they're coming from. I get up very early (5-6am) here in the UK, and that's when I see these spam messages most of the time.
The vast majority of spambots are birthed out of Fujian, China. There is a small cluster that originate out of various places in Pakistan, and the rest are of sufficiently better programming quality that they "spoof" their IP's so that they appear to come from innocuous places in the U.S., Canada, the U.K. etc. Our current wave of spambots is of the "better" variety so the filter is having a hard time catching them.
Part of the reason you see them is that you get up early in the UK. I'm usually the last mod to go to bed at night because I'm the farthest west, here in California. The farthest east mods are the Trians, but they're not much farther east than you are, so if you get up early, you still might beat them. All that has to happen is for the spambots to post after I've gone to bed and before the Trians are up, and China is a great place to do that from. You probably wouldn't see anywhere near as many spambot messages if we had a mod in, say, Japan or Australia. They could be up and active while the spam messages are posted, and so would be able to catch them before you see them. That's why it's desirable to have mods around the world, so all time zones can be covered.
Well, with all respect to Chinese people, this might explain why the English in some of these spam messages is so incredibly fractured. It's not their first language. One of the spams I awoke to a couple of days ago was so fractured I thought it might be some new-wave kind of writing! It wasn't till I got to the link at the end that it dawned on me....
A few of us should post spam on Chinese writing sites. We could post stuff like "Why buy a fake passport when you can buy a real one?"
I haven't used this forum software, but the ones I have used (phpBB and Simple Machines) have the option of a Q&A captcha, which, in my experience, stops spambots cold. The sort of captchas being used here are broken not long after they appear, especially if they are supplied by the forum software author. You won't stop a spammer willing to manually create an account, but they're rare. I can set up a forum and forget to set the captcha to Q&A, and come back a few days later and have one to two hundred 'members' signed up. With Q&A - zero. A Q&A captcha asks an usually easy question and accepts whatever answers you supply when you set it up. One I use: Q - "How many in a quartet?" A - 'four' or '4'. It doesn't really matter what the question is; the point is that the spammer would need one powerful AI-enabled computer to solve it. There are also captchas which display an image of, say, a book cover and ask the registrant to enter a word from the book's title - a variation on the Q&A method. I really wish we could figure out a way to solve the spam problem on the demand side. If users would stop clicking on the links in spam messages, these creeps would disappear. Edit: I just created a demo XenForo installation and took a look at the admin options, and the option to use Q&A captchas is built in, under Options - User Registration. Setup is straightforward.
the fractured English is somewhat deliberate. it's randomly jumbled from a bunch of copy-pasted sources, so that the posts aren't identical (one thing spam filters catch easily) but they look like "real" posts. nobody's sitting down and writing them; the program generates chunks of text itself. the person running the program doesn't even have to understand English. most of the time the output is complete garbled nonsense, but sometimes it's kind of poetic.
Yes it is. Appreciate spambots. The last resort for chat. Keep their comments. For me, their arrival on the blog can be hard to fathom: alluring, mysterious, although sometimes they are very kind. Suppose, I really yearn 'Your great advices of priced maximum life.' 'You're so pretty.' But they never return.
there's been more than one SF story based on the idea that the first AI emerges from a spambot--usually a spambot/virus combination in the wilds of the internet. I rather like that idea, though it's probably the worst-case AI emergence scenario for humanity...
I hope it never comes true, as it would be sad that the first signs of artificial intelligence is to sell me a porn subscription.
"Superintelligent AI emerges from internet, takes control of world's computer systems, and subjugates humanity, forcing everyone on the planet to buy Viagra and enroll in diet programs."
Just in case there were any curiosity, this is what the Moderation Queue looks like. These are the ones no one ever sees but the mods. This is a thread of my own I just created and forced into the queue. They each get a do nothing (not sure the purpose of that), approve, and delete option. We just go through them, read them, check to see if they are a go or not and click the appropriate option. If they are new RPG games waiting to be given the green-light, we leave that for the RPG mods.
Whoever's behind all the bots is a real spammer in the works. ...It's alright, I can show myself out.
