Spambot Prevention Suggestions

FIRST General Forum
#59

One thing I’ve noticed is that most of the spam has upwards of a dozen hyperlinks in it (most somewhere around 20). They also all seem to appear as new threads.

Would it be possible to put an additional level of security/captcha/mod approval type thing on any new threads that get started with more than 10 hyperlinks? I rarely see any threads started by actual forum users that have that many hyperlinks.

It’s a bit of a stopgap, but should at least slow the current influx of spam threads we’ve been getting recently. Mod approval might be the best way. If someone REALLY wants to get a thread started with lots of hyperlinks, and needs it fast they can always send Brandon Martus or another mod a PM/email. Otherwise they remove the hyperlinks or wait.

#60

This is a good idea, but in reality, a spammer often starts a thread and edits in the hyperlinks a few minutes later. That would be a little harder to deal with.

#61

I thought that the forums had a ‘you must have this many posts before you start a thread’ rule. It seems that there are a lot of spam threads that are a first post by the user. Could something like ‘you must post 10 times before you start a thread’ rule be put in place? Or something that required mod screening then locking the thread for editing?

#62

There isn’t one, and I don’t think there will ever be one. Most folks making a first post as a thread are trying to find help; even if there’s a thread on that topic on the portal they might not know to look at it until AFTER their first thread.

But the bigger reason for not having a “X posts to start thread” is this: They’ll (both spammers and non-spammers) just post all over OTHER threads looking for whatever they’re looking for, often wildly offtopic–and the spammers do that already. So instead of seeing new thread with a lot of links pop up, you’ll see an older thread pop up and think that “Oh, someone just posted, let’s see what’s up–Rats, spam” after opening the thread.

There is an “X posts to Y” rule, but it’s 50 posts and positive reputation to be able to give non-neutral reputation.

#63

Am I just crazy, or does the “this post was edited by…” text rarely show up when that happens?

Also, I was looking at Alexarankings today, looked up Chief Delphi, and some interesting results:

http://www.alexa.com/siteinfo/chiefdelphi.com

43.8% of CD visitors have an IP address in India.

#64

That text only shows up after a certain amount of time has passed. I’ve gone back and immediately edited a post after I submitted it because I noticed something I wanted to change, and the “edited by…[date]” doesn’t show up.

#65

It’s not very long, though. Typically about 1-2 minutes, I think, from when you hit the post button to when you hit the edit button/save changes button.

#66

Some of the other info there is really weird too.
The top three search terms are
1.delphi tableviewer
2.delphi forums (Ok, makes a little sense)
3.driver swag (???)
4.chief delphi (why is this number 4?)

#67

That is cool.
OFF TOPIC: Also, i wonder where the web industry is going. My site running on a Pi has quite a high rank and is 50% speed rank.

#68

I just went through 10 pages of search results for “Driver swag.” I found nothing but lost hopes for humanity. All the others have this forum pop up on the first page.

Spock eyebrow raise
Fascinating.

#69

Tangent: About half of the spam posts titles are mildly informative about world news events. If you’re too immersed in robots, it’s a good way to keep up to date on current events.

That being said, it’s still a net detriment. (no pun intended)

#70

Kind of off topic:
I placed a link to my forum (devstuff.no-ip.info/forum/) and said how wimpy it was (it runs on a raspberry pi). Apparently, the forum members are so helpful, they DoSed it! After reviewing the log files and other stuff, I found out that it was just because it was overloaded with requests. That gave me an idea of where I need to patch it up and now I am working on adding a maximum processes limit that will prevent apache from hogging all the resources from the kernel. The funny thing was that this time, the crash was so severe, even my UART connection was useless for degugging :). That made me do a hard reset :(. I do not think that that is a problem to the forum, but admins, you may want to place a resource cap to prevent the server from hogging all the resources, causing a kernel panic. I wish that there was a way to recover the logs from the kernel panic. However, that isn’t safe as writing to the disk after the computer has crashed is VERY risky!

#71

You aren’t exactly familiar with CD history, so here’s a couple of tidbits.

  1. Legend has it that Brandon (the webmaster) used to put the CD server in a bulletproof case shortly before Kickoff, so that damage to surrounding things was minimal when it exploded. Not sure how true the exploding part is, but we’ll go with that.

  2. There was at one time a resource cap. Due to the amount of traffic, Brandon at one point locked the number of users online to a certain maximum–spiders were banned temporarily, guests had lower priority, registered users who were over the max couldn’t get on for a while until someone else got off. Shortly afterward, the server got more space, so the limits were lifted.

And 3): The current solution is a bit of cloud-based server. Resources can be added in or removed to handle increased traffic (say, adding another couple of servers around Kickoff for the traffic, and letting them slide out in April or so). No issues other than the odd forum downtime, generally for an upgrade or to add in a server or three.

#72

How about setting up a system where every post/edit MUST be reviewed by a moderator before being put up. It is easy to get past a filter, but people are much harder to trick.

#73

The mods do a great job cleaning up the spam that gets through. They do not have the time to approve everything. When you report spam or questionable content, they get notified by email.

There needs to be a balance. I think it works here.

#74

I think the webmasters have a life! Having to moderate every post will mean ChiefDelphi will need an army of moderators to approve the posts. Also, The posts won’t show up until they are approved, which can be after something like possibly 24 hours! Instead, the first few posts should be moderated

#75

How about (when somebody registers), putting up a picture of someone like Dean Kamen or Woodie Flowers and asking for the person’s last name. It’s something that everyone involved in our community should know, but not so much outside. If they honestly don’t know, make them register with admin approval.

#76

Probably not the best thing to hear, but I don’t know how these people look like even though I am on an FRC team. :yikes:

#77

Then your registration would need an admin approval.

#78

I’m pretty sure many people do not know how these people look. Most newbies won’t know it either. maybe, we could have a captcha gif instead, with moving characters to make OCRs have a harder time!