| 1 person marked this as a favorite. |
I'm no computer programmer, so I have no idea if its as easy as I think it would be, but I wanted to suggest something regarding the recent spate of spammers as its presumably only going to get worse.
I wondered if you could easily add a flag category (ie "post is spam") and set some kind of conservative threshold (maybe ten?) such that, if any post is flagged as spam by ten posters it is automatically deleted. I figured that might cut down on staff workload, help when the spam attack is 2:00 AM Seattle time and not be easily abusable nor give too many false positives.
Presumably, you can still see deleted posts, so an incorrectly purged post could be restored easily enough. If there was any concern about misuse, perhaps one could restrict that flag option only to "established posters" with over a hundred posts, who signed up over a year ago, or something similar.
|
| 5 people marked this as a favorite. |
A new flag category for "spam" is a good idea, but I don't the an auto-delete would be very wise. Once people realize there's a threshold at which posts would get auto-deleted, they would flag anything they didn't like as "spam" in hopes of it disappearing before mods intervene.
For the flagging system to work correctly, there basically has to be a human-in-the-loop.
-Skeld
I considered the abuse thing but figured that the threshold would guard against that, plus it would be easily reversible if that did become a problem (or the threshold could be lifted). Also, I assumed that the mods can see deleted posts - so they'd still log on six hours later, see a ten-flagged deleted post that shouldn't have been and then be able to restore it.
My thought is that the human is still "in the loop" but only has to act if ten paizonians act together as jerks.
since they like to put the same phone number in thread subject, can you block any message from being posted that has that number?
That's what they're doing at the moment, but it isn't always the same guy.
| 2 people marked this as a favorite. |
We just need someone to set up an autodialler to call the number on his phone once every 15 seconds
To prevent folks from abusing the auto removal against posters they simply disagree with, the auto-remove could be limited to first-post-in-thread and then also delete the thread.
To prevent folks from abusing the auto removal against posters they simply disagree with, the auto-remove could be limited to first-post-in-thread and then also delete the thread.
Whilst I agree the abuse is a risk, I dont think it's that great a deal is it? I mean it would be visible to Paizo and easy enough to reverse (no harder than what they have to do now anyway).
I'm presuming they can still see deleted posts and that "delete" actually means "dont let non-staff see it".
To prevent folks from abusing the auto removal against posters they simply disagree with, the auto-remove could be limited to first-post-in-thread and then also delete the thread.Whilst I agree the abuse is a risk, I dont think it's that great a deal is it? I mean it would be visible to Paizo and easy enough to reverse (no harder than what they have to do now anyway).
I'm presuming they can still see deleted posts and that "delete" actually means "dont let non-staff see it".
I don't particularly see it as a big risk either, but if I understand the way the site is constructed correctly (and admittedly, that is a BIG if) then adding a must=post#1 line into the script wouldn't be very difficult and would prevent almost all conceivable possibilities of abuse. At the same time it would help to allay any fears that posters might have about being ganged up on.
|
Not really. You just open the floor up to full thread deletion for controversial topics. And instead of having to flag each post, someone who is vehemently against topic XYZ would only need to get enough folks to flag the first post and the entire thread disappears.
If you really want to reduce the amount of spam in a forum, one really needs to incorporate captcha (be it traditional, math based, etc.). However, this is something I believe Gary has said they would avoid implementing.
Not really. You just open the floor up to full thread deletion for controversial topics. And instead of having to flag each post, someone who is vehemently against topic XYZ would only need to get enough folks to flag the first post and the entire thread disappears.
If you really want to reduce the amount of spam in a forum, one really needs to incorporate captcha (be it traditional, math based, etc.). However, this is something I believe Gary has said they would avoid implementing.
It could be limited to threads which only have 1 post.
Am I wrong that if a post is deleted, the staff can still see it? I think this "what if a bunch of paizonians all decide to abuse the process in a way that's traceable and really obvious?" Concern is a bit overblown, to be frank.
|
|
| 1 person marked this as a favorite. |
Bottom line really is this. What leads to less work on their end?
I don't think allowing the deletion (even temporary deletion) of threads will reduce their workload in regards to the spam accounts in the slightest. They will still need to review those posts which were removed by the flags, deactivate the spammer's account and delete the offending posts. Kinda like what they do now.
Except if the system is abused even slightly, they then would also need to restore the deleted threads and potentially send out warning emails to those who were responsible for the abuse of the system. Sure it may happen only once in a while, but over time that adds up to hours that could have been spent on other efforts.
The entire goal for coding a solution to a problem is to decrease the workload from what is currently required to deal with the situation and I don't see flag-->thread deletion by visitors accomplishing that goal.
EDIT: and why put a mechanism in place that could lead to an automated means of hobbling new posts on the forums? If a mechanism has the potential to be exploited, some ass out there will exploit it for no purpose other than they can.
You may be right. If it doesn't help, it doesn't help obviously.
I'd be surprised if zealous anti-X posters started flagging any post about X as spam. Partly since I think people are basically decent but also because its a really obvious and trackable abuse. As you say though, if its not any easier the suggestion is a waste of pixels.
|
|
| 2 people marked this as a favorite. |
I agree that the vast majority of folks would not flag a thread with which they disagree as spam in hopes of having it deleted, even temporarily. All it takes though would be one person to throw together a script to concurrently log in to a series of bogus accounts and flag all threads with a single post as spam and in the matter of minutes you would have a freaking mess to deal with on the back end.
Wow. Yeah. I don't have the mind for this! I never even thought about that. :o
| 1 person marked this as a favorite. |
Thank you for the suggestion. We are brainstorming on some ways that we can make moderation a bit better and how we can deal with spammers. I'm not sure that we'll ever have a mechanism that automatically deletes posts without a human reviewing it first, but we'll see :)
| 6 people marked this as a favorite. |
Gamers gonna game the system.
is there any validation element on the account registration form, I cannot remember
Gamers gonna game the system.
So, spammers be min-maxing?
Wouldn't a CAPTCHA when registering an account be a viable solution? It's only a one time "annoyance" for actual people instead of e.g. putting in a CAPTCHA when posting a new thread.
No idea how it would fit in with the rest of the coding.
IP blocking then? If they're not bots then they shouldn't be able to use zombie computers. They might still spoof IP addresses, of course (go through proxies etc.).
They are at it again, in the off-topic.
What about limiting all users to three new threads per hour? or day even?
|
Looks like the Loveking* has taken over the OT section of the forums.
I have to say though, that section of the forums haven't looked so good in a long time so maybe it's an improvement.
Here's a potential fix
All new joiners must fill out a small, non-dropped down character sheet when they join. Something to fill out, sort of like a super-capcha in the form of a character sheet?
Anyway, I may need to call this guy for some advice on black magic.
*sorry, the name made me think of the cult song - lemons into lemonade if you will?
Am I wrong that if a post is deleted, the staff can still see it? I think this "what if a bunch of paizonians all decide to abuse the process in a way that's traceable and really obvious?" Concern is a bit overblown, to be frank.
And what happens when the spammers find out and start marking every post they can reach as spam?
"Hey, where'd the forum go? Now all that's left is the Poodle thread and a bunch of posts from KC."
| 4 people marked this as a favorite. |
Can I recommend that we hunt them down, chase them through the streets with automatic weapons, corner them, cut their fingers off and make them eat them, dip their bleeding hands in salty lemon juice, throw them in a cage full of starving chipmunks, mince the remains and burn the mince with napalm, scatter the ashes on holy ground, and warn them not to do it again or we'll get nasty?
How about just auto-deleting anything with the text "babaji" in it? :P
But what if someone wants to discuss the alignment of babaji and whether or not it can fall?
Delete any spammer accounts, put newly registered people on test period (a dayor three), where system would track their activity and allow mod to delete them and everything they have posted at once?
Or use the search engine to find all posts by users and give mod a delete all posts button.
I know it's been mooted and declared unlikely before, but community based moderators selected from longstanding account holders (with clearly delineated roles) would probably help.
I can't imagine anything bad would happen if Joana (for example) was given full mod powers and told "just delete spam accounts, please".
Autokill every thread that contains a phone number in the title? Or at least this specific one?
Automatic check on any new thread started by a poster until they've hit a certain post count.
If a poster starts enough threads within the space of an hour that have similar words in the titles (or simply limit the total number of new threads they can start within an hour) , block them from starting new threads for an hour.
I think that ought to be non-obtrusive enough on the rights of new posters while being relatively easy to code.
|
rig something up that when the spam, SOMEHOW, it hacks into there computer and kills it, the router, and all other computers for good. Thus permanently stopping spam, because they are now out of money trying to buy new computers and routers, and there service provider is told to deny them acess to a new router. Boom.
Or just have me hopped on sugar running into peoples houses and beating them over the head with a sack of paizo merchandise
| 1 person marked this as a favorite. |
IP based Satellite weapons system?
| 1 person marked this as a favorite. |
I was just reading about a service to help get ride of enemies, where is that number...
| 3 people marked this as a favorite. |
The spam problem has already been solved:
xkcd 810: Constructive
We're all just waiting on an implementation...
| 2 people marked this as a favorite. |
rig something up that when the spam, SOMEHOW, it hacks into there computer and kills it, the router, and all other computers for good. Thus permanently stopping spam, because they are now out of money trying to buy new computers and routers, and there service provider is told to deny them acess to a new router. Boom.
Or just have me hopped on sugar running into peoples houses and beating them over the head with a sack of paizo merchandise
I'm not replying to you in particular, but stop.
If spam was a problem that could be eliminated by a few forum goons brainstorming, it would've been dealt with a decade ago. It's a complicated issue with serious technical and logistical difficulties.
I get this all day, every day from my customers whose anti-spam packages are blocking upwards of 90% of ALL e-mail trying to enter the building. Not 90% of unwanted mail. 90% of ALL. Because that's how bad the spam problem is.
It doesn't matter that each day the CEO's mailbox has 840 messages automatically blocked and zero false-positives. All I hear about is the two messages a month that make it through.
Paizo does an excellent job of dealing with this. I trust in them to adjust their manpower and technology as they see fit.
| 4 people marked this as a favorite. |
I can't imagine anything bad would happen if Joana (for example) was given full mod powers and told "just delete spam accounts, please".
I can't imagine anything bad would happen if Joana (for example) was given full mod powers and told "just delete spam accounts, please".Mwahahaha! The power...!!! :D
There's already double-post prevention that won't let you submit the exact same text twice on a page. Could you make it where you can't submit the same text twice in an hour, say, anywhere? Would interfere with Liz's "Now available" posts unless there's a manual override, and I don't know how long it would take the spammers to adapt.
There's already double-post prevention that won't let you submit the exact same text twice on a page. Could you make it where you can't submit the same text twice in an hour, say, anywhere? Would interfere with Liz's "Now available" posts unless there's a manual override, and I don't know how long it would take the spammers to adapt.
Roughly 30 seconds.
Seriously, pattern-matching heuristics have been in play against unwanted messages for well over a decade. They're decreasingly useful. Long gone are the days were "I'll just block anything with the word Viagra in it!" makes any sense. As soon as you do that, you're guaranteed to get mail about "Vi4gra".
You put in place a system that watches for 95% content repetition and I guarantee you'll see 94% content repetition moments later. These guys aren't stupid... this is a huge money-making deal for them and it's all about getting their crap in front of enough eyes to keep it profitable.
If this keeps up, the only real solution would be to block new users from posting until one of the website moderators can review their initial posts. One quick look should be enough to choose between "let in, no further review necessary" and "ban immediately and permanantly". We don't seem to have any cases of established accounts being taken over by spammers so far.
If this keeps up, the only real solution would be to block new users from posting until one of the website moderators can review their initial posts. One quick look should be enough to choose between "let in, no further review necessary" and "ban immediately and permanantly". We don't seem to have any cases of established accounts being taken over by spammers so far.
Presumably we'll then get some new posters making comments such as "I really enjoyed this product, looking forward to the next!" followed once they're approved by the usual guff.
Yes. Honestly, it IS more important to them that you see their spam than for you NOT to see it. They are ready, willing and able to adapt. It isn't a problem that's going to go away.
|
|
I know it's been mooted and declared unlikely before, but community based moderators selected from longstanding account holders (with clearly delineated roles) would probably help.
If they wanted to do something along those lines then the easiest option would probably be to give certain specific permissions to VCs, VLs, 5* GMs, Superstar finalists etc -- people who Paizo have had direct contact with and who've already demonstrated trustworthiness in other contexts. It would also have the advantage of providing a reasonable number of people in non-US timezones.
However, having managed various forums in the past, superusers can cause problems (unintentionally, indirectly, by accident or not), and even with limited permissions, when mini-mods attack it causes headaches.
| 2 people marked this as a favorite. |
Well, solution is actually simple. We need to start executing people that respond to spam providing the spammer overlords with money. After some time the spammers stop profiting because people learn not to respond to spam (or only those that knew that already survive) so there will be no motivation for most spammers to spam.
For those who think that this is an overkill a milder option is just taking away credit cards from those that respond to spammers.