Discussion in 'Web Feedback' started by ScottulusMaximus, Sep 9, 2015.
4 pages today, 2 hours and counting from first post
As has been said. Report what you can, or at least 1 or 2 for each spambot. That way the moderators can get it all.
I've reported them all, that doesn't change the fact that there are no mods or spam killers active for 2 hours and counting... That is the issue
Wow that was the most I've seen yet.
Well that isn't really FTB's fault.
As has been said before, people aren't chosen based on their timezone, they are chosen based on their activity on the forum, time they can commit, etc.
Also, it isn't their job they can't be expected to be on 24/7, but also this isn't something they should ever have to deal with, as the spam filter for the forum should stop the spam, not the mods having to delete it.
Please stop defending them because you're starting to annoy me again... Yes you worked for FTB, yay for you, but right now this has nothing to do with you. I am speaking to the CURRENT mods and staff and saying something needs to be done.
And I will keep saying something needs to be done until the excuses stop and the problem is fixed.
Sure, go for it.
Trust me, we hear everyone on this. As of right now we're talking daily about this and what's going to be done to fix it, but our hands are tied. As soon as something changes I'll drop by here and let everyone know. It's not like we don't care about when it gets fixed, it's that we have no control over that right now.
(FULL DISCLOSURE: Despite the forum tag I have basically been retired since May and have zero voice in the governance or management of the forums here. Please note that I am sharing a personal perspective, not dictating policy. Also, this is the first time I've actually browsed the forum for more than two minutes at a time since May, so my observations may involve superficial data. The blowhardiness of the post of course is the usual ration you'll get whenever I say something. =))
Spam Preventor's Maxim: A given spamming operation will figure out a way around almost any given human verification step added within 3-6 months of its implementation, if not sooner. This includes human verification questions tuned towards specialist questions. The problem here isn't that the admins haven't figured out the perfect cocktail of flaming hoops to make new registrations jump through; the problem is that the solutions all have shelf lives and the site doesn't have an admin who is focused primarily on the forum versus being involved in eleven other FTB core projects that require higher-level attention. Things slip through the cracks in volunteer organizations because you can't just hand out all the keys to the city and every keyholder is spread among general task lists.
On the human verification question suggestion: Any question sufficiently simple enough for this site's target audience to reply to correctly can be answered with a google search. Anything obscure enough to actually keep the bots fooled would keep far more legitimate users fooled as well and would be counterproductive to the purpose of having a forum. The only way to make this work, in the long term, is to have somebody refresh the human verification questions every time the cipher has been cracked. This presents a problem in this community, though, in that there's only a very finite amount of information one can expect a new member to know and answer with correctly and without research; once you've cycled through that information, you've run out of questions.
(And yes, I can confirm that they will get the answers on these. One of my other hats has been as the primary administrator of a rather large RPG site, where I once attempted to curtail bot activity by putting in human verification questions that required actual research in the site's lore compendium; if they're going to look up and catalog in their answer sheet who the fifth Empress of a second dynasty that reigned 700 years ago in a fantasy world was, they will definitely find Minecraft trivia a much simpler nut to crack. Once a solution has been found it gets propagated among the bot script-writers, and they're right back at it.)
When I used to do the rounds on this, I was quite literally eyeballs-on-site from about 10pm to around 4am if I wanted to whack all the moles before anybody really noticed. That's not a sustainable pace to expect unpaid volunteers to do in the long haul, which is why that often slipped and why other background solutions were supplied. An interesting note to make as I write this: even with the ones getting through, there are 90 new threads that the system has trapped tonight in queue. So the system is filtering out a lot of crap, it just needs to do better.
One thing to note is that the spam currently coming through is of a particularly challenging type to curtail. Site SEO-mongering spam needs to include hyperlinks, so a simple bar on those posts is to limit who can hyperlink (another example of the adaptations the bot-writers make: I've noticed that in this war that a number of the bots now have been programmed to the extent that they make an initial innocuous post to get past the queue filter and then edit in the preferred content; there probably should be a restriction set on edit permissions for new users for this reason, as obnoxious as it may be). These ones, however, are a different breed of spam: the search result blanketer. This is distinguished from the Site SEO spammer, whose goal is to increase linkbacks and hook an occasional lackadaisical click-through. Search result blanketers just want a particular term set and/or set of real-life references (addresses, phone numbers) embedded throughout the web in search results.
This set is particularly difficult because they're in Korean and Chinese. Another layer of the spam filtration process besides link restriction here is (or at least was) a list of keywords that were common to much of the spam being posted. It's not that easy to come up with a keyword list for these languages for a lot of reasons (and I could go into this at length but I don't think anybody really cares about the quirks of East Asian ideographic linguistics) and so the filtration is difficult to work on that dimension. Even the old Indian "black magic love" spam was easier to target than these are.
Anyway, that's a lot of trivia. What it boils down to is that a rounded solution involves several steps and that all of these require some amount of attention and manpower applied to them:
Registration filtration through human verification and other steps (for example, I've personally found it useful elsewhere to simply ban a number of foreign+free e-mail hosts or set restrictions on that basis from registration).
Access curtailment for suspect/new accounts (longer between-new-thread post timers, restrictions on editing, other tricks that stymie the bot's ability to succeed).
Content filtration via URL and keyword blacklists for content from new accounts (requiring a minimal post count before permitting unqualified URL-inclusive posts, barring very common terms that aren't applicable to the forum and pretty much only get used commercially).
Human intervention to catch the ones that slip through the cracks and to further refine the previous three processes.
Please keep in mind that only #4 is the actual domain of moderators, and only to the extent of those currently with the tools doing what they can; the others are admin domain concerns, including adding more people for #4. When you rag on moderators to deal with the first three things, you will get absolutely nowhere, because they can't do anything about it besides pass along feedback just like you already can.
Beefing up #4 can be useful but can also be problematic. You have to strike a balance on who you give tools to in an environment like this. I'm totally a fan of the suggestion of lifting the report cooldown from basically all user categories except the spam suspect category, as long as it's also made clear that anybody who goes hogwild in an abusive manner is basically going to end up banned for it (I think during my tenure I banned two people for this kind of behavior, so yes, it does happen). I'm hesitant about expanding the spam cleaner group too far because you are giving the power of banning new accounts to new spam cleaners; somebody in that group going rogue can seriously damage the forum's reputation, so you have to take care on handing out that tool. 10x that concern for additional moderators (and with moderators you also have to strike a balance of day-to-day workload; if you select a moderator cadre to fill in at the level of an every-six-month spike, you're going to have a lot of bored and useless feeling moderators who just wander off and won't handle the spike anyway; you need to keep a moderation team lean enough that there's something they can always be working on and understand that occasional effluent piles will temporarily hit the forum).
There. I've iterated entirely too much and it's probably a giant wreck with little coherence. There is obviously a problem, there is more than can be done, and suggestions and feedback on it being a continuing problem are valuable; but please make sure the tree you're barking up is the correct one, though, otherwise it's just going to become a big circle of frustration in an already frustrating situation (and it is very, very frustrating for the moderators on this too, I will iterate again from my own experience).
(On that note, shouldn't this thread get located to Web Feedback, perhaps? =))
I read that first paragraph and blanked out...
Vauthil is right about where this should be located, and now that people have seen this and commented a bit, I'll toss it over there with an active redirect in general FTB for a couple days.
I can assure you I am not an "ex-staff" member who would actually care whether FTB's reputation was great or terrible (in terms of spam), I only say they are doing what they can because I know they are.
It had been said multiple times before as to reasons why things can/cannot be done, and why spam is the way it is right now.
Vauthill did a great job of explaining things, albeit quite a long way of putting, but still true.
As for the
It's not just the current FTB staff that have the power to help remove spam, and heck report it.
There's a fair amount of people who are in direct contact with the FTB staff, and moderator team, on a daily basis, which aren't classed as mods/staff, and they are all trying to help. And trying to remove spam is a top priority on the forums. You think you're fed up with it? So are they. The mods don't want to deal with 50+ spam bots when they wake up, not does anyone on the forums. And when you go about complaining, and then not seeing what everyone else had said already (and repeating the same things/causing people to repeat themselves), it is just annoying.
You don't need to remind anyone. Everyone sees it. People report the spam when it appears.
If people stopped reporting the spam, and mods didn't delete it. Then yes, moan away.
But these "excuses" (read: reasons) are why the spam is here. There are many ways to stop it, but as vauthil said, it's all temporary, as it will be sussed out eventually.
Paraphrasing what Vauthil said above, any fix applied might last about 6 months if we're lucky.
Yes and no. You can make it unprofitable for them, however.
For example, flood control on new users, requiring a ten minute to one hour cooldown before next post for the first ten or so posts. The whole reason for this spam is for the linkbacks. If they can only get one linkback per account, this site is going to be less attractive to attack when numbers count. Combine with restrictions on editing for your first x number of posts and you've got a pretty solid way of putting a cramp in this sort of stuff.
Alternately, auto-ban any post which uses oriental characters. This forum is largely English speaking, although many of our users are multilingual. This puts a kibosh on the most problematic trouble-spammers.
That works for this wave. There was another wave of spam some time ago that was all Roman characters, and it's only a matter of time until it comes back.
Everything he said... A first poster has absolutely no need to create 20 threads in half an hour, in fact no one does.
Create a new group of "demi-mods" whose only allowed function is to remove spam. I understand that the power to delete threads is a big one so use the following rules:
Make it very clear they are NOT moderators, if they ever attempt to behave like moderators(warning people etc) they get an infraction, 3 infractions and they're perma-banned from the forum.
If they ever delete a post that is not spam it's an instant perma-ban.
Not sure what platform this runs on but if it's possible instead of deleting the posts/threads they're just hidden until a proper mod comes on and can kill them.
Open up voting like you did the current mods and make sure that there are a couple of them active IN DIFFERENT TIME ZONES.
I'm already one of your demi-mods. Those exist. All I can do is clean and block, and ONLY on new posters.
Honestly, I'm seeing way less spam than I did several months ago. It used to be I killed 30+ a day. Now I'm lucky if I find 2 (during North American work hours) and trust me, I check at LEAST once an hour, and usually way more.
Preventing new posters from using links helped a lot. It needs to be improved so they can't edit in links too, which is simple enough.
I like the idea of them not being able to post too many times per hour on their first day.
That's the problem
Yep. One of the key solutions is really simply to find another member or two who is happy being a garbage collector.
For the record: I love being a garbage collector. Most of the time its blatantly simple to determine spam, but every once in a blue moon I'll escalate it and I'll get a note from padfoote telling me if I was dumb or not.
Exactly, this is all that's needed but any criticism(constructive or not) of this forum and you get ganged up on by the loyal cadre bellowing how dare you insult the forum and it's not the mods job etc etc all the while spouting excuses.
huh? by whom?
Granted I skimmed some posts here (Vaudril's was even more tl/dr than usual) but i thought the gist was all constructive.
I certainly don't feel like anyone's insulting me. I do my part, and I'm reasonably sure the rest of the team appreciates it