Re: How big a problem is spam
You are not logged in. Please login or register.
Spam is not a problem IF you can control your user list. That is, if you reasonable assurance that all users are legit, you don't need to worry about spam fighting techniques. So the question really boils down to keeping spammers from signing up.
My experience managing lots of people's blogs tells me that nothing comes close to Akismet for efficacy. With boards, it's a bit different. Assuming you don't allow guest comments, the game is really about controlling user registrations. With that done, you don't need to worry too much about comment spam.
If Akismet or a similar system could deny bogus signups in real time via the "hive mind," I think we'd have a good solution on our hands.
It would be good to be able to delete a user account with a single click on a button next to a message they have posted. Currently it involves a lot of clicks and screens and a wildcard search to get a list of users in admin. Which gets a bit tedious if you have several to do at once.
How about a check to delete box next to each user in the list and a confirm button at the bottom?
Also I think someone else suggested the option to have unconfirmed accounts deleted after a certain number of days?
Absolutely brilliant - just what I was looking for. Note I couldn't find this at punres anywhere. For future searchers, the direct link is:
Fake registrations and spam are such a problem in everything these days that I think these tools are a basic which should be included in the main board software. To be honest I'm not too keen on or confident about adding mods. I find it boring. I just want something that I can install and run without any tweaking :-)
rgsuk - I agree about mods - and unfortunately there seem to be a whole lot more mods than plugins for punbb for some reason. However, User Management IS a plugin, not a mod, so I can't fault the devs for not including it in core.
Smartys - Interesting distinction -- very different from many/most systems, where a plugin can modify any aspect of the system (for example WordPress plugins can affect virtually any aspect of WP, not just the back-end). Out of curiosity, why is this different with punBB?
They were added as "Administration interface plugins", in the 1.2 release, to help admins with "simple" tasks, and has since then been 'shortened' to plugins when ppl talk about it
The extensions is a further step in that direction, with possibility to manipulate data all over the place ... and it will really "extend" the forum, so the name seems logical to me ...
"Administration interface plugins"
Yes I see why you shortened that :-D
I'm not sure I'm can tell my mods from my plugins. I just want to be able to install the board and touch it as little as possible after that. I guess I'm not exactly a power user ;-)
On page 1 AlanCollier suggested 'people power' vs spam sentiment by relying on an "active community of moderators". Its funny because about 2 post before his I was wondering the same thing.
I mean, some forms of automation may be helpful, but the challenge seems to be how to not be bogging down the systems and administrators with complex automated utils that then require more sophisticated administrative management, and don't totally serve becasue they cannot 'think' on their own well enough). (sigh - @#$%@ BOTS!)
Maybe something really simple like using an "automated system for creating and managing an internal-network of would-be moderators out of existing members" would actually provide a smart and low overhead defense?
But we already have moderator group availability you say...
Maybe it would work something like this (and maybe this would be different?):
- Create criteria for allowing members who fulfill the criteria to join 'anti-spam moderator group'.
- Obviously it would have to be a system based in/on trust... (but consider, trust is a human characteristic bot-spam will probably never have or be able to emulate/automate (in our lifetimes anyway?), thus a system based in trust might well be a most powerful and simple tool (if we could figure out the best way to be leveraging that)...
- Those qulaifying to elect to volunteer to join this anti-spam moderator group would only have the responsibility and priviledges for deleting spam-posts, and possibly even spam-members (however you choose to set the criteria for that). They would not have any other moderator priviledges. (Some sort of checks and balances system could be discussed and sussed out for what might be the best ways for managing the members of this group, of course.)
- The idea here would be to enable the ability for the automated evolvement and maintenance of a virtual network of human thinking anti-spam tools (every fool is a tool, eh?). (As opposed to the automated evolvement and maintenance of all sorts of blacklists and such (for ips, urls, words, ad nauseum) which isn't really so automated, because the minute you build something to think a certain way, it starts becoming outdated (but a basically intelligent human being seem sto always recognize what spam is or isn't ;^)...
It seems this is the sort of "feature-set" for a forum that would require very little:
- very basic modification to the forum itself,
- without requiring continual updates and maintenance of a/some utility(s) (because its not an anti-spam, or group of, anti-spam utilities, it would be a simple feature),
- as membership grows, the shear number of people-force would prpbably outweigh the ability of the spam-bot force out there to be potentially overwhelming or taking down your system(?),
- and then who knows what other way-more simple protection tools might evolve to for enabling better leverage of/for the human driven anti-spam-moderator group?
- oh, and it may allow for additional usage of other existing anti-spam tools with less maintenance overhead?
Does this really sound all that -stupid- or -unviable-?
Am I overlooking something obvious that renders this idea 'a non-idea'? (Maybe this was thought out in the stone-henge days of anit-spam dev, and they tossed it because...?
...added later-however, I know they're thinking about this art punres, because I just went there and read this http://www.punres.org/viewtopic.php?id=1848 ;^P
What do you think?
And if its a good idea, then maybe this feature capability should be built into 1.3, if only to facilitate the potential viability to prove itself out or not - lets see who uses it and how well it adapts to fighting spam?
PS: if you think I got my head up 'it', please be gentle pulling out, okay?
Maybe I'm not understanding correctly, but how is what you're describing different from Akismet?
Hmmm.. I have to admit I am unfamiliar with akismet and had not really read up on it at all until you just brought it up again (my bad - I am only loosely keeping up with newer anti-spam tech at best... you just get a bit jaded...).
Its a bit late right now, but I believe I witnessed a 'fond' post from Richard at punres with regard to Per Soderlind's akismet mod he is working on. Akismet integration with a mod like this sounds pretty cool, to be sure, but people are still having some issues with it... need to get a leash on that automation yet (but it sure sounds promising)...
Regarding your actual question, after (getting bonked on the head and then) reading some about it I would have to admit the relevance (if indeed it is functioning with that virtual 'hive mind'-edness as you sort of put it); albeit then... the differences would (obviously) be (from what I could briely gather):
1) - akismet is (still) automated, and so it still requires regular human oversight anyway (but admittedly not a lot, as you and others interestingly keep pointing out),
2) - it relies on yet another spam-server of sorts basically, really,
3) - as well, what you point out in your comments (in post 53 above) regarding "controlling user registrations",
...well it seems neither of these methods (live human or akismet) can really handle that right now...
But I have to thank you for bringing my attention to this (again)... I bet between the two, akismet might have the upperhand to potentially be dealing with bogus sign ups, and that would really be something (but at what point do you arrive at being able to rely on that? ;^)
AND it is interesting, the way in which your, and others who are using akismet, comments I read are expressing their 'unique' pleasure with the product.
After looking at it I still think there may be a value (viability worth testing) in a mod for punbb (or any forum) to allow a real-human collaborative system (internally) which does not rely on any automated system or other servers or more server overhead. I wouldn't go so far as to argue one over the other mind you... I am only trying to look at the real similarities and difference, and saying I wonder if it may still be an unknown worth validating (just as askimet seems to be)?
Well, I would not say that Akismet is automated - it depends entirely on users identifying comments as spam. It's automated in the sense that it does not require your intervention for each comment, but not automated in that its "intelligence" is purely based on what groups of humans designate as spam.
However, Akismet does require reliance on an external server / service. But I don't have a problem with that - it's been great, and with a very low failure rate.
That image verification mod is an old rls.
It is imposible for me to use it on v1.2.14
If Dzuma is talking of Antibot then Smartys is correct, I tested myself on a board.