@cigien That pattern looks like it's already caught by Bad keyword in answer and Bad keyword in body Append -force to the command word(s) if you really want to add the pattern you provided.
@MartinSleziak it's a good discussion to have, we don't currently have any systematic approach to following up on deletion. I created a ticket last year but there is not yet any real consensus as to what a proper solution should look like
over the past month or two I've kept track of the posts I flagged as TP and generally followed up here with "still up" and/or escalated to the Tavern where generally some additional flaggers are available and happy to help out
@cigien That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body Append -force to the command word(s) if you really want to add the pattern you provided.
we sometimes see posts which were reported and TP:ed months ago turn up again when the post is rescanned for other reasons, which ideally we would like to avoid going forward if there is a clear consensus that a post is spam
my proposal at the time whs to create a separate dashboard on metasmoke for these posts, but perhaps it would be enough, or even a better solution, if the review queue would prioritize posts which are still up, and encourage reviewers to also flag them
nothing much is likely to happen until we figure out how to attract Rails developers to work on metasmoke (or teach ourselves enough to start limping ahead again)
@SmokeDetector naa- User has at least 4 answers posted yesterday linking to the same blog article. Article was posted a few months ago, and no other hits for the blog on SE, but user appears to have other reasonable contributions, and I can't find any affiliation, so giving this the benefit of doubt for now. Watching the domain just in case.
@Ethan No. The post is very spammy by itself, the username matches the website name; the account will likely be destroyed along with the post. If not, they'll be handled in due time. Just red-flag it.
@Ethan @Ollie Some moderators won't mind and not even close to all accounts are destroyed. It differs between moderators and sites how this would be handled.
@Mast That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body Append -force to the command word(s) if you really want to add the pattern you provided.
@Ollie Accounts are only destroyed by manual action from moderators/CMs. If a post is red-flagged out, there's no active notification about the post or account made to moderators, so no action is likely to be taken on the account. On SO, there are multiple moderators who look at SD reports and who tend to destroy spammer accounts (largely because userscripts make it really easy, even easier than just suspending). On other sites, it's substantially less likely.
Red-flagging nuking a post will usually result in the user being blocked from posting additional content for 7 or 14 days, due to the IP address being fed into SpamRam, which tends to be sufficient to prevent most spammers from posting additional spam from that account. A notable number of spammers for whom that's not sufficient tend to not care about the destruction of the account, as they will just create a new one when they need it.
OTOH, it does help to prevent some spammers, and more commonly trolls, for the account to be destroyed. I probably wouldn't raise a custom mod flag for a single spam post, but if there are a few by the same account, then it's probably a good idea to go ahead and raise a custom mod flag to specifically let the mods know.
Yet another side of the decision is that it's actually substantially easier for us to detect spam when the spammer uses the same account (i.e. that's what blacklisting the user is for), so there are advantages to not deleting/destroying the account.
As for spammy profiles without any posts, the general SE policy is "meh". Some moderators will care and some will not. It will also depend on the site.
On SO, there are at least 10s of thousands of such accounts, if not substantially more than a million. SO moderators generally don't want to see custom mod flags regarding spam profiles which have never posted a question or answer, largely because if we wanted to track such things down we could do it with a lot less effort than having someone flag each one individually. Such a flag is good if there's something particularly egregious (e.g. child porn, human trafficking; yep, those happen).
room topic changed to Charcoal HQ: Where smoke is detected, diamonds are made, and we break things by developing on production. This room is for Charcoal, a volunteer organization focused on detecting and eliminating spam and rude/abusive posts on all SE sites. 220k true positives and counting. For more information, see charcoal-se.org/about. Handy links: metasmoke.erwaysoftware.com; github.com/Charcoal-SE [dev-on-prod] [flag-based-fraternity] [programmatic-crowbars]
@Ethan I assume you mean 254 Ki flags (i.e. kibiflags), as I'm seeing 260333 flags on the MS homepage.
@Ethan That's the number of autoflags successfully cast. There's a 13-grand-odd deficit because Smokey sometimes fails to cast flags, for various reasons (e.g. because you don't have an account on the site where the spam was posted or because the post is already deleted). See metasmoke.erwaysoftware.com/flagging/logs for more info.
I'm no expert though, so please correct me if I've gotten anything wrong.
Well, that's strange. The flagging page shows 254959 flags cast and 13713 failures, but the homepage shows "80618 posts have been autoflagged, totalling 260333 flags". I don't really expect those numbers to be equal, as the numbers on the flagging page appear to be all flags (i.e. including manual flags). However, if they are showing autoflags vs autoflags + manual flags, then I'd actually expect the numbers to be different by substantially more than 6k.
I'd note, however, that whatever those numbers are reported as, they are inaccurate, as we lost a few/several months worth of flagging data from the MS database.
I have a handful of commands to test out. Is it fine if I use Charcoal Test for that? It says it is for unofficial Charcoal clients, but I'd like to avoid cluttering up here
> Bad keyword in body, bad keyword in title, bad keyword in username, blacklisted username, potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Bad keyword in body - Position 0-23: Coinbase support number Bad keyword in title - Position 0-23: Coinbase support number Bad keyword in username - Position 0-23: Coinbase support number Blacklisted username - Position 0-23: Coinbase support number Potentially bad keyword in body - Position 0-23: Coinbase support number
@cocomac Your question is unclear. Are you wanting to perform commands which are run on your own instance? Are you wanting to test some commands which are run on the main currently active SD instance? If it is that you're wanting to run your own SD instance and run commands against that instance, then the best thing to do is create your own room for testing your instance. Other than a few commands, any command will be executed by all instances listening in a room.
@Makyen I’d like to run a few commands against the main SD instance. I’m working on making it scan a data dump but it is giving me some odd reasons so I’d like to compare what I get locally to what it actually is
That effectively means that there shouldn't be more than one active SD instance listening for commands in the same room. Primarily the only commands which allow specifying which instance to run on are those which are used to control moving an instance to the standby or off states.
That makes sense, but if I have some !!/test commands to run that will clutter that chat, is here fine or should I head somewhere else? I don't intent to run any commands that will switch instances or run any other instances myself
@cocomac That sounds like it's more appropriate for one of the Charcoal Test rooms, rather than here, as it would be less disruptive in one of those than it would be here.
@cocomac If you're concerned about disrupting other users' use of Charcoal Test, then you might consider Charcoal Test SO, which has relatively little use other than SD posting "debug" messages.
