> Potentially bad ns for domain in body, potentially bad ns for domain in title, potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Potentially bad keyword in body - Position 0-14: thepennsyl.com Potentially bad keyword in title - Position 0-14: thepennsyl.com Potentially bad keyword in username - Position 0-14: thepennsyl.com Potentially bad ns for domain in body - thepennsyl.com NS suspicious ns2.hawkhost.com.,ns1.hawkhost.com. Potentially bad ns for domain in title - thepennsyl.com NS suspicious ns2.hawkhost.com.,ns1.hawkhost…
Scan's used for scanning a linked post, or posts, to see if it's spam but not caught yet. Bisect's for checking if it's on the watch/blacklists, and if so, what regex it's caught under. Test is used to see if it's caught by any spam check, and if so, which.
@JeffSchaller 10 Rep guessed that it might be a non-serious troll question based on the question itself, which the user's subsequent behavior (as far as I can tell without being able to see the deleted post) seemingly confirmed
Well, I'm not able to keep a finger on the pulse of every SE post, so I'm going to leave my fp feedback unless there a consensus that I'm wrong -- which I will do my best to acknowledge, if so!
In my mind that question is a little offensive and seems troll-like. But fp feedback makes sense as the question isn't directly offensive. I'll keep my feedback as well.
@RyanM Interesting. I understand the desire to focus on spam, but it's a bit of a pain to give an NAA feedback on MS and then flag on the actual site (or flag on the actual site and then give SD an NAA feedback). It'd be nice to be able to do it all at once...
@SmokeDetector tpu- undisclosed affiliation. The answer is written in first person, but it's not at all clear which papers are theirs, and the answer seems promotional. Also, it's their second offense (other answer had affiliation added in by another user). Custom flagged.
@Spevacus Thanks for taking the effort (and it looks to be quite a lot) at making the PR to address the whitespace issue in watches. Getting that in, in some form, would be very nice.
@cigien The code doesn't feel too impressive (though I will say this is my first venture with python beyond "Hello world!") But the real venture was setting up and running Smokey. When I got it to run, I was surprised how easy it was, despite it taking... A little longer than I care to admit ;) Thanks :D
Also the code requirements were a new thing to me as well. Line character length requirements in particular. Let me have my god lines, dangit! ;)
@MatthewDaniels Hi! This room is for Charcoal, which focuses on detecting and eliminating spam and rude/abusive posts on all SE sites. If you need an MSE mod, you can flag any post on MSE as "in need of moderator intervention" and explain your issue. Or the Tavern, as Spevacus suggests.
err, actually, you probably can't do either of those things.
I could probably get the hang of it. I'm just not used to the restrictions. I like the freedom to have insanely long lines so that future me hates past me. ... You know, writing that out now, I'm thinking that maybe that is a bad practice...
I'm actually used to auto-formatters, so I, or any particular project, can have whatever column limits it wants, and it doesn't matter to anyone else. Of course, this may not be straightforward, or even possible, to do in languages where whitespace has semantics.
@RyanM Yeah, I was about to go through the domain on MS which you linked to. Done now, thanks.
Hmm, for the last couple of weeks I've been reading the transcript in reverse, and on the whole it's been an improvement, but it does have some downsides, like with this example :/
@Nick I also considered watching Mike Hunt (5 TP/0 FP in MS), but it turns out there are a surprising number of parents with the last name Hunt who made poor naming choices for their children.
@cigien Yep, I checked 'em all. Only one has any posts, and based on its one post having been acted upon by multiple CMs, I suspect that account was owned by a troll who happened to post one useful thing.
@Spevacus At this point, only SO (3) and Mathematics (2) have a queue depth which isn't 1. On sites other than those, a post being alive for 3 or 4 minutes and not being detected tends to be because the portions which would have been detected were added during a grace period edit that happened after SD originally scanned the post.
and because...well, I have no idea what the code search option on Stack Exchange is supposed to do, because its behavior is wildly inconsistent and unpredictable
Yeah for articles or any sort of prose text (including code comments) it's great
The trouble is when you've got a user that you know is plagiarizing a large portion of their answers and you're trying to check them all
!!/scan-force android.stackexchange.com/a/142850 "Promotes known spammed site with horrible, useless advice. As noted in the comments, it omits how to factory reset a phone in a boot loop, and the gist of the answer is that you should just delete all your data (and 'don't worry about your data'!) because you can just use recovery software."
!!/scan softwarerecs.stackexchange.com/a/33748 "Suspected spam: Stellar are known spammers, and this user's profile links datarecoverymethods.blogspot.com, the author of which has a bunch of weird spammy-looking "blogs" each with a single post consisting of text that appears to be copied from somewhere, with the "blogs" having been created with minutes of each other based on the "welcome" post on each."
