@10Repsaysgetvaccinated That pattern looks like it's already caught by Potentially bad keyword in body and Potentially bad keyword in answer; in addition, the regex contains an unescaped "."; in most cases, it should be "\."; append -force if you really want to do that.
@10Repsaysgetvaccinated That pattern looks like it's already caught by Potentially bad keyword in body and Potentially bad keyword in answer; append -force if you really want to do that.
:56886078 That pattern looks like it's already caught by Potentially bad keyword in body and Potentially bad keyword in answer; append -force if you really want to do that.
@10Repsaysgetvaccinated yeah, I'm inclined to agree...it looks like they searched amazon.in for "jumper wires" and picked the cheapest one out of the first row... cc @Machavity
@CodyGray That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; append -force if you really want to do that.
my vague recollection is that the identity of a reason needs to be coordinated in complex ways between metasmoke and Smoke Detector so once you have a reason -> number mapping you basically can't tear that up any longer
there was one instance where I changed the name of a rule when I slightly changed its semantics and so now the old version hangs around forever in the reasons listing on metasmoke
@cigien @tripleee US phone numbers are commonly written by spammers (and everyone else in the US) both with and without the "1" country code. The numbers detection code doesn't automatically add/remove it. Adding one version as a detection will not, necessarily, detect some ways the spammer may write the other version.
I'd note, however, that including a "+1" is becoming much more common, because including it results in more phones auto-detecting that it's a phone number and allowing the user to dial by clicking the number.
@CodyGray I haven't seen a human hand write that, but I have seen it used by people in device assisted communication (i.e. email, texts, postings, etc.). My impression has been that it's progressively becoming more common.
> Potentially bad ns for domain in answer, potentially bad keyword in answer ---------- Potentially bad keyword in answer - Position 4-12: 21yunbox Potentially bad ns for domain in answer - 21yunbox.com NS suspicious f1g1ns2.dnspod.net.,f1g1ns1.dnspod.net.
And suddenly decided in the last 3 hours to post links to what is definitely spam? based on the other appearances of the link, that link looks fishy at the very least.
To me, the simplest explanation is that a high rep user decided to create multiple other accounts, and then post answers from all accounts sharing that link, and with a bit of voting to increase rep. Looks like Spam.
I had already marked all of the MS reports as TP; it was just a matter of whether we would be losing value by deleting what appeared to be a legitimate, relevant recommendation.
@Makyen Thanks, that's interesting. Is this strictly a US +1 thing? If so, would it make sense to change how watch-number transforms the number to account for this?
It just seems like you did precisely what you were reminding yourself not to do: adding a watch that would catch both the company name and the website URL.
Yeah, mine do, but that's a separate issue. I sometimes sleep with my eyelids just barely cracked open, apparently, so by the morning, my eyes will be dried out, and blinking will scratch them. :-(
They make goo that you can put into your eyes before going to sleep. My uncle uses that (he has a particularly bad case). But that sounds like a bad idea, to me.
My dad wears a mask over his eyes to sleep, which helps (and also blocks out light). But I can't manage to keep the mask on.
@cigien Mostly. I'm definitely much more familiar with the practices in the North American Numbering Plan (i.e. "1"). From what I've seen, businesses in some other areas/countries will sometimes leave off the country code. For people and businesses in the North American Numbering Plan, it is very common not to include the "1" when providing a phone number.
Yes, it would be reasonable, under many circumstances, for the numbers detection code to automatically account for it, but nobody's gotten around to writing that code.
@Makyen I was not aware of NANP, thanks for that :) Ok, makes sense. If there's more than a few of these cases, then automating it might make sense. I think I'd prefer that to remembering to split those every time, but not a big deal if it's just one or two.
it should be a reasonably simple task for someone who would like to start contributing to the code base, though of course you'd have to understand the overall design of the phone number normalizations we already do
some sort of normalization around the drugs patterns would also be welcome; the patterns we routinely put in for those are ridiculously complex
@CodyGray The - should be replaced with \W, it catches more text in the markdown. Maybe even replace it with \W* though I don't know how useful that would be.
@RyanM That pattern looks like it's already caught by Mostly punctuation marks in answer and Mostly punctuation marks in body; append -force if you really want to do that.
@SmokeDetector hmmmm....it seems like that couldn't have worked when that answer was posted, based on the other answers (and it doesn't work now)...possibly suspicious, given the article link
