@K.Dᴀᴠɪs No, it didn't. It should not have detected the post. Our feedback is supposed to be based on the post, even if that's based on the Markdown, instead of the HTML which SD looks at.
So what if someone spelled out an obscene word using html that wouldn't have been detectable just looking at the source code, but when formatted it was rude?
@SmokeDetector The post had formatting problems which resulted in the majority of the Markdown being stripped from the HTML. The post was intended to include several lines in code format. That was probably not done by the author because of the code-only filter.
@K.Dᴀᴠɪs Yes, we are supposed to base our feedback on both. In other words, the totality of the post, even if that's more than the HTML which SD looks at (e.g. that it would have needed to look at the HTML too, or instead). Or even if it's HTML that pulls in an R/A image, etc.
Two things. Firstly, NAA doesn't affect the reason weight - making it a relatively harmless response. Also, if someone intentionally posted a bunch of @@ @ @ @ @@@ as an answer, and we tell Smokey that is a false positive, then what? I still feel that Smokey correctly identified a post - even if the post was incorrectly formatted by the user - Smokey did what it was supposed to do.
@K.Dᴀᴠɪs SD did what it was programmed to do, but not what it's "supposed" to do. Ideally, it should also recognize that the formatting caused HTML to be stripped. I'd agree that it should be NAA, if we were supposed to feedback based just on the HTML that SD looks at, but we are supposed to consider the entire post. It's not clear to me that this is the best choice from the POV of getting data which can be used for training ML.
@K.Dᴀᴠɪs HEAD isn't at tip of origin's master branch (local [2d3342e3]ArtOfCode-: Auto watch of 88situsbandarq\.info by K.Dᴀᴠɪs --autopull, remote [6728007f]SmokeDetector: Merge branch 'master' of github.com:Charcoal-SE/SmokeDetector)
@K.Dᴀᴠɪs HEAD isn't at tip of origin's master branch (local [2d3342e3]ArtOfCode-: Auto watch of 88situsbandarq\.info by K.Dᴀᴠɪs --autopull, remote [6728007f]SmokeDetector: Merge branch 'master' of github.com:Charcoal-SE/SmokeDetector)
> Potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Title - Position 0-8: trimfire Body - Position 0-8: trimfire Username - Position 0-8: trimfire
> Would not be caught as a post, title or username.
@iBug HEAD isn't at tip of origin's master branch (local [2d3342e3]ArtOfCode-: Auto watch of 88situsbandarq\.info by K.Dᴀᴠɪs --autopull, remote [6728007f]SmokeDetector: Merge branch 'master' of github.com:Charcoal-SE/SmokeDetector)
@ThomasWard It looks like git got messed-up on Solar due to the master & deploy branch histories being rewritten (via push --force). You'll probably need to manually do a pull --force, at a minimum.
I agree that overriding CI/Travis failures is dangerous. It should only be done in very limited circumstances, but we've had periods when CI/Travis have failed for unrelated reasons (e.g. limited login retries). Yes, we should be very cautious about doing so for non-trivial changes, but for automatic changes to the watch/backlists, the risk is low.
@Makyen In that case, we try adding code that bypasses the failing checks instead of pulling in spite of the failures, which IMO is much safer, though it requires push access to handle.
@DJMcSpookem SD is undergoing some maintenance, this command may or may not have gone through. This command may need to be repeated once SD is back up and running well.
@ArtOfCode Can you check the git config on your instance? It seems you committed the blacklist changes (Smokey author'd them), better git config user.name ...
at least I use two different clones for production and development, and the production clone has committer info set to Smokey
@Makyen Removed /(?:(?:how-to-)?fix-(?:error-(?:\d+-in-)?)?)?(?:asus|canon|epson|fire-tv|ipad|hotmail|hp-printer|lenovo|lexmark|quickbooks)- from blacklist
@Makyen Hmmmm... that should probably not be done as an edit. It's better to have the record if someone goes searching for it. Although, it's in the SD reloaded text. That domain is now at 7/4/1, which isn't good enough for the blacklist.
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, bad pattern in url body, blacklisted website in body, body starts with title and ends in url, +3 more (751): Testo Drive 365 Reviews by testodrive365 on meta.SE
[ SmokeDetector | MS ] Url-only title, bad keyword in body, bad keyword in title, blacklisted website in body, blacklisted website in title, +4 more (693): www.bluesupplement.com/ketozin/ by picasper on askubuntu.com
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, pattern-matching product name in body, pattern-matching website in body, +1 more (492): The 90-10 Rule For Fast Weight Loss by littelirow on graphicdesign.SE
[ SmokeDetector | MS ] Blacklisted website in body, potentially bad ns for domain in body, potentially bad keyword in body, potentially bad keyword in title (102): PropseRX Hair ReGrowth by qpelaas on apple.SE
