@MattSamuel SmokeDetector's (SD's) user blacklist is a highly dynamic list of users who have gotten feedback on a post they authored which explicitly directs SD to add the user to the user blacklist (or the user has been added by explicit command, but that's relatively rare). The feedback is tpu, which is True Positive, with the u indicating to put the user on the user blacklist.
Any time a post the user authored receives FP (false positive) feedback, the user is automatically removed from the user blacklist. Being on the user blacklist causes SmokeDetector to report any posts SD scans which are authored by that user, and adds a bit of "weight" to the detection of the post. The extra "weight" is the historical True Positive percentage of posts which were detected by a "blacklisted user".
In the specific case of that user, the first time that post was reported it received tpu feedback from one user, which put the author on the user blacklist (that feedback has since been retracted). I saw the post a bit after it was reported a second time after being edited 8 hours later. At that point, I gave FP feedback to both the first report and the second report. The first FP feedback I gave removed the author from the user blacklist.
@Magic_Isa "Bait", as I have used it, means words or phrases which don't convey the content of a link, but which are meant to lure (or have the effect of luring people) into clicking on that link.
See "clickbait".
I find clickbaity links or question titles offensive. They seek to get pageviews by intentionally hiding or obfuscating the actual content of a question or page. Titles should be descriptive. Clickbaity titles are the opposite of descriptive.
@Magic_Isa Also, to address the implied meta question: I very much want the HNQ to advertise the very best that MSE has to offer. I don't want users from other sites to be lured in by useless clickbaity titles, and I particularly don't want users from other sites to be presented with low- to no-context PSQs via clickbaity titles. On any given day, there are dozens of posts on MSE which achieve the "hotness" required for the HNQ, but only five slots.
Better to remove poor questions from the HNQ (or, at the very least, retitle them so that the titles are informative) in order to make room for the good ones.
I don't know about you, but I care about the way in which MSE is perceived by the users of other SE sites, and by other people in the mathematical community. Since the HNQ is one of the principle ways in which MSE reaches out to other communities, I think that it is worth taking some time to curate it.
In case anyone still has deletion votes left, most of the above old posts, i.e. request for deletion of which was made before today, needs only one more vote.
@Makyen @Saad One of the nice things about the form [C###](link) is that the userscript which helps to manage this room formats the ink in a nice way. Links of the form [c####-##](link) don't play nice with the userscript. Is it possible for you, @Saad, to use the C### and D### format.
Alternatively, @Makyen, can the script be modified to play nice with the [cmmdd-##] format?
I'd do it myself, but I am not clever nor invested enough to take the time to figure out what needs to be changed.
@XanderHenderson I can make that change. the script understanding the other formats wasn't intended as a restriction on what could be useful in this room, but merely what I'd seen as the formats which were commonly used here. The primary limitation is to not expand what the script chooses as acceptable to change past the point where the displayed information that's replaced is what the user really wants to see.
math.stackexchange.com/questions/3653570/… I feel this question should be reopened. It was closed as a duplicate. There is a subtle difference, I feel. The first one is asked for a soft answer, the second one wants a hard answer.
The question asks why z^n has an antiderivative for all n but n=-1. He wants a soft answer and doesn't literally want a proof that 1/z has no antiderivative. I think it is different enough to not be a duplicate.
@Saad No, it can be as many digits as you want. Specifically, the displayed link text has to match the regular expression ^\s*[cd]?(?:\d+|[a-z])\s*$, or be just the partial URL which SE Chat uses for when a user posts a bare URL, which is matched by the regular expression ^\s*(?:math\.stackexchange.com\/(?:q(?:uestions)?|a(?:answers)?|p(?:osts)?)(?:\/\d+(?:\/\d+)?)?(?:\/…)?|question|answer)\s*$
@Saad np. As you can probably tell, it's very easy to change the first regular expression to match having a "hyphen-minus" character in the link's displayed text. Are there other additional formats that are desired to be used/matched (i.e. be replaced by the title text) (cc @XanderHenderson)?
@Makyen I haven't seen any other commonly used formats. That said, I'm pants at regular expressions, thus your code looks like gobble-dee-gook to me. Thanks for maintaining it. :)
@AlexanderGruber You might want to look at the page Set Up and Run SmokeDetector, which has copy-&-paste instructions for getting SD basically running. For what's been described as desired here, you'll need to: Change the "config" file to use a bot account of your choice. Also in that file point it to the GitHub fork you've made and supply it with some credentials, if you want it to be able to do actions on GitHub.
You'll want to create your own rooms.yml, or rooms_custom.yml, which defines only the chatroom where you're going to have it report. You'll want to clear, but not delete, all the bad*.txt, blacklisted*.txt, and watch.txt files.
The actual detections are all in "findspam.py". You'll probably want to delete almost all of the lines which are in blocks annotated by @create_rule, or that is a call to create_rule(), probably except the blacklist/watch stuff.
You can then create your own rules, which can be just a regular expression, or a function, using either create_rule() or @create_rule.
Without additional changes, your bot will scan all posts on SE, but you can limit what's reported to only Mathematics through the configuration defined in rooms.yml. If I was doing it, I'd likely make the quick change of only actually scanning Mathematics, but that's an optimization that doesn't have to happen. To do so, it's probably easiest to have your bot still scan the same feeds, but just toss out anything not from math.se.
(cc: @MartinSleziak)
@XanderHenderson np. I'm happy to help. The line that I'll be changing is this one, which will change to if (/^\s*[cd]?(?:[\d-]+|[a-z])\s*$/i.test(link.textContent)) {
The CRUDE version of the URRS has been updated to Version 2.1.0.101 (GitHub) (install).
While I'm making changes, it looks like there was an old stated desire to have the script operational in other rooms associated with Mathematics. If there's a list of rooms in which people would like the script to run, I can easily add additional rooms to the base configuration.
For ex if we have equation like..$y^2=4ax$.
And we wish to find the equation through any point $(h,k)$ which lies on tangent line rather than parabola. Actually my brain stuck with this😂
Request for y'all. I need help collecting examples of: crank questions, annoying duplicates (and associated duplicate targets), and PSQ red flag phrases (such as "i have no idea" or "urgent"). To prevent this from derailing the room, please discuss here.
@EnlightenedFunky The close box for that question reads
> This question is missing context or other details: Please improve the question by providing additional context, which ideally includes your thoughts on the problem and any attempts you have made to solve it. This information helps others identify where you have difficulties and helps them write answers appropriate to your experience level.
It was closed three years ago, when the generic close reason was "Off-Topic". With respect to the new closing dialogs, the corresponding close reason is "Community Specific Reason".
Recently published on the arxiv : https://arxiv.org/pdf/2005.00346.pdf
The paper gives a proof of the Collatz conjecture by analyzing Diophantine equations related to a slight generalization of the Collatz map.
The look of the paper is very professional and seems really well written which gives...
Regarding the specifics of that question, it is a PSQ (a problem statement question). The asker has given no context (in the form of definitions, theorems, other results they know, hints, a reference to a text or set of lecture notes, or even an attempt at the problem).
@XanderHenderson "I find clickbaity links or question titles offensive." I think you are taking it too personal. I don't think it's intentional, not in the particular case above. On the other side, you might be right for some other cases. Question: if the question above would have a descriptive title (no MathJax) would have you agree on having it on the HNQ?
@Magic_Isa The entirety of my comment read "Any question with "hard", "weird", "tricky", etc in the title is (1) bait for the HNQ and (2) likely not of very good quality. :\"
So, no, I don't think that the question belongs on the HNQ (or on the site at all).
@XanderHenderson I know, but my question was for the last version, the one with attempt. So again: if the question above would have a descriptive title (no MathJax) would have you agree on having it on the HNQ?
For ex if we have equation like..$y^2=4ax$.
And we wish to find the equation through any point $(h,k)$ which lies on tangent line rather than parabola. Actually my brain stuck with this😂
