> Few unique characters in body, repeating characters in body, repeating characters in title ---------- Title - Repeated character: *FFFFFFFFFFFF* Post - Contains 2 unique characters Post - Repeated character: *FFFFFFFFFFFF*
Okay, I'm trying something. Temporarily enabling real flags (not dry run). I'm going to abstain from flagging anything it does so we're not actually adding any flags. Also, I'm logged in an Smokey in a browser so we can revert any bad flags (which shouldn't happen anyway).
Louis Alexander Slotin (1 December 1910 – 30 May 1946) was a Canadian physicist and chemist who worked on the Manhattan Project. During World War II, Slotin conducted research at Los Alamos National Laboratory. He performed experiments with uranium and plutonium cores to determine their critical mass values. On 21 May 1946, Slotin was conducting a demonstration when he accidentally initiated a fission reaction, which released a burst of hard radiation. He received a lethal dose of radiation and died of acute radiation syndrome nine days later. Slotin was the second person to die from a criticality...
Okay, so here's a potential path forward on the integration thing: As a no-effort option for SE, just allow us to cast up to 5 (maybe 6) flags at our option. Boom, spam lasts no time at all.
I would have no issue casting 5 flags at 99.9x% accuracy
For a project I'm working on, it'd be nice to be able to retract flags through the API.
We allow people to retract flags from the web. Could this functionality be extended to the API? It could work such that you receive a 'Retract Flags' option on calls to /questions/{id}/flags/options, then pas...
metasmoke sorta has three levels of importance... If it's on the tools page, it's useful but mostly unimportant. If it's in the dropdown, it's important, but not a big thing. Top level menu items are big things.
@Undo could add a status indicator to the menu item, like status
Another feature that would help auto flagging is being able to determine the status of a flag. Then we could automatically look into ones that were declined
As a companion to my other feature request regarding flag statuses, and a previous feature request for the same thing, I would like to request the ability to get a flag's status via the API.
There are applications that allow you to flag questions, answers and comments without being on a Stack Ex...
@Undo @ArtOfCode I'd like to see an "Active", "Dry Run" or "Disabled" status indicator too, on the page, to show whether actual flagging is enabled or not
@JanDvorak closevoting is something that requires site-by-site knowledge of what is on-topic and what is not. it therefore is quite hard to automate using one single bot
@JanDvorak but you deal with less code at a time, plus the respectable members of each community can customize their autoclosevoting bot to make it cast close votes as accurately as possible without messing with other communities' configurations
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, link at end of body, pattern-matching website in body, repeated URL at end of long post: We’ll cowl: T Perfect Youth Serum by jeyixigu on askubuntu.com (@ThomasWard)
[ SmokeDetector | MS ] URL in title, bad keyword in body, bad keyword in title, pattern-matching website in body, pattern-matching website in title, repeated URL at end of long post: nutrahealthtrimsite.com/vinetics-c/ by user83064 on graphicdesign.stackexchange.com
@ArtOfCode Done. Currently accessible by all users, I figure we can yell at anyone who abuses it and it gives us some assurance that someone should be around to kill it at any time.
@Rob given 100 posts out of which 95 got flagged, what could be done about the final 5 to hopefully have similar posts flagged in the future? task 1, find them. task 2, analyze them for patterns which are common in some types of spam and uncommon otherwise
as a quick exercise, find the Smokey reports above which correspond to messages which were not autoflagged
so far, the metric looks like it's working fine, but there is rather significant overlap between some of the heuristics, so there's a risk that the weight could be inflated by multiple rules tagging on to the same pattern
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, link at end of body, pattern-matching website in body, repeated URL at end of long post: perfect youth serum by user62177 on workplace.stackexchange.com