Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

I was allowed to set a flagging condition that was below the permitted threshold #945

Open
gparyani opened this issue Aug 28, 2022 · 0 comments

Comments

@gparyani
Copy link

gparyani commented Aug 28, 2022

Currently, I have an autoflagging condition set for Meta Stack Exchange. The allowable minimum threshold for accuracy rate, according to the site, is 99.75% or higher, and what's supposed to happen if I attempt to set a lower rate is that I get an error message stating it's not allowed (either immediately if I check the box to enable flagging, or once I try to edit the condition again if I don't).

However, the system permits me to set certain flagging conditions that are below the allowed threshold and doesn't give any error message. In particular, it allowed me to set one that was only 99.71% accurate (minimum weight 103, maximum author rep 1, and minimum reason count 1). As soon as I tried to set it with a minimum weight of 102 (99.67% accurate), I got an error message that it was too low. However, increasing it to 103 allowed me to set it even though the accuracy rate was below 99.75%.

I found out this bug when I tried to set a condition that was (seemingly) exactly 99.75% accurate, but after I successfully set it, I immediately got pinged in chat that SmokeDetector had disabled my condition because it was too low. I was initially going to file a bug report that metasmoke should always round down/truncate rather than use standard rounding rules (it was actually 99.748%), but I ended up discovering that even a lower rate was accepted as long as it was above 99.68%.

Can this please be fixed to properly show error messages for such conditions, and also tweaked to always round down (truncate) the percentage rather than use standard rules (so users aren't pinged out of the blue that their condition isn't accurate enough even though the system told them it was)?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

No branches or pull requests

1 participant