@Feeds The lack of trigger discipline in that thumbnail really irks me. Dude's got a lotta experience with guns and should not be demonstrating such unsafe behavior. (It's probably intended to get more clicks, methinks.)
@DJMcMayhem I've never been to a gun range, though I'd like to at some point. (I have fired guns way out in the country three times though.)
@SriotchilismO'Zaic Of all the things to harp on repeatedly, trigger discipline is a pretty good one. Guns Are Cool™ but also Guns Are Dangerous™ and reminding people to be as safe as possible with them is always a good thing.
@SriotchilismO'Zaic Out of pure curiosity, what would you say are the most important rules of gun safety? There's typically four and I'm really curious as to whether you've been exposed to them enough - despite not knowing much about guns - to remember them.
1) Do not point a gun at something you do not want to shoot. 2) Do not put your finger on the trigger if you do not want to shoot. 3) Keep your gun locked and secured when you are not using it. 4) Idk wear ear and eye protection?
You definitely got the first two right. Third is also vital and fourth is good, but not strictly necessary. (Well, ear protection is pretty good to have for most people.)
I was actually trying to remember them myself, and one of them is: don't hand a loaded gun to someone else. (Always let them load the gun themselves.)
Corollary to #1 and #2 is that the safety doesn't necessarily work. I remember a demonstration where someone had a gun (without ammo!) that had a worn down safety (due to lots of use) and they showed that it was still possible to pull the trigger despite the safety being on.
Is it possible to arbitrarily overfit a sine wave? You're given N points (x_i, y_i) and you want to fit them with a function y = a*sin(bx+c) so that all residuals |f(x_i) - y_i| are at most epsilon for some fixed epsilon. I wonder if this is possible using some extremely large frequency b, assuming the inputs x_i are linearly independent over the rationals.
