My article To make a function nicer without changing it much, convolve it with an approximate delta function is doing very strange things – somehow too strange for me to be able to do just by typing in the LaTeX wrongly (not that I can see anything wrong with it). Is there some more serious bug?

It seems like you have a mistaken dollar into a "maths" environment

(in \[ maths]f*g(x)=\int_{-\infty}^\infty f(x-y)g(y)dy$.\[ /maths]).

I'm going to correct it.

Thanks!

Olof and I are planning to do some work next week which will result in such errors being handled much more gracefully.