jrullan/neotimer

Millis overflow and signed longs

Opened this issue · 1 comments

mtib commented

This is how it is determined that the time interval has expired. I am uncertain whether that will behave correctly when millis rolls over. (Like described here)

Shouldn't the time and last variable be stored as unsigned long to work? (ref)

But I might also just have a faulty reasoning, but I'd like to use this library in a long running application, so I wanted to bring this up.

Great question! I'm also testing the same!