[steerablesafe]

> I understand there's a good chance the next standard will specify that signed integer overflow results in an unspecified value

Do you have a source for this?

[lmm]

It was linked from a discussion on HN but I didn't record where.