Comment by saagarjha

You can see that the blog post does not claim to solve the halting problem; instead, they pick a subset of behavior that is likely to be useful.