[antidesitter]

Nearly all the objections people have raised are directly addressed in the post.

Are they? I see nothing in the article about the negative consequences of surprising users with silent failure, for example. Or the fact that it makes little sense in real-world scenarios.

would you object to me defining the piecewise function f:R->R defined to be 1/x for x =/= 0 and 0 for when x=0?

What do you mean by objecting to you defining a function?

[Zalastax]

The post is not about whether this is a good choice for its users; the post is about whether there are nathematical objections to it. The author even writes that they don't agree with the engineering decision to define it like that, but they can't object with an appeal to mathematics. It lists three valid choices for handling division by zero and Pony picked one of them. My personal favorite is to shrink the domain but that's not supported in most languages and those that support it are often not meant for number crunching.

[a_wild_dandan]

> "see nothing in the article about the negative consequences of surprising users with silent failure, for example."

That's because the article is purely about the mathematical consistency of using 1 / 0 = 0. It explicitly and repeatedly mentions that it's not about real-world logistics.

> "What do you mean by objecting to you defining a function?"

The author is defining the division function with this special mapping, f, which includes 0 in the domain. Damn near everyone in this thread is confusing that definition with defining the multiplicative inverse, 0⁻.

It's an understandable confusion, but also honestly getting pretty frustrating. That's what he or she is getting at.