2 ms·
In math, it's common to extend the numbers, but the problem is that most of the times you lose important useful properties. For example in quaternions you add i
by gus_massa 4y ago
In math, it's common to extend the numbers, but the problem is that most of the times you lose important useful properties. For example in quaternions you add i^2=j^2=k^2=-1, https://en.wikipedia.org/wiki/Quaternion https://en.wikipedia.org/wiki/Quaternion but the problem is that ij=-ji, so the multiplication in new set of numbers is not commutative. It's possible to continue extending the numbers adding more fake "roots of -1", but the properties of the operations are even worse.
There are a few ways to add infinite to the operations, and add some operations for them, but the extensions breaks the properties of the operations https://en.wikipedia.org/wiki/Cardinal_number https://en.wikipedia.org/wiki/Cardinal_number
Physicist use something like what you are proposing, using Taylor and hand waving. I read a few tries to formalize that, but they are just too weird and no one use them. I think the siplest example is https://en.wikipedia.org/wiki/Graded_ring https://en.wikipedia.org/wiki/Graded_ring that you can use to define ϵ and a number is something like 7.9+3.1ϵ+27.4ϵ^2+... So you have that ϵ is almost like 0, and ϵ/ϵ=1, but 7/ϵ is undefined.
It look hard to make an extension where ?/0=1. I think it's better to be happy with something similar enough.