I disagree and you have not given me a reason to agree. You can't even stick to one definition of space and so you don't make a clear point. I disagree. $$\mathbb{R}^n$$ doesn't have a metric until you specify one. You are assuming the Euclidean metric $$\sqrt{\sum_{k=1}^{n} \left(\Delta x_k \right)^2 }$$ but so far have failed to rule out the pseudo-metric $$\left(\Delta x_n \right)^2 - \sum_{k=1}^{n-1} \left(\Delta x_k \right)^2 $$ which
induces the Euclidean metric $$\sqrt{\sum_{k=1}^{n-1} \left(\Delta x_k \right)^2 }$$ on any slice with fixed value of $$x_n$$. The Euclidean metric is unsuitable for 4-dimensional space-time because it would allow treating time exactly the same as any spatial direction, which is not observed; however the pseudo-metric you object to without cause allows us to induce a Euclidean metric in
any slice of simultaneity and to induce a 1-D Euclidean metric along any time-like direction.
But $$\mathbb{R}^n$$ is not a metric space until you specify a metric.
That cannot be the issue since Minkowski space uses a 4-dimensional Cartesian coordinate system just like a Euclidean 4-dimensional space.
No it doesn't. You are just advertising your ignorance and lack of experience with Minkowski space. You are also confusing if you are talking about an address or a vector which is modeled as a
difference of addresses. This means we don't know if you are talking about a vector space (which has a well-defined origin) or an affine space (which has no concept of an origin).
It's like you can't do geometry at all, but that is not a failure of Minkowski space but merely a statement of your inability.
A function must map one element of its domain to exactly one element of its range. Famously, the Lorentz transforms and Euclidean rotation maps are functions -- indeed, they are one-to-one functions so their inverse mappings are functions. Specifically, there are unit vectors in Minkowski space such that any address turns into location via $$\mathcal{X}(x_1, x_2, x_3, x_4) = \sum_{k=1}^{4} x_k \hat{e}_k$$ which is as functional as any function can be.
//Edit:
That's $$\mathcal{X}(x_1, x_2, x_3, x_4) = \mathcal{O} + \sum_{k=1}^{4} x_k \hat{e}_k$$ for an address or $$\mathcal{X}(x_1, x_2, x_3, x_4) = \sum_{k=1}^{4} x_k \hat{e}_k$$ for a vector but the distinction is blurred when $$\mathcal{O}$$ takes on the properties of 0 as it does in a vector space.
I never claimed it was a manifold -- I claimed it was pseudo-Riemannian manifold.
But this can be done if you consider first the hyperbolic neighborhood of a point a with "radius" r. Thus this is all points with $$\left( <x-a, x-a> \right)^2 < r^2$$. Then we choice as our basis all the non-empty intersections of two or more of these hyperbolic neighborhoods where coordinate values don't run off to infinity. This it turns out results in a topological space on Minkowski space where the basis sets have the property of being Lorentz-invariant. Further, this is a topology we are well-versed in as it is the Euclidean topology.
http://mathpages.com/rr/s9-01/9-01.htm
And for a pseudo-Riemannian manifold, Minkowski space is the canonical example since the whole point of a pseudo-Riemannian manifold is to show that locally it approximates a Minkowski space and nothing approximates a Minkowski space like a Minkowski space.
Lies! You don't have the math credentials to do multivariate calculus, so you can't follow logical axioms to theorems.
Lies! (As demonstrated above.) That does not follow, so you have perverted the term "logic" by your use here.