@MikeMiller It's probably for the greater good, in a way; now, whenever someone Googles "books to learn [X] from," they can easily find a nicely compiled list
guys how can I prove that there exists $ab = nq$ for all $a,n,b,q \in \mathbb{Z}$?, assuming $a,n$ are fixed integers and $1 \leq a \leq n$ and $1 < n$?
yes but it seems there is no such $b$ for that case. I see, it makes sense that they prefaced it by saying $(a,n) \ne 1$ I thought it wouldn't matter if they had a common factor or not :p
So, a Riemannian metric on a manifold is a smooth choice of inner products at each tangent space $TM_p$ where $p$ is point of $M$, smooth in the sense that given any two smooth vector fields $X, Y$ on $M$, my metric eats $X(p), Y(p)$, spits a scalar in $\Bbb R$ and this gives me a smooth function $M \to \Bbb R$.
@AkivaWeinberger I don't feel comfortable telling stories about things I don't know well to someone who isn't familiar with that topic, because I may tell them my personal wrong point of views about things (resultant of not knowing the story well) or even worse, wrong facts about things.
So, I don't remember how to prove that the metric induced by the Riemannian metric is actually a metric. I think $d(p, p') > 0$ for $p \neq p'$ is the hardest part.
@Sᴋᴜʟʟᴘᴇᴛʀᴏʟ I never encountered any math txtbook asking me to create computer programs before. What is the point of that? Perhaps coding something procedurally gives more insight on what they are trying to understand. What about functional programming languages like haskell then? I don't know much about them but they seem to take out a lot of the 'effort'.
work in "normal coordinates", and use these to prove that the shortest path in these coordinates is an actual geodesic (as opposed to a broken one); see that any geodesic from the center of the coords to your point x that leaves the chart must have length at least, say, C; and therefore any geodesic is at least the length of the unique geodesic from the center to the point in your normal coords
something like this
only works for close points, but if the points aren't close... :)
@AkivaWeinberger You should work out why inf of sum of all piecewise linear partitions of a curve $\gamma$ in $\Bbb R^n$ is the same as $\int_0^1 \|\gamma'(t)\| dt$ at some point of time. not really complicated, just a proof-worthy fact you didn't seem to know.
Alright, so I guess I will get to sleep today instead of more math. Will start on serious math tomorrow.
I should probably stop discussing fairy tale math and head to bed. That should make me less guilty, because discussing fairy tales while not doing the real work is probably more sinful.