I recently say, some arguments, in a problem. The arguments, made me ponder upon it for hours. The question was:
Find number of triplets of positive integers satisfying
$2^a-5^b\cdot 7^c=1$
Given options are $0 , 1 , 2$ or infinite.
The solution goes like this in the thread,https://math.stackexchange.com/questions/1827071/number-theory-problem-finding-triplets?noredirect=1 :
Note that $ 2^a\equiv 1\pmod 5$ implies $a\equiv 0\pmod4$, so write $a=4m$. Now $$5^b7^c=2^a-1=(2^{2m}-1)(2^{2m}+1)$$ and the two factors are coprime. Take it further, $$5^b7^c=(2^{m}-1)(2^m+1)(2^{2m}+1)$$ and all three factors on the right are pairwise coprime. So at most one of them is divisible by $5$ and at most one is divisible by $7$, meaning that the smallest one must be $=1$, i.e., $m=1$. This gives us $$5^b7^c=(2^1-1)(2^1+1)(2^2+1)=15$$ which is absurd.
Note that we used $b\ge 1$ to conclude $2^a\equiv 1\pmod 5$. If we were to allow $b=0$, we find additional solutions such as $a=1$, $b=c=0$ or $a=3$, $b=0$, $c=1$ and beyond that enter the realm of Catalan's conjecture.
But I am particularly, confused, about this statement "Note that $ 2^a\equiv 1\pmod 5$ implies $a\equiv 0\pmod4$, so write $a=4m$." ? Is this sort of, always true? For example, if $a^m\equiv c\pmod n$ and $a^n\equiv c\pmod n$ such that (WLOG) $m<n$ then it is not always true, that $m|n.$ So, I want to ask, under which circumstances can we make these sort implication? What are the criterias (if any) that must be satisfied, to go with this analogous reasonings?