(a,b)=(0,0) and (a,b)=(1,0).
By adding 1 to both sides of the given equation, we can rewrite it as
a8−a7+1=(19b+1)2.
Left hand side of this equation has the following factorization:
a8−a7+1=(a8−a7+a6)−(a6−1)=(a2−a+1)⋅(a6−(a+1)(a3−1)).
Let d=gcd(a2−a+1,a6−(a+1)(a3−1)). Since d∣a2−a+1 we get d∣(a+1)(a2−a+1) and d∣a3+1. Thus, a3≡−1(modd). Therefore,
a6−(a+1)(a3−1)≡1+2(a+1)≡2a+3(modd).
Now since d∣a(2a+3) and d∣2(a2−a+1) we get that d∣(5a−2). Finally, d∣2(5a−2) and d∣5(2a+3) yield d∣19, which means that either d=1 or d=19. Note that since d2∣(19b+1)2, d=19 is not possible.
Thus, a2−a+1 and a6−(a+1)(a3−1) are coprime integers whose product is a perfect square. Therefore, each of them should be a perfect square. Note that (a−1)2<a2−a+1<a2 for a>1, and (a−1)2>a2−a+1>a2 for a<0. Hence, a2−a+1 is not a perfect square except a=0,1. Inserting a=0 and a=1 to the main equation we get two solutions: (a,b)=(0,0) and (a,b)=(1,0).