i'm not sure; i made a few mistakes in each of them as i was writing them, but they were both working by the time i first tested them. i do feel like the insertion sort was maybe a little more challenging?

edited to add: someonefromca spotted a bug in the insertion sort; where it says

    if (tmp > a[j-1]) break;

it should say

    if (tmp >= a[j-1]) break;

which is both a performance bug and also breaks the stability property the algorithm otherwise has

this is some evidence in favor of your point that insertion sort is more error-prone; even if i could have made the same error in the bubble sort, i didn't, and possibly that was because the rest of the algorithm required less mental effort?