Mid point in divide and conquer algorithm implementations

Question

I've seen many implementations using below to find mid point of two indices:

int mid = lo + (hi - lo) / 2;

instead of

int mid = (lo + hi) / 2;

Mathematically, I see no difference and yet, I've never seen anyone using the below one. Is there a difference between the two computationally?

And this [oldie, but goodie](https://research.googleblog.com/2006/06/extra-extra-read-all-about-it-nearly.html) — OneCricketeer, Nov 17 '16 at 02:31

score 1 · Accepted Answer · edited Nov 17 '16 at 02:28

1

There exist a maximum positive value for a 32-bit signed binary integer in computing.

We assume this value is 100.

int lo = 60;
int hi = 80;

then lo + hi = 60 + 80 = 140 > 100, it is dangerous to do so because it will cause a integer overflow error.

edited Nov 17 '16 at 02:28

OneCricketeer

answered Nov 17 '16 at 02:25

Vau

"it will cause a integer overflow error" Integer overflow isn't really an error, is it? It's just the way computers handle the erroneous situation of a number requiring too many bits to represent. – The SE I loved is dead Nov 17 '16 at 02:50

1 Answers1