I think it's better try to contact everybody no matter how minor the contribution, because (a) it's polite and reduces the chances of somebody getting annoyed a

I think it's better try to contact everybody no matter how minor the contribution, because (a) it's polite and reduces the chances of somebody getting annoyed and (b) it is easier to tick somebody's name off a list than to go in and look at their change and make a judgement call about how minor 'minor' is in each case. Better to save that extra effort for the cases where you get no response or a negative response. (I worked on getting a portion of QEMU code relicensed -- that had a fairly small number of authors to track down and get permission from, but it was still pretty painful. I'm impressed with anybody with the perseverance to attempt it for an entire large and old codebase.)