what, "hard to get correct"? Are we talking about fundamental computer science that literally consists of implementing an algorithm made out of compare and swap/move elements...?
I can't imagine any sort algorithm for which it can possibly be legitimate to say it's 'hard to get correct' unless it's in the sentence 'it's hard for a first-year computer science student to get correct.'
i mean, what is more traditional and basic than implementing tricky sort algorithms?
"When Jon Bentley assigned it as a problem in a course for professional programmers, he found that an astounding ninety percent failed to code a binary search correctly after several hours of working on it,[7] and another study shows that accurate code for it is only found in five out of twenty textbooks.[8] Furthermore, Bentley's own implementation of binary search, published in his 1986 book Programming Pearls, contains an error that remained undetected for over twenty years.[9]"
I just learned that Java has an unsigned right shift operator >>> from an article cited as a reference in that Wikipedia section. The article has some more details on the subtleties of binary search. For instance, the 20 year old bug in an algorithm proven to be correct was down to integer ranges:
int mid = (low + high)/2;
breaks (obviously in retrospect, am I missing something?) for
low + high > Integer.MAX_INTEGER
It can be replaced by the elegant (but to me, somewhat oblique)
Also from that page, here's the link to the bug in Sun's tracker, priority 2-High, evaluation: "Can't even compute average of two ints" is pretty embarrassing.
The video is not demonstrating an in-place merge sort. Notice how the sorted pairs suddenly merge. There's another buffer (not included in the visualization) where each merged list is temporarily stored while it is being built up. The sudden merger is when that temporary buffer is copied back into the main list.
