Quote:
|
Originally Posted by smogo
|
I hope you don't take offense at my posting the following comment, since it is clear that your heart is in the right place.
I just downloaded and examined some of your code and it is the funniest thing I've seen in a while. The "hashtable" in particular is the funniest data structure I've ever seen.
You state in the code that it should have O(n) inserts and O(log n) searches. This is the first clue that something is seriously wrong. A proper hash table should have O(1) inserts, O(1) deletes and O(1) lookups (assuming good hash functions). What you have is really a sorted vector. The really funny part, though, is when you realize that it isn't even a good sorted vector! The search function is recursive, which is a little ugly. The really funny part is that you walk the vector to insert in sorted order instead of using the binary search function you've already created. And all of these numbers are biased by using comparisons as our atomic operation. If you look at overall efficiency, the story saddens considerably more.
A good knowledge of data structures is the most dramatic difference between a good programmer and a poor one. In my experience, 80% of all programming tasks are solved by setting up appropriate data structures and then manipulating them (classes are data structures when using this view). I commend you for trying to write your own, since that can be an excellent way to learn, but still felt compelled to rip on you a bit for such a funny implementation.
Threaded Mode
