To quote my former computer science professor, the only good thing about bubble sort is it's name.
It's a very slow algorithm although it's conceptually easy to understand
So if you're working with a very large dataset it can run painfully slow
However, with a small dataset and with the faster CPUs' these days I guess the difference isn't that noticeable
Try quicksort. At least it can perform much faster unless the dataset is sorted in reverse order already