You save an awful lot of time when you know have an idea of what's going to be slow up front.
I've spent so much time trying to chase down bugs caused by junior programmers because something that worked fine with 100 rows in db, slows to a crawl when we move a feature out of beta and we suddenly have 1000 rows.
If they had some idea that the query they were writing was O(n^3) they could have done a quick estimate and saved us all the trouble.
I've spent so much time trying to chase down bugs caused by junior programmers because something that worked fine with 100 rows in db, slows to a crawl when we move a feature out of beta and we suddenly have 1000 rows.
If they had some idea that the query they were writing was O(n^3) they could have done a quick estimate and saved us all the trouble.