Um, not to pick on this too much (but after all it's a WTF), but I just realized that your division was reversed anyway; the long one takes 24.64 times as long as the short one. That doesn't translate to the short one being 2464% faster.
D
DWalker
@DWalker
0
Reputation
2
Posts
27
Profile views
0
Followers
0
Following
Best posts made by DWalker
This user hasn't posted anything yet.
Latest posts made by DWalker
-
RE: Even Middle School Kids know better than Oracle
-
RE: Even Middle School Kids know better than Oracle
"4344, which is 2464% faster"
Um, what does x% faster mean? I take it to mean that something occurs in x% less time. Clearly this isn't the case. Although "x% faster" may not actually have a strict definition. Much clearer would be to say "This shows a cost of 4344, which runs in 4% of the time as the full table scan".
This reminds me of a SyncSort ad back in the mainframe days, in the Computerworld newspaper, which quoted a mainframe sysadmin saying "Now that we use Syncsort, we get our daily workload done in 240% less time than before". Ha! I'm sure that the Syncsort company employs some mathematicians, but obviously they don't check the ad copy.