On (23 Jul 97) Thomas Maeder wrote to Cameron Clark...
 TM> Have  you  tested the speed of exception handling? I haven't, neither,
 TM> but it's said to be slow, slow, slow.
I participated in a benchmark on the speed of exception handling about a
year ago or so.  Overall, the speed was found to be far less dismal than
expected -- while exiting from some level via an exception was
relatively slow, it frequently wasn't much (if any) slower than creating
a code to indicate the error, then exiting via multiple levels of normal
exits back to the point that the error could be handled.  The exact
difference in speed often depended upon exactly how many levels there
were between the exception and the handler.  If there were very few
levels, normal handling tended to be faster.  If there were many levels
between the two, exception handling tended to be faster.
    Later,
    Jerry.
... The Universe is a figment of its own imagination.
--- PPoint 1.90
---------------