http://blogs.msdn.com/b/ericlippert/archive/2003/10/16/53232.aspxIn the dark days before exception handling you could always use setjmp/longjmp to do non-local gotos. In fact, the script engines are compiled in C++ with exception handling turned off and the mainline loop of the bytecode interpreter uses setjmp-longjmpen-USTelligent Evolution Platform Developer Build (Build: 5.6.50428.7875)http://blogs.msdn.com/b/ericlippert/archive/2003/10/16/53232.aspx#53236Fri, 17 Oct 2003 20:12:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:53236RJSorry if it appears I am being awkward, I'm not - I'm just interested in learning to be optimal :)<div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=53236" width="1" height="1">http://blogs.msdn.com/b/ericlippert/archive/2003/10/16/53232.aspx#53235Fri, 17 Oct 2003 18:29:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:53235Eric Lippert&gt; What was the measured cost when C++ exceptions were turned on? You're asking me for the numbers which we measured on Pentium 1's seven years ago? I haven't got them on me, sorry. But it was significant. &gt; It would be interesting to know *which* performance reasons these were The overall cost, not the per-exception cost, killed our performance on the server. Exception handling adds additional code to every function, and that additional code has a nonzero run time. Since the purpose of the code -- handling exceptions -- was NEVER going to be fulfilled becasue we never THREW an exception, we turned it off. <div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=53235" width="1" height="1">http://blogs.msdn.com/b/ericlippert/archive/2003/10/16/53232.aspx#53234Fri, 17 Oct 2003 11:30:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:53234RJ&quot;In fact, the script engines are compiled in C++ with exception handling turned off (for performance reasons)&quot; It would be interesting to know *which* performance reasons these were after your reply to my comment in &quot;I take exception to that&quot; - sorry for being facetious 8*) <div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=53234" width="1" height="1">http://blogs.msdn.com/b/ericlippert/archive/2003/10/16/53232.aspx#53233Fri, 17 Oct 2003 07:35:00 GMT91d46819-8472-40ad-a661-2c78acb4018c:53233MMWhat was the measured cost when C++ exceptions were turned on?<div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=53233" width="1" height="1">