Path: utzoo!dciem!nrcaer!scs!spl1!lll-winken!lll-tis!ames!ncar!oddjob!uwvax!rutgers!columbia!read.columbia.edu!kearns
From: kearns@read.columbia.edu (Steve Kearns)
Newsgroups: comp.sys.mac.programmer
Subject: trap efficiency : explain?
Message-ID: <5673@columbia.edu>
Date: 7 Jun 88 15:38:56 GMT
Article-I.D.: columbia.5673
References: <7212@watdragon.waterloo.edu> <8796@dartvax.Dartmouth.EDU> <11713@Apple.Apple.COM>
Sender: nobody@columbia.edu
Reply-To: kearns@read.UUCP ()
Distribution: comp
Organization: Columbia University Department of Computer Science
Lines: 10

I have seen in several recent letters the suggestion 
"get the trap address..... and call it directly".
The implication here is that calling through a trap is inefficient.
How inefficient?  

(I do not understand microseconds, so please give me the time
 in multiples of a "normal" JSR)

 Thanks,
 -steve
 kearns@cs.columbia.edu