Re: Multiple returns vs. Single exit

"Mann, Ivan H" <[email protected]> Fri, 28 Dec 2007 15:30:40 -0600
Newsgroups gmane.comp.windows.devel.java.advanced
Message-ID <3E47D20FD1918B48AC434AA83B57529601BA0CC6@usplm215.amer.corp.eds.com>
 
One more comment:

Way back when, when Djikstra first wrote that gotos were considered
harmful a debate started about coding style, pragmatics, etc.  The
reason standards like "single exit" are created is because programmers
who are not real good can make programs that don't work in spectacular
ways and that one problem the person debugging has is knowing how
control exited from this subroutine.

Making it easy to find the exits (by having only one) does nothing to
clarify the program.  A bad programmer can still fail in spectacular
ways, but now his program will not have that one problem (it will
probably be worse).

A much better approach would be to hire (and pay the price for) expert
programmers.  Another approach would be training to improve less good
programmers.  Obviously some managers feel it easier to produce edicts
than to pay for good programmers.  There are innumerable Dilbert
cartoons on this subject.

Back to the original question: it's sort of lame and not a substitute
for good programming.  They don't understand that.

Ivan Mann

===================================
This list is hosted by DevelopMentorĀ®  http://www.develop.com

View archives and manage your subscription(s) at http://discuss.develop.com