Thursday October 27 1994 07:41, Murray Lesser wrote to Jonathan Markevich:

 >>
 JM>> Are there still some GOTO-lovers?  I thought they all died off!
 >>

 ML>    The logical fallacies in Prof Dijkstra's 1965 "Goto is Harmful"
 ML> papers live on in several generations of programmers trained by
 ML> computer-science profs who didn't read those papers carefully enough, so
 ML> accepted the assertions in them.  Consequently, those programmers never
 ML> learned how and when to use a GOTO.  However, there are some of us, too
 ML> old to have fallen for the "structured programming" nonsense, still
 ML> alive.

This is sadly true.  Proper use of a GOTO can streamline code and actually
make it easier to maintain.  However, it is a statement that has been
widely abused, hence the reasons that companies frown upon the use of them.
 I'd like to see someone try to write a good old BAL program without using
a GOTO equilivent, (IE: B, BC, etc..)  For one thing, it would be
impossible, since a simple DTFCD generates an unconditional branch for the
EOFADDR parameter.

Now, in the case of the COBOL ALTER statement, I am glad to see it gone
from coding practices.  That made maintaining code a nightmare! :)

--- GoldED/2 2.42.G0615