🚀 go-pugleaf

In article <41l688$csv@wormhole.cognos.com>,
Mark Ingram <ingramm@Cognos.COM> wrote re a proposal for a "leave"
statement:
>
>I recall a language very similar to C in which it was possible to
>label blocks, and break optionally took such a label.  Except it
>was "quit", not "break".  Also there was a combined if()quit structure
>which was called "quif".

I have used an interpreted database language with supported the syntax
(roughly):
	break [number of levels] [and execute this code as if it occurs
		directly after the (multilevel) break happens

It's a powerful feature -- but I doubt that its power justifies the cost
of the associated complexity.

Breaking out of a doubly-nexted loop (which is what this whole thread is
about) is clearly best solved with a goto, according to such authorities
as Dijkstra (author of the letter "GOTO Statement Considered Harmful"
which started this whole goto debate) and Donald Knuth (see "Literate
Programming", in which he discusses just this example).  On the other
hand, this thread *is* lending support to Dijkstra's claim that
programmers, and especially students, are not able to determine how
to use goto *correctly* :-).

-- Rich Miller (rmiller@wolfenet.com)