Newsgroups: comp.object,comp.lang.c++,comp.lang.ada,comp.lang.clos
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!gatech!news.mathworks.com!uhog.mit.edu!news.mtholyoke.edu!world!bobduff
From: bobduff@world.std.com (Robert A Duff)
Subject: Re: C++ not OOP? (Was: Language Efficiency
Message-ID: <D8vr59.HKn@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
References: <dewar.797512974@gnat> <dewar.800927060@gnat> <3pjio9$b4u@disunms.epfl.ch> <dewar.800941444@gnat>
Date: Sat, 20 May 1995 14:05:33 GMT
Lines: 27
Xref: glinda.oz.cs.cmu.edu comp.object:31206 comp.lang.c++:129215 comp.lang.ada:30137 comp.lang.clos:2972

In article <dewar.800941444@gnat>, Robert Dewar <dewar@cs.nyu.edu> wrote:
>Hmm! There certainly is not such restriction in the GNAT fix and continue,
>and, knowing the way F&C works, I can't see why there would be any
>such restriction in the C++ version either. Basically what you can do
>is any change you like to a subprogram body, but you can't change specs.

If you add, say, "X: Integer := ...;", and add a statement after the
current point of execution, is X initialized?

How does it keep track of where you are, especially when you do
something like delete the statement you're currently executing in?

Are there documents available on what it does, and how it works?
Where?

- Bob

P.S. I agree that "fix and continue" can be useful -- I've used it in
Lisp environments before.  But there were some circumstances where it
got confusing, and I had to start the program over to understand what
was going on.  E.g. suppose a function runs 3 times, and then blows up
the third time.  If you change the function to be correct, it still
might not work to "continue" it, because whatever it was supposed to do
the first two times didn't get done.  I found it hard to keep track of
that sort of complication in me head -- sometimes.

- Bob
