Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!columbia!heathcliff.columbia.edu!metzger
From: metzger@heathcliff.columbia.edu (Perry Metzger)
Newsgroups: net.lang.mod2
Subject: Re: SetJmp implemented in M2 (Really: Why no exception handlers)
Message-ID: <2859@columbia.UUCP>
Date: Tue, 5-Aug-86 12:22:49 EDT
Article-I.D.: columbia.2859
Posted: Tue Aug  5 12:22:49 1986
Date-Received: Tue, 5-Aug-86 22:51:31 EDT
References: <8607310607.AA28309@bugs.ARPA>
Sender: nobody@columbia.UUCP
Reply-To: metzger@heathcliff.UUCP (Perry Metzger)
Organization: Columbia University CS Department
Lines: 17


 Hello, all.

The posting of a SetJmp-LongJmp brings to mind a question I have had for a
long time. Why is it that when he designed Modula-2, Wirth explicitly
left out any exception handler mechanism like SetJmp? I have heard that he
hates exception handlers. Why is this? They seem at times to be extremely
useful. Does anyone know why he thinks so poorly of such mechanisms? And
if he does have a good reason for disliking them, should we be encouraging
non-portability by using privately created versions?

.pm

Perry Metzger

Arpanet: metzger@heathcliff.columbia.edu
UUCP: ...![seismo, topaz]!columbia!heathcliff!metzger