[boost] Re: Invariants Library

From: Thorsten Ottosen (nesotto_at_[hidden])
Date: 2004-01-03 03:46:28

"Dan W." <danw_at_[hidden]> wrote in message
news:bt5sc7$qjd$1_at_sea.gmane.org...
> >>Precisely, the DBC
> >>'fuses' are only compiled in debug mode because you don't want to stop
> >>the application out there in the real world.
> > Often it is better to terminate() than letting a program continue on
> > undefined behavior; especially in
> > languages with unsafe pointers.
>
> That's what I've been saying all along. OFTEN, but not always. That's
> why you want to use exceptions, and have a *policy*. And preferably a
> run-time customizable policy.
>
> But I've also been saying, DBC has only one policy: to stop execution.
> And the intent for DBC is to help debug as well as document.  It is not
> meant to replace exceptions.
>
> Using exceptions for debugging sucks. Exceptions are good at dealing
> with exceptions in the field, not with bugs.
>
> DBC is for dealing with bugs, NOT with exceptions in the field.
>
> That's why I've been saying _please_forget_exceptions_ when talking
> about implementing DBC.  Just leave them alone.  Let's just talk DBC.
> Which is for dealing with bugs.

Other people disagree:

http://www.digitalmars.com/ctg/designbycontract.html

Quote:

"Currently, contracts rely on the assert() macro aborting the program with a
message. Perhaps this needs to be revisited, and changed to throw an
exception instead. "

http://archive.eiffel.com/doc/manuals/technology/contract/

Quote:

"The benefits of Design by Contract include the following: [...] A technique
for dealing with abnormal cases, leading to a safe and effective language
construct for exception handling."

You will also see that the invariant is kept in the short form.

br

Thorsten