Re: my $@ - a proposal

[Abigail <abigail@abigail.be>]

On Mon, Jun 30, 2008 at 09:52:33AM +1200, Sam Vilain wrote:
> Moritz Lenz wrote:
> 
> > Warnings should be reserved for stuff that is very likely an error, not
> > for stuff where some cases might be errors.
> 
> Yes, and relying on $@ after a scope destruction is possibly an error,
> because it can unexpectedly change.

It is *POSSIBLY* an error. Not likely. In the majority of the cases, the
current usage of $@ does not a problem at all. When I asked how often people
encountered problems, people reported numbers than can be counted on one
hand. 

IMO, that means adding a warning causes more problems than it fixes. People
will find themselves with an extra hurdle when the time is ready to upgrade
to 5.12 - code that has worked fine will suddenly start emitting warnings.



Abigail

---

Follow-Ups from:

Aristotle Pagaltzis <pagaltzis@gmx.de>
Mark Mielke <mark@mark.mielke.cc>

---

[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index][Thread Index][Top&Search][Original]

---