Skip Menu |

This queue is for tickets about the Gtk2-Ex-NumAxis CPAN distribution.

Report information
The Basics
Id: 99342
Status: resolved
Priority: 0/
Queue: Gtk2-Ex-NumAxis

People
Owner: Nobody in particular
Requestors: JHI [...] cpan.org
Cc:
AdminCc:

Bug Information
Severity: Important
Broken in: (no value)
Fixed in: (no value)



Subject: upcoming problem with round()
http://www.cpantesters.org/cpan/report/099ad582-4b64-11e4-b9f8-7b352c6f0924 Forewarning about Perl 5.22 (due out spring 2015). The good news is that the math APIs of the POSIX module will be expanded, one addition will be round(). (To complement the ceil() and floor, all definitions taken as from the C99 standard.) The bad news is that by default the POSIX module exports *everything*. Including the round(). If someone uses both Math::Round *and* POSIX *and* they do the export-all (which cannot be changed, because of backward compatibility), there will be a naming conflict and crash in module startup. Suggested cure: instead of exporting everything from the POSIX module, export nothing, and explicitly prefix the uses of POSIX interfaces, e.g. use POSIX (); # The () is the explicit nothing. ... .... POSIX::whatever() ...
CC: bug-Gtk2-Ex-NumAxis [...] rt.cpan.org
Subject: Re: [rt.cpan.org #99342] upcoming problem with round()
Date: Wed, 08 Oct 2014 18:07:36 +1100
To: JHI [...] cpan.org
From: Kevin Ryde <user42_kevin [...] yahoo.com.au>
"Jarkko_Hietaniemi via RT" <bug-Gtk2-Ex-NumAxis@rt.cpan.org> writes: Show quoted text
> > use POSIX (); # The () is the explicit nothing. > ... > .... POSIX::whatever() ...
I think I already do that, so for my code it would just be to ask for a new enough Math::Round. (I see I import Math::Round's defaults without using them though.)
Math::Round 0.07 has fixed this, I believe.