Bug #107053 for Net-DNS: useless statement

Sat Sep 12 06:18:44 2015 kes-kes [...] yandex.ru - Ticket created

Subject:

useless statement

Net::DNS::Domain:57 use constant LIBIDN => UTF8 && defined eval { require Net::LibIDN; }; can be rewriten as: use constant LIBIDN => UTF8; eval { require Net::LibIDN; }; may be you mean: defined eval { require Net::LibIDN; } && use constant LIBIDN => UTF8;

Sat Sep 12 18:06:43 2015 rwfranks [...] acm.org - Correspondence added

From:

rwfranks [...] acm.org

On Sat Sep 12 06:18:44 2015, kes-kes@yandex.ru wrote: Show quoted text

> Net::DNS::Domain:57 > > use constant LIBIDN => UTF8 && defined eval { require Net::LibIDN; }; > > can be rewriten as: > > use constant LIBIDN => UTF8; > eval { require Net::LibIDN; }; > > may be you mean: > > defined eval { require Net::LibIDN; } && use constant LIBIDN => UTF8;

This is not a beauty contest. Every version of every compiler from 4.4.4 to 5.23.2 understands the meaning of this code. Every ugly feature got there for a reason.

Sat Sep 12 18:06:44 2015 The RT System itself - Status changed from 'new' to 'open'

Sun Sep 13 13:18:18 2015 SREZIC [...] cpan.org - Correspondence added

On 2015-09-12 06:18:44, kes-kes@yandex.ru wrote: Show quoted text

> Net::DNS::Domain:57 > > use constant LIBIDN => UTF8 && defined eval { require Net::LibIDN; }; > > can be rewriten as: > > use constant LIBIDN => UTF8; > eval { require Net::LibIDN; }; > > may be you mean: > > defined eval { require Net::LibIDN; } && use constant LIBIDN => UTF8;

This does not work, because you try to do a compile-time statement depending on a runtime condition. Also, it's inefficient if UTF8 is false; in this case the code would needlessly load Net::LibIDN. But the current code in Net::DNS::Domain actually works, I think.

Mon Sep 14 16:07:05 2015 kes-kes [...] yandex.ru - Correspondence added

From:

kes-kes [...] yandex.ru

Show quoted text

> > use constant LIBIDN => UTF8 && defined eval { require Net::LibIDN; };

Sorry, I lose priority. That mean: use constant LIBIDN => ( UTF8 && defined eval { require Net::LibIDN; } ); But I think: ( use constant LIBIDN => UTF8 ) && ( defined eval { require Net::LibIDN; } );

Tue Sep 15 16:38:14 2015 rwfranks [...] acm.org - Correspondence added

From:

rwfranks [...] acm.org

On Mon Sep 14 16:07:05 2015, kes-kes@yandex.ru wrote: Show quoted text

> > > use constant LIBIDN => UTF8 && defined eval { require Net::LibIDN; > > > };

> > Sorry, I lose priority. That mean: > > use constant LIBIDN => ( UTF8 && defined eval { require Net::LibIDN; } > ); > > But I think: > ( use constant LIBIDN => UTF8 ) && ( defined eval { require > Net::LibIDN; } );

You are arguing the unarguable! 1) The code is good and there are at least 40000 successful test results in CPANtesters.org which serve to underline the point. 2) I wrote the code, understand very well how it works, and will fix any faults if there ever are any. 3) Unless you can write a test script that demonstrates a bug, there is nothing more to be said.

Wed Oct 07 04:28:27 2015 NLNETLABS [...] cpan.org - Status changed from 'open' to 'rejected'