Hey everyone,

Following is the p5p (Perl 5 Porters) mailing list summary for the past two weeks. Enjoy!

September 29th - October 10th

News and highlights

Grant reports

• Dave Mitchell TPF Grant #2 September 2016 report.

Issues

New issues

• Perl #129757: for loop doesn't update correct variable.
• Perl #129770: Null ptr deref, segfault Perl_mess_sv.
• Perl #129777: Caller return unexpected info for subs called from condition under while.
• Perl #129794: Debugger stops on use statement.
• Perl #129802: Regexp performance problem with swiped strings (?).
• Perl #129803: Regexp syntax check when variables contained with \Q...\E.
• Perl #129821: Bleadperl v5.25.5-65-g5012eeb breaks Devel::OptreeDiff.
• Perl #129825: Overloaded integer constants are prematurely deallocated.
• Perl #129826: Invalid read of size 8 in S_parse_body when argv ends with --.

Resolved issues

• Perl #116250: Skip count is wrong for timeout in dist/Tie-File/t/29_downcopy.t.
• Perl #117223: Remove IO::File example from perlfunc.
• Perl #125702: Garbage Collection Segv in 5.21.10+.

Rejected issues

• Perl #129368: Error while trying to install perl 5.8.7 on RHEL 6.

Suggested patches

Dave Mitchell merged Reini Urban's patches in Perl #125792 to fix sassign to be declared as BASEOP instead of BINOP and to be optimized.

Andy Lester provided a patch in Perl #129766 to cleanup numeric.c.

Sergey Aleynikov provided both a patch and a test in Perl #129788 to fix a possible memory leak in IO::Poll.

Sergey also provided a patch in Perl #129789 to fix Time::HiRes in blead on Darwin.

Discussion

Jakko Hietaniemi is working on fixing building perl in a vax-bsd box. He hit a problem with POSIX, which was eventually resolved.

Todd Rinaldo asked about the struct definition of a hash key in hv.h, which raises interesting confusing points in C.

Klemen Markovic asked about Windows W APIs which were replaced in favor of Windows A APIs. An exhaustive research by Steve Hay into the history of it showed that it never really worked properly with the Unicode changes in 5.8 and was therefore removed.

Dave Mitchell asked ((\@a) = \($x,$y) in non-void context) about a current behavior of the refaliasing feature, which Father Chrysostomos sees as a bug.

Andy Lester asked (Tabs and spaces in patches) whether he should provide a patch with spaces or tabs and what is desired now. Karl Williamson notes that we are moving the code to spaces and no tabs.

Sam Kington suggested extending the fat comma syntax to handle situations of undefined or empty list variables on the right-hand-side.