status on 5.18.0

This weekend, I made the final release of 5.17, marking only the second time a
version of perl 5 has made it to patchlevel 11.  Exciting, nu?

Now it's 5.18.0 or bust.  I'm shooting for "not bust."  This means that some
"blockers" are likely to end up being called "not blocky enough" and listed in
the "Known Problems" section of perl5180delta.  Nobody really wants that!

I'm calling seven things in our list "real" blockers:

### "in cleanup" warnings generated from destructor
    https://rt.perl.org/rt3/Ticket/Display.html?id=113794

    Zefram has supplied a patch.  It breaks a test, which he says just needs
    "updating, similarly to t/op/die_keeperr.t"; he said that while he's
    crushed lately, it should be no problem for someone else to sort out.

### S_croak_memory_wrap breaks gcc warning flags detection
    https://rt.perl.org/rt3/Ticket/Display.html?id=116989

    This had been de-blocked, but we have reports that it's still failing on
    Solaris 11 with Sun CC.  Tony had a suggestion to which there has yet been
    no reply.

### Sequence (?#...) not recognized in regex
    https://rt.perl.org/rt3/Ticket/Display.html?id=117327

    Having been asked which of the two ways to resolve this bug, I said
    "preferably X, but I'll take either one for now!"  So far, neither fix is
    merged.

### Perl 5.16 regression: lost warning for -l on filehandle
    https://rt.perl.org/rt3/Ticket/Display.html?id=117595

    David Golden supplied a patch and said that someone more comfortable with
    the guts should give it a once over before applying it.  Still needs both
    those things.

### v5.17.6+ segfaulting on simple (buggy) code
    https://rt.perl.org/rt3/Ticket/Display.html?id=117607

    perl -e 'use strict; $anyvar; &CORE::lc;' # Boom!

    It's been bisected, just not fixed.

### Do a better job of escaping components of ldlibpth for the shell
    https://rt.perl.org/rt3/Ticket/Display.html?id=115708

    I think this just needs to be sanity-checked and applied.  I may well do
    that myself when it's not so close to bedtime.

### caller behaving unexpectedly in re-evals
    https://rt.perl.org/rt3/Ticket/Display.html?id=113928

    No progress.  This is drifting toward being a "Known Problem."

-- 
rjbs

• status on 5.18.0 by Ricardo Signes
• Re: status on 5.18.0 by Dave Mitchell
• Re: status on 5.18.0 by Ricardo Signes
• Re: status on 5.18.0 by H.Merijn Brand