From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andrey Rachitskiy <pl0h0yp1@gmail.com>
cc: andrew@dunslane.net, pgsql-hackers@lists.postgresql.org
Subject: Re: Residual cleanups for tied objects in PL/Perl
In-reply-to: 
 <CAB8bMivHNm5ySE_krD7KfKsCzA_7FpDppwypf0bz5oeQ+VKz7Q@mail.gmail.com>
References: <569769.1786901901@sss.pgh.pa.us>
 <CAB8bMivzt9CGJWM--i3V5MUSXL2FSK2b1DPiJYow2rd9y6pu9A@mail.gmail.com>
 <579928.1786912017@sss.pgh.pa.us>
 <CAB8bMisdLJVgRt5nvUfU=ivCG6V778x225m93fwKXcq2zFvA-w@mail.gmail.com>
 <CAB8bMisM-qSPw2qVnyjTp862LVAKpfX33_y-H1MBGzjAfwAQcg@mail.gmail.com>
 <CAB8bMivZFmWh9vvkF5HBT+Cy9y=nA2NbGhxQapGPUaZ-mNerSA@mail.gmail.com>
 <756684.1786981296@sss.pgh.pa.us> <1132241.1786999322@sss.pgh.pa.us>
 <CAB8bMivTgGQrLzvRTf3q7FdyZQk_onxL9sBbmhpGvMGaGFEDBw@mail.gmail.com>
 <CAB8bMivHNm5ySE_krD7KfKsCzA_7FpDppwypf0bz5oeQ+VKz7Q@mail.gmail.com>
Comments: In-reply-to Andrey Rachitskiy <pl0h0yp1@gmail.com>
	message dated "Tue, 18 Aug 2026 16:58:08 +0500"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1674509.1787090283.1@sss.pgh.pa.us>
Date: Tue, 18 Aug 2026 17:58:03 -0400
Message-ID: <1674510.1787090283@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1674510.1787090283%40sss.pgh.pa.us>
Precedence: bulk

Andrey Rachitskiy <pl0h0yp1@gmail.com> writes:
>>> Here's a patch to try to clean that up.  I think it is wrong that
>>> the extension modules don't use hek2cstr, so I made them do so.
>>> (But we probably shouldn't back-patch that: it's a user-visible
>>> behavioral change and we've not gotten actual field complaints AFAIR.)

>> Agreed
>> path looks good  for me

OK, applied that to HEAD only.  I've been dithering about how much
behavioral change we want to put into the back branches, and have
gotten cold feet about changing anything we don't clearly have to;
particularly in v14 whose next release will be its last.  So what
I've done is to back-patch a fix for just the null-pointer dereference
problems (see c0641ae4e and siblings).  The hek2cstr change and
hash-usage consistency fixes went into HEAD only, and I'm intending
the same for the fixes to allow fetching from tied arrays/hashes.
I'd possibly be more excited about treating those as back-patch
material, if we'd had more than zero complaints about plperl not
working with tied objects.  But we haven't AFAIR, and there might
even be someone who's surprised if it starts to work.

I plan to study your patch more closely tomorrow.

			regards, tom lane