On Tue, 19 Dec 2023 at 16:52, Nathan Bossart <nathandbossart@gmail.com> wrote:
> I'm not sure we should proceed with rewriting most/all eligible foreach
> loops. I think it's fine to use the new macros in new code or to update
> existing loops in passing when changing nearby code, but rewriting
> everything likely just introduces back-patching pain in return for little
> discernible gain.
To clarify: I totally agree that if we're not backpatching this we
shouldn't do bulk changes on existing loops to avoid pain when
backpatching other patches.
> Unless there's some way to argue this is a bug, security issue, or data
> corruption problem [0], I seriously doubt we will back-patch this.
In the past some tooling changes have been backpatched, e.g.
isolationtester has received various updates over the years (I know
because this broke Citus its isolationtester tests a few times because
the output files changed slightly). In some sense this patch could be
considered tooling too. Again: not saying we should back-patch this,
but we could only realistically bulk update loops if we do.