Stephen Frost <sfrost@snowman.net> writes:
> * Tom Lane (tgl@sss.pgh.pa.us) wrote:
>> So are we at a consensus yet?
> You had me at "make public less special", I was just trying to make sure
> we all understand what that means.
> +1 from me for moving forward.
Applying this patch will leave us with the original pg_dump misbehavior
removed, AFAICS. So now we are hard up against the question of whether
--no-comments can support its weight as a pure feature, rather than a
workaround for a misbehavior/bug.
I've been pretty much on the negative side of that question, but
I just posted a patch here:
https://postgr.es/m/32668.1516848577@sss.pgh.pa.us
that will have the effect of causing pg_restore to print ACLs,
comments, and seclabels in cases where it formerly didn't.
If you want to get back to the old behavior in those cases,
we have you covered with --no-acl and --no-security-labels ...
but not so much for comments.
Between that and Robert's point that we have --no-foo for every
other subsidiary object property, so why not comments, I'm prepared
to change my vote.
This isn't a positive review of the contents of the patch,
because I've not read it. But I'm on board with the goal now.
regards, tom lane