Tom,
* Tom Lane (tgl@sss.pgh.pa.us) wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
> > On Tue, May 30, 2017 at 8:55 PM, David G. Johnston
> > <david.g.johnston@gmail.com> wrote:
> >>> Having --no-comments seems generally useful to me, in any case.
>
> >> It smacks of being excessive to me.
>
> > It sounds perfectly sensible to me. It's not exactly an elegant
> > solution to the original problem, but it's a reasonable switch on its
> > own merits.
>
> I dunno. What's the actual use-case, other than as a bad workaround
> to a problem we should fix a different way?
Perhaps it's a bit of a stretch, I'll admit, but certainly "minmization"
and "obfuscation" come to mind, which are often done in other fields and
might well apply in very specific cases to PG schemas. I can certainly
also see a case being made that you'd like to extract a schema-only dump
which doesn't include any comments because the comments have information
that you'd rather not share publicly, while the schema itself is fine to
share. Again, a bit of a stretch, but not unreasonable.
Otherwise, well, for my 2c anyway, feels like it's simply fleshing out
the options which correspond to the different components of an object.
We provide similar for ACLs, security labels, and tablespace
association. If there are other components of an object which we should
consider adding an option to exclude, I'm all ears, personally
(indexes?). Also, with the changes that I've made to pg_dump, I'm
hopeful that such options will end up requiring a very minor amount of
code to implement. There's more work to be done in that area too,
certainly, but I do feel like it's better than it was.
I definitely would like to see more flexibility in this area in general.
Thanks!
Stephen
