Re: Mark all GUC variable as PGDLLIMPORT

On Mon, Feb 14, 2022 at 10:38 AM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> > I don't particularly like Chapman's solution, but given that you've
> > repeatedly blocked every effort to just apply PGDLLIMPORT markings
> > across the board, I'm not sure what the realistic alternative is.
>
> You do realize that I just have one vote in these matters? If I'm
> outvoted then so be it. The impression I have though is that a
> number of other people don't like the extra notational cruft either.

Hmm, I guess I'd need to know who those people are in order to be able
to review their comments. I don't *like* the extra notational cruft,
but applying it inconsistently isn't better than being consistent. As
I see it, we have four choices: (1) apply PGDLLIMPORT markings
relatively broadly so that people can get extensions to work on
Windows, (2) continue to apply them inconsistently, thus slightly
reducing notational clutter at the cost of breaking lots of extensions
on Windows, (3) put some complex system in place like what Chapman
proposes and get all extension authors to adopt it, and (4) remove the
Windows port. To the best of my current knowledge, everyone other than
you prefers (1), you prefer (2) or (4), and (3) is an attempt at
compromise that is nobody's first choice.

If that is correct, then I think we should do (1). If it's incorrect
then I think we should do our best to find a choice other than (2)
that does attract a consensus. The current situation, which is (2),
must be the worst of all possible options because it manages to bother
the people who dislike the clutter AND ALSO bother the people who want
to have their extensions work on Windows. Any other choice has a
chance of reaching a state where only one of those groups of people is
annoyed.

--
Robert Haas
EDB: http://www.enterprisedb.com