Tom Lane <tgl@sss.pgh.pa.us> writes:
> Just to point out one concrete problem: the postmaster reads
> postgresql.conf too, so it would have to do this as well in order to
> parse postgresql.conf correctly.
Well, if I parse you correctly, in my patch, it does. There's another
GUC array where to store invalid placeholders so that we can revisit
later. That even allows custom_variable_classes not to be parsed first
in the list, even if that's not put into use in the patch.
> Please remove it. If you think of a non-broken way to do this later,
> we can revisit the problem then.
Sure. I'm curious about understanding how it's broken and if there's
any will at all to have something better (and what better would be), but
it's now obvious that it's not going to help the extension patch.
Regards,
--
Dimitri Fontaine
http://2ndQuadrant.fr PostgreSQL : Expertise, Formation et Support