Re: consider including server_version in explain(settings)

Justin Pryzby <pryzby(at)telsasoft(dot)com> writes:
> This is a "concept" patch to show the version, which is frequently requested on
> -performance list and other support requests. If someone sends
> explain(settings), they don't need to also (remember to) send the version..

I'm not really on board with the proposal at all here; I think it'll
be useless clutter most of the time. I do not agree with the position
that the only use-case for explain(settings) is performance trouble
reports. Moreover, if we start including fixed settings then where
do we stop? People might also want "pg_config" output for example,
and that's surely not reasonable to include in EXPLAIN.

Independently of that, however:

> /* skip GUC variables that match the built-in default */
> - if (!modified)
> + if (!modified && strcmp(conf->name, "server_version_num"))
> continue;

This is both horribly contorted logic (it could at least do with a
comment) and against project coding conventions (do not use the result
of strcmp() as if it were a boolean).

regards, tom lane