Re: Trouble with amcheck

On 2017-09-14 22:36:38 -0400, Stephen Frost wrote:
> Andres,
>
> * Andres Freund (andres(at)anarazel(dot)de) wrote:
> > On 2017-09-15 02:22:49 +0000, Douglas Doole wrote:
> > > Thanks all. Making and installing the contribs got me rolling again. (I
> > > tried "make world" but ran into trouble with the XML docs. But that's pain
> > > and suffering for another day.)
> > >
> > > I'd agree that "make installcheck-world" should imply that all prereqs are
> > > met - that's certainsly the normal behaviour for make.
> >
> > I'm very unconvinced by this, given that one use of installcheck is to
> > run against an existing server. For which one might not even have access
> > to the relevant directories to install extensions into.
>
> Sure, but if the extensions aren't in place and you're trying to run
> make installcheck-world, it's not like it's somehow going to succeed.
>
> Failing earlier on the install seems like a reasonable thing to do
> rather than failing later halfway through the check process.

But, uh, aren't you now provoking errors because of non-existing rights
to install stuff on the system? I'm all for improving error messages
and/or adding extra checks, but this doesn't seem to be a solution.

> Now, that said, perhaps a bit more smarts would be in order here to,
> instead, check that the extensions are available before trying to run
> the checks for them. I'm thinking about something like this: check if
> the extension is available and, if not, skip the check of that module,
> with a warning or notification that it was skipped because it wasn't
> available.

I think that'd just lead to people not noticing that they're not
executing all tests. I'm ok with adding a hard error with a better
message or such tho.

Greetings,

Andres Freund