Re: Improving connection scalability: GetSnapshotData()

Hi,

On 2020-09-15 11:56:24 +0900, Michael Paquier wrote:
> On Mon, Sep 14, 2020 at 05:42:51PM -0700, Andres Freund wrote:
> > My test uses IPC::Run - although I'm indirectly 'use'ing, which I guess
> > isn't pretty. Just as 013_crash_restart.pl already did (even before
> > psql/t/010_tab_completion.pl). I am mostly wondering whether we could
> > avoid copying the utility functions into multiple test files...
> >
> > Does IO::Pty work on windows? Given that currently the test doesn't use
> > a pty and that there's no benefit I can see in requiring one, I'm a bit
> > hesitant to go there?
>
> Per https://metacpan.org/pod/IO::Tty:
> "Windows is now supported, but ONLY under the Cygwin environment, see
> http://sources.redhat.com/cygwin/."
>
> So I would suggest to make stuff a soft dependency (as Tom is
> hinting?), and not worry about Windows specifically. It is not like
> what we are dealing with here is specific to Windows anyway, so you
> would have already sufficient coverage. I would not mind if any
> refactoring is done later, once we know that the proposed test is
> stable in the buildfarm as we would get a better image of what part of
> the facility overlaps across multiple tests.

I'm confused - the test as posted should work on windows, and we already
do this in an existing test (src/test/recovery/t/013_crash_restart.pl). What's
the point in adding a platforms specific dependency here?

Greetings,

Andres Freund