| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | Andrew Dunstan <andrew(at)dunslane(dot)net>, pgsql-committers(at)lists(dot)postgresql(dot)org |
| Subject: | Re: pgsql: Close stdin where it's not needed in TestLib.pm procedures |
| Date: | 2019-11-26 08:47:02 |
| Message-ID: | 20191126084702.GF5435@paquier.xyz |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-committers |
On Tue, Nov 26, 2019 at 02:13:50AM -0500, Tom Lane wrote:
> Perhaps the "where possible" caveat needs to include a test whether
> IO::Pty is installed? It's evidently not there by default everywhere.
>
> It's possible that we should just move the goalposts and say IO::Pty
> is required for TAP testing. I can foresee that we'll have to do
> that if we ever want automated tests of psql's tab completion,
> for example.
Hmm. I am not sure that we are at that point yet (in which case we'd
need to check after it in configure). Any other modules we require to
be installed are directly imported to our code, but here we visibly
have a dependency forced by the loaded code. My point is that it does
not impact platforms that do not need it, so it seems like a waste to
require for everybody something which is actually necessary only for a
portion of users running the TAP tests.
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Peter Eisentraut | 2019-11-26 08:51:22 | Re: pgsql: Close stdin where it's not needed in TestLib.pm procedures |
| Previous Message | Tom Lane | 2019-11-26 07:13:50 | Re: pgsql: Close stdin where it's not needed in TestLib.pm procedures |