Re: Using LibPq in TAP tests via FFI - Mailing list pgsql-hackers

From Andres Freund
Subject Re: Using LibPq in TAP tests via FFI
Date
Msg-id 20240616232435.n4p3d6j6nbdnyiku@awork3.anarazel.de
Whole thread Raw
In response to Re: Using LibPq in TAP tests via FFI  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Hi,

On 2024-06-16 19:07:49 -0400, Tom Lane wrote:
> Thomas Munro <thomas.munro@gmail.com> writes:
> > On Mon, Jun 17, 2024 at 10:38 AM Andres Freund <andres@anarazel.de> wrote:
> >> I'm somewhat doubtful we'll just agree on making FFI::Platypus a hard
> >> dependency, but if we agree to do so...
> 
> > Why can't we just do that?  I mean, do we have any concrete reason to
> > think that it'll block a supported platform?
> 
> IIUC, this would only be a hard dependency if you want to run certain
> TAP tests (maybe eventually all of them).

I think it'd be all of them within a very short timeframe. IMO we'd want to
convert a bunch of the code in Cluster.pm to use psql-less connections to
maximize the benefit across all tests, without needing to modify all of them.

Greetings,

Andres Freund



pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Using LibPq in TAP tests via FFI
Next
From: David Rowley
Date:
Subject: Re: assertion failure at cost_memoize_rescan()