Re: [HACKERS] Preliminary results for proposed new pgindent implementation - Mailing list pgsql-hackers

From Tom Lane
Subject Re: [HACKERS] Preliminary results for proposed new pgindent implementation
Date
Msg-id 30855.1497800534@sss.pgh.pa.us
Whole thread Raw
In response to Re: [HACKERS] Preliminary results for proposed new pgindentimplementation  (Peter Eisentraut <peter.eisentraut@2ndquadrant.com>)
List pgsql-hackers
Peter Eisentraut <peter.eisentraut@2ndquadrant.com> writes:
> On 6/16/17 10:51, Tom Lane wrote:
>> So I'm back to the position that we ought to stick the indent
>> code under src/tools/ in our main repo.  Is anyone really
>> seriously against that?

> I think it would be better to have it separate.

> Other than for reasons of principle and general modularity of the world,
> I would like this to be available separately for separate download,
> packaging, etc. to it can be applied to extension projects without
> having to download and build (a specific version of) PostgreSQL.  The
> code formatting in extension projects is, um, questionable.  In fact, if
> it's a better indent period, I would like to package it for the general
> public.

Well, the direction I'm headed in for addressing the portability issues
is to make it depend on the usual PG build environment, notably c.h and
libpgport.  If we don't want it in-tree, it can be built using PGXS,
but it'll still require a PG installation somewhere in order to get built.
Making it independent of both FreeBSD and PG is a significantly larger
project, and one I don't personally intend to tackle.  (And, if someone
does tackle that, I don't exactly see why having our own copy in-tree
would stop them.)

However ... off-list discussion with Piotr indicates that he's unwilling
to touch the license text without permission from FreeBSD core and/or
legal teams.  While the 4-clause license is certainly no impediment to
using indent, we don't want any such text in our tree, so that seems
like a showstopper, at least until the license question is resolved.

Accordingly, I'll proceed with setting up a repo for it on
git.postgresql.org.

> If the vote is to put it into the tree, I would request not to do it in
> PG10.  At this point, we should be winding things down and not open up
> new areas of activity.

I'm confused by this.  Are you objecting to switching to the new indent
version for v10?
        regards, tom lane



pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [HACKERS] initdb initalization failure for collation "ja_JP"
Next
From: Satyanarayana Narlapuram
Date:
Subject: Re: [HACKERS] Making server name part of the startup message