Re: Disable alternate locations on Win32 - Mailing list pgsql-patches

From Tom Lane
Subject Re: Disable alternate locations on Win32
Date
Msg-id 24566.1052280716@sss.pgh.pa.us
Whole thread Raw
In response to Re: Disable alternate locations on Win32  (Bruce Momjian <pgman@candle.pha.pa.us>)
List pgsql-patches
Bruce Momjian <pgman@candle.pha.pa.us> writes:
> Tom Lane wrote:
>> Why not do it with one ifdef in one place?

> I checked this again and the test is at the top, while the symlink()
> call is at the bottom after the database has already been copied.  I
> think we do need to error out before we start copying the database.

Huh?  We haven't copied anything at that point.  We have done the
mkdir() though.  On spec-compliant platforms it'd be possible to
switch the order of these operations, so that the symlink comes before
mkdir.  Does anyone know of a platform on which symlink() fails if the
referenced pathname doesn't exist yet?

            regards, tom lane


pgsql-patches by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: Disable alternate locations on Win32
Next
From: "Matthew T. O'Connor"
Date:
Subject: pg_autovacuum update