Message-ID: <gh-pgjdbc-pgjdbc-3837-c3361406249@github.com>
From: "m-van-tilburg (@m-van-tilburg)" <noreply+m-van-tilburg@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Thu, 02 Oct 2025 14:12:14 +0000
Subject: Re: [pgjdbc/pgjdbc] issue #3837: ResultSet.getTimestamp() returns incorrect point in time for historical dates
In-Reply-To: <gh-pgjdbc-pgjdbc-3837@github.com>
References: <gh-pgjdbc-pgjdbc-3837@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: m-van-tilburg
X-GitHub-Comment-Id: 3361406249
X-GitHub-Comment-Type: issue_comment
X-GitHub-Issue: 3837
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/issues/3837#issuecomment-3361406249
Content-Type: text/plain; charset=utf-8

> Interesting problem. How would you propose we solve how to figure out what calendar to use?

According to https://www.postgresql.org/docs/17/datetime-units-history.html all dates are treated as (proleptic) Gregorian. The JDBC API also suggests that provided `Calendar` objects are used to only get time zone information from. Code like `private final Calendar calendarWithUserTz = new GregorianCalendar();` in `org.postgresql.jdbc.TimestampUtils` appears to have been written in the assumption that this creates a pure (proleptic) Gregorian calendar. However, using `new GregorianCalendar()` actually creates a calendar that is a hybrid of the Julian and the Gregorian calendar using the default change date of October 15, 1582 (Gregorian). Read `java.util.GregorianCalendar#setGregorianChange(java.util.Date)` for more information.