Message-ID: <gh-pgjdbc-pgjdbc-3845-c3431203957@github.com>
From: "victornoel (@victornoel)" <noreply+victornoel@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Wed, 22 Oct 2025 08:57:33 +0000
Subject: Re: [pgjdbc/pgjdbc] issue #3845: Wrong result on equality between Date and Timestamp given as query parameter
In-Reply-To: <gh-pgjdbc-pgjdbc-3845@github.com>
References: <gh-pgjdbc-pgjdbc-3845@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: victornoel
X-GitHub-Comment-Id: 3431203957
X-GitHub-Comment-Type: issue_comment
X-GitHub-Edited-At: 2025-10-22T08:58:03Z
X-GitHub-Issue: 3845
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/issues/3845#issuecomment-3431203957
Content-Type: text/plain; charset=utf-8

@vlsi 

> What do you mean by "properly"?

I guess I mean that when you look at the implementation of `PgPreparedStatement#setDate(int, java.sql.Date, java.util.Calendar)` or `PgPreparedStatement#setTimestamp(int, java.sql.Timestamp, java.util.Calendar)`, you can see it's using `Oid.UNSPECIFIED` most of the time to propagate the type to the server.

I just discovered this, I didn't know it when I wrote my first comment but I was intuiting this kind of thing.

> Frankly, I would suggest to use java.time types instead of java.sql types.

So you mean one should be doing `setObject(LocalDate.of(2013, 03, 21))` instead of `setDate(java.sql.Date.valueOf(LocalDate.of(2013, 03, 21)))` right?

It's indeed a good workaround right now, it's unfortunately not followed by libs like Hibernate that will directly use `setTimestamp()` or `setDate()` but I guess it's also workaroundable.

>  Is there a reason you have to use java.sql types without casts?

Is there a reason that we should be using casts? I mean: the only workaround currently is to do use casts, so that's what is going to be done, but we thought there was maybe something wrong going on a the driver level and decided to start a discussion here about it.