Mailing-List: contact dev-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@jackrabbit.apache.org
Received-SPF: pass (nike.apache.org: domain of jukka.zitting@gmail.com
 designates 209.85.220.216 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :content-type:content-transfer-encoding;
        b=WxTvFQdpwHywmEAzrNWnWAFgE6I1tblSdja1uh+WFUEMn/NfPKoOUERIx9M2m94DWz
         iOCumyi4tpKmcnsMHwTIMwzWNdKsu+qmjBMON9PUjb30tFv5HCjPpVoahU8vcWjpGdCW
         fp5wVTaeqXr8mIh3KaSknq628ClK0toNtkXeM=
MIME-Version: 1.0
In-Reply-To: <7A5837CEF4D4834E97721F3BCC46D7F71BC513DA52@darth-malak.gx.local>
References: <7A5837CEF4D4834E97721F3BCC46D7F71BC513DA52@darth-malak.gx.local>
From: Jukka Zitting <jukka.zitting@gmail.com>
Date: Mon, 11 May 2009 09:00:23 +0200
Message-ID: <510143ac0905110000h2b65efddr1218211277e3c7ba@mail.gmail.com>
Subject: Re: Lucene 2.4.1 and jackrabbit-core-1.5.4
To: dev@jackrabbit.apache.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Hi,

On Fri, May 8, 2009 at 8:00 AM, Martijn Hendriks <martijnh@gx.nl> wrote:
> I've been looking at merging the patch for issue JCR-1855 to jackrabbit-core-1.5.4 to
> get rid of a lucene memory leak, but for me that's not straightforward. I noticed,
> however, that the core compiles and builds (including all tests) if you only change
> the lucene dependency to 2.4.1. The patch of JCR-1855 contains a lot of code that
> replaces depercated method calls. Is that all that happens? Is it safe to only upgrade
> the version number and let the jackrabbit query code as it is?

Marcel would know best, but he's currently otherwise occupied.

AFAIK the main problem with upgrading to Lucene 2.4.x was the bug in
2.4.0. The upgrade to 2.4.1 in revision 756442 seems to have been all
that was needed, and as you notice, revision 756444 looks more like
related cleanups.

The query functionality is reasonably well covered by the test cases,
so if your backport passes all the tests then it's reasonably safe to
use as-is.

BR,

Jukka Zitting