Mailing-List: contact derby-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Derby Development" <derby-dev@db.apache.org>
Received-SPF: pass (hermes.apache.org: local policy)
Message-ID: <41EEB836.60506@Golux.Com>
Date: Wed, 19 Jan 2005 11:42:46 -0800
From: myrna <myrna@Golux.Com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
 rv:1.7.1) Gecko/20040707
MIME-Version: 1.0
To: Derby Development <derby-dev@db.apache.org>
Subject: Re: [jira] Commented: (DERBY-95) NPE when passing in url instead
 of database name to EmbeddedXADataSource
References: <1688777574.1105615579485.JavaMail.jira@ajax.apache.org>
 <41EB75A9.2070302@Sun.com> <41EC827C.1000805@Golux.Com>
 <41EEB267.1070804@debrunners.com>
In-Reply-To: <41EEB267.1070804@debrunners.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Daniel John Debrunner wrote:

>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>myrna wrote:
>
>  
>
>>It does no longer throw an NPE, and it passes all tests....
>>
>>Could we leave EmbeddedDataSource alone & only change
>>EmbeddedXADataSource? Only with the xa datasource did we get a NPE...
>>Although strictly speaking, it should behave the same, so, with
>>EmbeddedDataSource too, we should accept the databasename only, it's
>>been able to resolve the url fine. I'd be apprehensive about any
>>possible cloudscape customers that have mistakenly used this
>>functionality...
>>    
>>
>
>I agree with Shreyas that his patch is logically correct but I want to
>ensure that your concerns are addressed. Can you be explicit about what
>you believe the patch breaks for existing users of Derby. I tried this
>and it results in a NullPointerException if the return from
>getConnection() is referenced, e.g. in the conn.close() here.
>(this is before the patch is applied)
>
>org.apache.derby.jdbc.EmbeddedDataSource ds = new
>org.apache.derby.jdbc.EmbeddedDataSource();
>
>ds.setDatabaseName("jdbc:derby:fred");
>
>Connection conn = ds.getConnection();
>
>// NPE here because conn is null
>conn.close();
>
>
>The only issue I have with the patch is that the error thrown is
>misleading, "Database Not Available" implies to the user or application
>developer that they need to do something to make the database available.
>A database not found exception with the name obtained from
>DataSource.getDatabaseName() might make more sense. Or maybe even
>better, invalid property setting error, since databaseName is a Java
>Bean property of the datasource. E.g. SQLState.PROPERTY_INVALID_VALUE,
>resulting in an error like
>
>Invalid value for property 'databaseName'='jdbc:derby:fred'.
>
>Thanks,
>Dan.
>-----BEGIN PGP SIGNATURE-----
>Version: GnuPG v1.2.5 (MingW32)
>Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
>iD8DBQFB7rJnIv0S4qsbfuQRApTGAKDj+MxB4kOY/GNFCMs/9R5Js3zK2wCgqYSm
>DrAWFRgPOfwMHin0vUwGi70=
>=zL6e
>-----END PGP SIGNATURE-----
>
>  
>
I see your point. My initial test case was overly simplistic - never 
actually did anything with the connection.

Myrna