PGXAConnection of Driver Version 9.3.1001 does not play nicely with Apache DBCP 1.4 about pgjdbc

[alexismeneses]

Hi,

It seems that this has already been fixed into the master (commit 8d74338 PR #125), at least for the equals() part.

Is there a point returning 0 on hashCode() when con is null?

Regards

[ctolkmit]

Not sure right now about the 0 if con is null. It's just a safety pattern I am used to type automatically ;)

[ctolkmit]

Oh, and yes, I saw the equals()-Part in the Master, however the hashCode() part is also important.
And I don't like branching from an unreleased version, and we are currently using my patched version in production.

[alexismeneses]

Unless the zero-if-null thing is useful for a case that might happen and for which we can write a test, I think the issue is fixed.

Letting hashCode() fall into method.invoke(con, args) will do the job as two proxies with the same underlying connection will return the same consistent value.

The only remaining question is: should 8d74338 be backported to older branches?

[davecramer]

Yes, it probably should be backported.

Dave Cramer

On 1 July 2014 06:48, Alexis Meneses [email protected] wrote:

Unless the zero-if-null thing is useful for a case that might happen
and for which we can write a test, I think the issue is fixed.

Letting hashCode() fall into method.invoke(con, args) will do the job as
two proxies with the same underlying connection will return the same
consistent value.

The only remaining question is: should 8d74338
8d74338
be backported to older branches?

—
Reply to this email directly or view it on GitHub
#162 (comment).

[davecramer]

historic, if this still exists in the latest driver please advise