On 6/11/14 9:32 AM, Lee Breisacher wrote:
Hi. I’m interested in having a DRDA server that invokes other
databases besides embedded-Derby. Perusing the Derby drda server-side
code, it appears to be somewhat bound to Derby. So I was thinking that
one approach would be to create a DRDA API that is not tied to Derby
(or any particular database), then refactor the Derby classes (from
package org.apache.derby.impl.drda) to cleanly separate the
Derby-specifics from the DRDA parsing.
So, I’m wondering what anyone thinks about this idea – how feasible,
is it worth bothering with, any other suggestions?
Thanks,
Lee
Hi Lee,
It's hard to say without seeing the proposed changes. It's possible that
a refactoring will make the server code easier to understand. But it's
also possible that it will introduce an extra level of indirection which
doesn't add any value for Derby.
Bear in mind that Derby does not make any guarantees about its
implementation of DRDA. Derby's network protocol was originally based on
DRDA, but the protocol has evolved to include extensions which aren't
part of any DRDA rev level. Adhering to DRDA is not part of Derby's charter.
You are welcome to create a JIRA and start discussing proposals for this
project.
Thanks,
-Rick