jsr343-experts@jms-spec.java.net

[jsr343-experts] Re: [jms-spec users] Re: Guidelines on backwards compatibility

From: Nigel Deakin <nigel.deakin_at_oracle.com>
Date: Mon, 31 Oct 2011 11:43:41 +0000

On 29/10/2011 00:51, John D. Ament wrote:
> The way I take it, based on the messages I've seen posted, is that there's a big difference between "JMS 2.0" and
> "JMS2". I guess like NIO, NIO2? I remember javaposse complaining about that approach on their podcast. I guess, can
> we make this "JMS2" at this point?

If you're referring to a package name, I think it's too early for that. So far the discussion has been hypothetical: I
wanted to make it clear that if we wanted to make incompatible changes to the API we could do so but would need to
define a new set of interfaces (in a new package) .

Since this would be a disruptive change I think we should only consider doing it if the new interfaces would offer
significant new features (as nio did) or a major improvement in ease of use. We've talked about a simplified API but
this is still at the "brainstorming" stage - something to consider if we can't make further progress in defining
dependency injection on top of the existing API

Nigel