e.g. if we add some funky AOP stuff or support for deploying POJOs, JavaBeans, Avalon & PicoContainer components and so forth - which is outside of the spec but useful to different users. If this were included in the RI then folks might get confused and think these features are in the J2EE spec.
It seems a common issue with RIs that 'extra stuff' is often not allowed and for good reason.
So maybe (if Sun are interested) the reference implementation could be *based on* Geronimo. i.e. Sun can cherry pick all the bits they want from Geronimo (and contribute to it of course too :) but they get to choose what goes into Geronimo, how its configured & what options are enabled out of the box.
That way Sun stays in full control over what the J2EE RI is and the Geronimo community decides on what features we implement in Geronimo (providing the features we add don't break certification). Then we'd get the best of both worlds.
The best first step would be for Sun to allow Geronimo to the be reference implementation of the J2EE APIs - then we can all freely use the J2EE APIs without all this silly by-hand-typing-in-of-APIs-from-documentation.
On Friday, August 8, 2003, at 10:11 am, Adrian Jackson wrote:
Surely the goal should be to create a *good* J2EE implementation, rather
than one that necessarily falls within the constraints of what Sun would
want from a reference implementation? I'd guess that more people use
Tomcat because it's good at what it does, not because it's the RI.
If a good implementation ends up being something that Sun would like to use as the RI, then all well and good, but I'd rather see us building something that the admins/developers here would like to use for themselves than something for Sun.
Adrian
-----Original Message----- From: Greg Stein [mailto:[EMAIL PROTECTED] Sent: 08 August 2003 08:43 To: [EMAIL PROTECTED] Subject: Re: J2EE reference implementation
On Fri, Aug 08, 2003 at 02:54:57AM -0400, Henri Yandell wrote:b) Wants to be the J2EE RI.
That's a good question for this community, along with the ramifications of what that might entail.
James ------- http://radio.weblogs.com/0112098/
