yes very interesting. i didn't realise there was so much history behind
it. i came from tomcat so never knew about geronimo. i heard the name
all over the place of course
On 22/03/2018 03:04, David Blevins wrote:
On Mar 19, 2018, at 2:45 AM, Mark Struberg <strub...@yahoo.de.INVALID> wrote:
Let's face it TomEE is mostly an aggregator. A great one, I really love it -
but still.
[...]
Folks, you have to stop thinking as TomEE as being the center of the world. I
love TomEE and it's a great aggregator and a great community.
Everyone is free to have a perspective on what this project is and it's ok for it not to
be the same. It almost never is. It ebbs and it flows and that is natural. As long as
we're clear when we say "TomEE is..." we are expressing our own opinions, we're
ok.
People tend to think of the project in terms of when they came in. Their "OpenEJB is x"
or "TomEE is x" seems to reflect around the time they got commit.
I've put together a timeline of how I've experienced the project. I think
people should dream for more. It's the best part about open source and what
got us even this far.
If you look at this timeline you see all the growth periods are people deciding
that this is where they wanted to work and they we're the most flexible on what
that was. Even to the point of turning an EJB container into a Java EE
platform.
------
1999-2001
Project is born as an EJB library. Not quite a full EJB implementation. It is
envisioned as an EJB Container with the EJB Server part being implemented by a
full app server. The vision was to strategically not implement the server part
so true app servers would integrate the project as a library. Resources are
abundant, people are everywhere.
Many people on the project want to implement the server portion and make the
project more than it is. Myself and Richard tell them no. Speaking for
myself, I was a newbie idiot in this phase of the project.
2001-2003
Funding has dropped from the project, everyone from the original community but
Daniel, Jacek and Alan have gone. I see it now as a full EJB implementation
and perhaps a bit more. The Tomcat integration is born so Tomcat can have an
EJB implementation. The embedded EJB container for testing is heavily pushed.
The full remote protocol is created. The project is now bigger than its
original scope, but only slightly. Generally, there are no resources and not
many people around.
2003-2006
Geronimo is launched and the project is flooded with new people excited about Geronimo's
future. OpenEJB 1.0 is abandoned for OpenEJB 2.0 which is a total rewrite of EJB on the
Geronimo kernel and module system. Tomcat integration and embedded EJB concepts are
trashed. I still see this as a project that can live on its own and be more and
something I'd love to see grow in scope. Everyone on Geronimo, but me, sees it as a
library for Geronimo only. At one point I pull the remote EJB code from OpenEJB 1.0 into
OpenEJB 2.0 and people got quietly mad for bringing "legacy" code forward. The
mailing list is dead in these years with most discussion and decisions made on the
Geronimo list. The project is now significantly smaller than its original scope,
everyone is telling me to stop trying to make it more. There is a lot of fighting in
this time frame.
2006-2010
Work on OpenEJB 3.0 starts and this project regains technical freedom from
Geronimo. Dain Sundstrom leaves Geronimo in this timeframe, wants to make up
for killing OpenEJB 1.0 and puts his weight behind OpenEJB 3.0. OpenEJB 3.0 is
based on OpenEJB 1.0 and the work towards an embeddable EJB container and a
Tomcat integration continue where they left off. There was some discomfort,
skepticism and grumbling in the Geronimo community but largely ripping out the
old EJB container and putting in the new old EJB container was tolerated.
Enjoyed, no, tolerated, yes. It was enjoyed perhaps a bit later. The
embeddable container is a strong feature and brings new people into the project.
The project is bigger than any of the scopes it has had previously. Codebase
grows 5x from roughly 50k lines of code to 250k lines of code.
2010-2014
At this point "OpenEJB" is just shy of a full Java EE implementation and desire
to push it to the next level is high. The Tomcat-OpenEJB integration is pushed.
Momentarily called Tomtom, then finally called TomEE. Certification happens, the first
TomEE releases are made. The project is renamed and the website referencing OpenEJB is
changed to TomEE. The project enjoys amazing success. Things get busy fast. Geronimo
says nothing about TomEE competing it its space, is supportive and begins using some
additional TomEE/OpenEJB libraries like the jaxb tree. Xbean-finder is born originally
created in OpenEJB/TomEE and moved to Geronimo.
The project is bigger than any of the scopes it has had previously and the
charter is updated. Codebase grows 2x from roughly 250k to 520k lines of code.
2014-2018
User base continues to grow dramatically. Codebase growth slows and lines of
code go from roughly 520k to 580k. Many desire to see the project get back to
its glory state and continue growing. TomEE-Geronimo relationship appears to
be getting complicated again with perspectives being expressed that limit what
TomEE should be and indicate Geronimo is the proper place for those things to
live. There is a lot of fighting.
------
Everything is temporary. We can be anything we want tomorrow. Where we are
now feels a little bit like the 2003-2006 range. Not exactly, but similar. If
that's the case, fantastic, because there was some really good times afterwards.
-David
