« Moshe Cohen is coming to FOO 2006! | Main | Java Attribution Timeline Edits »

November 13, 2006


Eduardo Pelegri-Llopart

Hi Danese. I think the timeline should at the least include these three more milestones:

* JavaOne'04 - JAX-RPC, JAXB, JSF available under JRL/JDL. Their TCKs are available too.
* JavaOne'05 - Java EE 5 RI available under CDDL.
* JavaOne'06 - GlassFish ships Java EE 5 fcs.

The success of GlassFish (internal and external) has been one of (many) internal contributors to making the rest of the Java platform available as OpenSource.

- eduard/o

eduardo Pelegri-Llopart

Ah! I looked more carefully and I see there is a reference to the J1'05 event. Thanks - eduard/o

Mark Wielaard

Some other interesting events in this stream:

- Kaffe - Nov 22 1996, Tim Wilkinson.
- GNU Classpath 0.00 - Feb 6, 1998, Geoff Berry, Jim Blair, Brian Jones, Paul Fisher, Aaron Renn and John Keiser
- The start of the GCJ project - June 30, 1998, Cygnus, now Red Hat, Per Bothner, Andrew Haley, Tom Tromey, Anthony Green.
- The merging of GNU Classpath and libgcj, plus the birth of the GNU Classpath exception - March 6, 2000, Anthony Green, Tom Tromey, Paul Fisher and Richard Stallman
- Start of Kaffe and Classpath cooperation at 14+15 Oct. 2003 Linux Kongress '03 in Saarbrücken, Germany, Dalibor Topic, Mark Wielaard, Sascha Brawer

And I will also have forgotten names and dates and projects. Apologies. More can be found at:
http://www.gnu.org/software/classpath/news.html and
http://www.gnu.org/software/classpath/events/events.html and

Ed Burns

Hi Danese,

Glad to see you still fighting the good fight! As others have done, I'd like to add another milestone to your timeline.

June 2004, Sun Open Sources the first part of any Java Edition: JavaServer Faces, which is part of the Java EE platform.



james governor

hey Danese. it was pretty amazing to have intel supporting sun on this one. is that your fingerprints i see in the dust? we were just talking about it in the redmonk irc channel, and wondered... one guy described it as eye-rubbing. indeed.

Danese Cooper


If by "Intel's support" you mean the fact that Geir and I were there? Heck, we're open source activists!! Sun is finally, finally doing something about the proprietary nature of Java, and that's gotta be a good thing. Besides, we're talking about nearly a decade of my life put into this cause, so you'd best *believe* I wanted to be there! I'm not sure that having us there constitues Intel support however, I would have wanted to be there no matter who was signing my paycheck :-).

Intel and IBM are supporters of Apache Harmony through direct code contributions. Although I'm sure Rich Green and Simon Phipps would publicly disagree, Harmony is clearly one of the catalysts for Sun's actions yesterday. I think Intel would have preferred that Sun donate its RI code to Harmony or that they license their project under the ASL. We are still trying to get them to dual-license javac under CDDL so that Harmony folks can use it more easily (feel free to help us with that one). Their decision to GPL the code, while inconsistent with their past involvement at Apache is however still a good thing. It means more choice for Java users, a possible boost for the GNUClasspath project (who are really nice people) and the possibility of more direct influence on the project from outside of Sun. Yeah!




Is it true that Graham has left Sun? (http://blogs.codehaus.org/people/geir/archives/001429_graham_too.html)


Per Bothner

Nice timeline, but GCJ actually goes back all the way to 1996. The first ChangeLog entry in gcc/java is 1996-10-24. The first ChangeLog entry in the libjava is from 1998, but it clearly describes changes to an existing tree; I don't know where the older versions were.

The earliest published paper is my IEEE Compcon (Feb 1997) paper, written in December 1996. A copy is available from http://per.bothner.com/papers .

Danese Cooper


Yes, it is indeed true. Dr. Hamilton was one of the principal voices against open-sourcing Java, and in my experience he's a "my way or the highway" kind of guy...so not surprising that he's opted out of the new "age of participation" Sun. I found it interesting that Rich Green publicly thanked him during the event on Monday. I know many at Sun and in the larger Java community were sad to see him go. So far no word on what he plans to do next.


Danese Cooper


I updated the timeline per your history. Thanks!


Dalibor Topic

It's missing one watershed event, in a sense. The OpenOffice.org/gcj collaboration resulting in OpenOffice.org first building with gcj in January 2005 (Caolan McNamara, Kai Ramme, Mark Wielaard, Tom Tromey, Simon Phipps, Richard Stallman, and a really tiny bit of me in there somewhere, I guess).

It was the first time we actually got a small project going between Sun & the Classpath community, I believe, and it went quite well thanks to Kai's and Caolan's efforts, and the respective teams supporting them.



Thanks for the explanation. But right above the comments, in this post, I read something like "Graham Hamilton and Jeff Jackson move from Java to Solaris group, Laurie Tolson takes charge of JavaLand". Does it mean that Graham has only left the JavaSoft division and not necessarily Sun?



Oops! seems like you are the owner of the blog. I thought that you are just a visitor like me :P


dave shields

No time to list Jikes? Eclipse?

Danese Cooper

Where would you like Jikes to be listed? Sorry, but I'm not sure I think Eclipse has anything to do with the open sourcing of Java.

dave shields

1998 Dec IBM releases the Jikes source-to-bytecode compiler as its first open-source project

Danese Cooper

Thanks Dave for giving me a citation for Jikes. Sorry it took so long for me to post it (read above for reasons).


We use that on our site.

John Gilmore

Mike Tiemann of the 1st free software company Cygnus strongly advocated to Sun to use the GPL for Java, before Java was ever announced (it was Oak then). Mike was a former Sun employee (on the Spring project), well known to Sun executives, and Cygnus was a Sun contractor for projects like the UltraSPARC GCC port. Sun went back and forth on the issue, and eventually decided to use the license that would succeed in punishing Microsoft for their strategy of embracing and smashing compatibility. Once that was handled, hey, no reason NOT to GPL it any more. (Mike is now at Red Hat and can probably provide more details.)

The comments to this entry are closed.