Calvin Austin's Blog

May 2005 Archives

Free event at my place (Bay Area)

Checkout the website, http://www.spikesource.com/testfest.html for registration and further details.

Harmony - Friend or Foe

Now, not to keep you in suspense, yes, I believe it is a good thing. Infact I would like to see more developer goodwill from Sun but I'll get to that later.

The J2SE 5.0 TCK terms opened the door for other Java ports like to BeOS, Kaffe and Apache etc to gain access to the Sun test suite for 'free', ok sponsored. Geir was a valued member on my expert group for J2SE 5.0 and with Pier made sure Sun were honest and didn't try and pull any licensing tricks with the terms.

Now having worked on the Sun JVM I would say yes, use the Sun JVM. It is well tested and stable. However there is an often quoted valid concern that Sun could change the Sun binary license at will and that change could be out of the hands of the folks that actually build the JDK.

That means having alternatives that are free of whatever happens to Sun as a corporation is really a good thing.

Now back to goodwill. I wrote in an earlier article that Sun should make whole on the promise of delivering the Java DNA, (the JVM and JLS specs) to a non-Java standards body. This not only would help open source alternatives but also follow through on the faith that developers put in the platform all those years ago.

Personally I don't think that would ever happen, but in this case I wouldn't mind being proved wrong

Harmony. Good luck!

Is this the end of J2EE?

Did Java do this first?

So where did this all come from? I would like to vote that it was a Java first from one of my old managers at Sun, Scott Rautmann. Scott came up with many great ideas and also empowered folks to act on them. He even turned down a director position (when director positions at Sun were hard to come by). Anyway, when Java appeared there was no way to prioritize that a bug was extra special because it came from a paying licensee, say IBM. So they introduced a points scoring mechanism to "weight" the bug higher. I had just pushed all the Java bugs onto the Java developer website (after multiple meetings to convince the Javasoft management that this was a good thing) in 1996. Sun had never published bugs like this before and very few other companies did (Microsoft were especially shy)

Scott suggested that we weight those bugs too so we could tell if a bug was popular or not and with the help from Tony Squier and Rick Levine, who came up with the name, "Bug Parade" was born.

Unfortunately or fortunately Scott decided to semi-retire in Hawaii soon after that so only really saw this and many other ideas from afar.

So do you know of an earlier public bug voting system? If so let me and everyone else know that they were really first!

This work is licensed under a Creative Commons License.