Triggered by a blog I looked at an article by Richard Stallmann

Well, I finally I got his point and now I am sure that I have to disagree with open-sourcing Java.

First, the article. Stallman speaks about free software. I didn’t put “free” in quotes, but I set it in italics to emphasize that it is a term. Search Wikipedia for what a term is. Stallmann’s term is not the “free” in “free beer” or “free thought”, it is the “free” from Free-Software-Foundation (the guys who brought to you the GPL). The FSF “free” is a label that declares the compliance with some principles which are more or less arbitrary - I like to compre it with the DOGMA-manifesto in films, interesting, but - as they knew themselves - a dogma, not something that ca can universal truth.

When I find something interesting on the net I usually check the license - if its GPL its over. I already had a hard time ripping out GPL-code from a product, no thx nothing for me who has to make a living out of software.

Anyway the probability that Sun would gpl Java is virtually zero. This would mean the end of all corporate Java use and we would enter the dark age of a Microsoft reign. Luckily there are nicer license-schemes, the LGPL, BSD, Apache to name a few.

Question that remains is why should the licensing be changed? Just to have 100% “free”-Software stack? What for? Wake up, man: Most software is written for money and is either sold or used internally where a “free”-stack is absolutly no concern.

One important point is that you can check the code yourself (if you really want to) - this is already possible, so this is not an issue. The only thing that is left is thus, that you must not change the code. You can get a “reseach-license” and do so if you what do something experimental or need a fix (See this post how to fix officially Java). The only hard restriction is that you must not redistribute you hack as “Java”.

I cannot imagine any sane reason for doing such a thing. The last “Java” that went out was Microsoft’s “Java” and Java still hasn’t recovered completly from this blow. I am grateful that Sun organizes the JCP and keeps an exe on the platform, because Java’s USP “write once, run everywhere” will quickly go astray when even a few half-baked “Javas” get around.

The main asset is the predicatability - Java is predictable over versions and plattforms (after some days again with gdb I know why I don’t like C++ anymore). This makes the process perhaps a bit slower as for some niche-tool, but this is it what makes Java 1st choice for 95% of the programmers. I doubt that an open-sourced Java could advance quicker with out any drawbacks.

I am even not sure if we want this. Java 5 is know in the beginning of being accepted as mainstream. Most installations are still running 1.4.2 and I don’t expect them to switch before end 2007. Perhaps this is the true reason for delaying Mustang (which runs fine already), the big Java users need simply a break to digest 1.5 before they can even consider thinking about a next version.

You might argue that this isn’t an argument as everyone is free to stay on the old version. Well, this is only half true. Right now lot of “backports” start to appear and these had been written for 1.4 users that cannot switch and feel being left in the cold. Also Sun now promises only 2 majors being supported, so releasing 1.6 makes the clock ticking for 1.4 and put investments there in danger.

The older might remember the Osborne case: They went bankrupt because the customers stopped buying the current model, because they thought there will be something better soon.