Open source Java is on the horizon following support for changes to a Sun Microsystems-backed industry body overseeing development of the platform and programming language. A new article on theregister.co.uk provides a light discussion of the JCP changges and their effect on J2EE 1.4 implementations.
- Posted by: john flinchbaugh
- Posted on: December 05 2002 09:00 EST
Here is the article at the The Register.
What does the term clean-room mean exactly?
"A specification lead could license the reference implementation and test kit together, so if you wanted a clean-room implementation you had to get the source code from the specification lead which tainted the code," Kluyt said.
My understanding of "clean-room" is that the derivative developers never come in contact with anything related to the original code.
For instance, let's go all the way back to the case of Compaq developing a compatible version of the IBM Personal Computer BIOS. Compaq put together two teams of engineers: group one reverse engineered the IBM PC BIOS and based on that source code wrote a software specification; group two was handed this spec and developed their own implementation of the BIOS that was then used in Compaq's IBM PC Compatible lunchbox portable computer. In this case, the developers of the Compaq BIOS never saw any of the IBM code, and had no contact with the engineers that reversed it. Their implementation wasn't "polluted" by any direct contact with IBM intellectual property, hence the term "clean room".
So what exactly does this mean for those of you involved in some form or another in the JCP (Or at least follow it closely.) Do we have another Microsoft like move on our hands. For those who don't know MS has realeased the .NET Framework as "Shared Source." Its billions of lines of code, some from the actual MS implememtation of the framework. Anyway, is this going to be a partial spec, like MS or is this the whole enchalada?(If that's not how you spell enchalada, I don't care.) Anyone know the answer?