>That's why I think work should be distributed.
>   - the basic introduction and reference (exists anyway, Walter)
>   - a tutorial (will exist sooner or later in sufficient quality)
>   - a complete library reference (must be created anyway with the library)
>   - a few real world sample applications
>will bring the authors a long way towards the finished book.
>
>A book is an important signal.

I've always liked the O'Reilly Nutshell books. I have purchased a few over the years. Since I don't program very often in Java, I tend to use it as a refresher. I have the Java nutshell 2nd edition. It is now on the 4th edition.

Maybe you could start by outlining the table of contents. http://www.oreilly.com/catalog/javanut4/ http://www.oreilly.com/catalog/javanut4/toc.html http://www.oreilly.com/catalog/cplsian/toc.html http://www.oreilly.com/catalog/delphi/toc.html http://www.oreilly.com/catalog/csharpnut2/toc.html

I'm sure O'Reilly would publish it after D went 1.0 and not all the chapters were previously published on the web.