-=-= a b o u t - h e r e =-=-

Cloneable/clone() is broken.

I totally agree with Josh Bloch on this.  Clone is broken.

JGraphT

I'm liking JGraphT lately.  The API offers some nice basic tools to model graphs. Hopefully it will be speedy enough for what I want to do with it.

The newer version of the JavaCC plugin for Eclipse and Java 5.

If you're using Java 5 generics in JavaCC: There are some bugs in the parsing of JavaCC files in some previous versions of the plugin (which causes the plugin to report parsing errors that JavaCC does not).  I suggest getting the newest version.

http://perso.orange.fr/eclipse_javacc/

J.

Ahhh... Java 7 time?

Well, I noticed a few blog entries about Java 7 floating around.  It seems the idea to hold back on the language extensions only lasted for a single iteration.  Can you believe consideration for method references exists?  Amazingly powerful and beautiful, but... really, I hope they end up being more than compile-time tricks (as in generics).  Here's two good blog entries about it:

http://www.bluej.org/mrt/?p=11#more-11

http://kodeape.com/?p=38

Experts and categories.

Oliver Steele has written post that categorizes programmers into two camps--language mavens and tool mavens.  I absolutely commend the idea, but I have a relative distaste for the catagorization of people in this manner.  Are there no exceptions to the people we're putting into two groups?  Given the categories it is difficult to be both a "tool-maven" and "language-maven", but any geninely open-minded programmer will research and discover the language(s) and tool(s) needed for a given job.  People focus their efforts and expertise differently throughout their professional life.  It seems as if a person could, at a defined instance, be a "language-maven" and at another instance be a "tool-maven".

An Eclipse error.

(click for full image)

Array analysis and annotation to remove bounds checking.

Soot has an interesting feature that checks whether arrays can be executed safely (i.e. without bounds checking) and marks them as such in bytecode.  This seems like a good thing.  But, I haven't tried it yet.  We'll see how it goes.

Bitwise, incomplete.

O'Reilly acts like I need help to "blog" their article.  They specifically say "Here's some handy html you can copy and paste to make blogging about this page a bit easier."  They assist you by providing an abstract of the article with some HTML formatting--it feels as they're politely asking me to use it to help advertise their web site.

Here's an advertisement for their web site. *grumble*  At least it's not their handy html.

There's an article I enjoyed reading called "Bitwise Optimization in Java: Bitfields, Bitboards and Beyond" by Glen Pepicelli.  I'd really like more of the "beyond".  It's a good read for those interested in speeding up Java code using bit operations.