Sorry I haven’t been bloggy lately – too much twittering I fear. I did cross a threshold of interesting Java tidbits culled from various mailing lists though so I thought I would share.
On the JCP Java 7 front, the JSR 292 invokedynamic Early Draft is out for review. This is a bit old but if you’re interested in dynamic languages on the JVM or the JVM in general, you should give a read and discuss on the jvm-languages mailing list. I’m still trying to find time to sit down and understand it properly myself. I had a thought that it would maybe allow for multi-methods on the JVM and thus allow you to write visitors without the visitor pattern. Something to ponder in the future.
Furthermore, the JSR 262 Web services connector for JMX public draft was approved! They’ve had a working version for a while and lots of good info. The ballot was kind of interesting though. Eclipse, IBM, Oracle, Fujistu, and Intel all voted “no” and several included lengthy comments regarding the reliance of the spec on non-standard or older incompatible versions of standards of WS-{Transfer, Eventing, Enumeration, Addressing}. Seems like the expectation is that this will be resolved in some way before final approval.
There have been a few interesting discussions lately on the NIO2 mailing lists (JSR 203):
- Exchange regarding a backport of NIO2 – seems NIO2 will likely be backported under a jsr203.nio.* package. Very cool.
- Q&A regarding the non-existence of createTempFile() and deleteOnExit() in the new file system API
- Excellent questions from Elliotte Rusty Harold on file system metadata which led to an interesting thread addressing the concerns.
- More excellent questions from Elliotte on .exists() and .notExists() and answers on their non-complementary nature.