Mark Reinhold, chief architect for Java at Oracle, gave details on developments in Java 8 and beyond, and announced the release of JavaFX 2.0 during his turn on the keynote stage at the JavaOne conference in San Francisco. Oracle cut back on its plans for Java 7 in order to get the new build released in July, adding three of the …
It's Fork/Join not Form/Join!
You tolled them, they lissened and fictstit! (They don't always....)
I want to believe!
Could Oracle really make JavaFX happen – and in the process turn JME into a relevant platform for once?
Despite what the press would have you believe, Java ME is stilling shipping on hundreds of millions of phones each year.
It's the apps
Oh, I know you cannot throw a dead cat to the air without it landing on top of a JME-enabled feature phone; then again, Flash Lite is just as widely deployed. The problem is, installed base does not necessarily equal market relevance.
Topping GroupLayout (Matisse) ?
Now, small disclaimer: I'm not a full time Java developer. I picked up on Java years ago ever since I started Solaris administration because it fascinated me how deep Java was "embedded" within Solaris. It didn't take me long to pick up on NetBeans, at a later time extended with VP's SDE-NB for my UML modelling needs (I actually enjoy working with UMLs) and that's basically where I am now; in the mean time I've developed several Java applications, some simple and some quite complex. Mostly aimed at server administration, but also some client/server related programs.
Alas; while I don't keep up with every detail I still remember when NetBeans' 'Project Matisse' was announced and demonstrated. That was IMO very heavy; an extension to Swing which actually allowed rapid and easily created GUI's in a way which was extremely user friendly. The main 'problem' here was that Java didn't support this new layout manager ("GroupLayout") out of the box. As such it was first shipped as a library with NetBeans and later (iirc around SE 6_10) embedded with Java.
I think that development is extremely had to top. Or what to think about the inclusion of the NetBeans platform in the form of the jVisualVM application; a program which allows you to monitor all running Java applications quickly and easily (checkup on cpu usage, heap size, threads, classes, etc.) ?
All without turning the JDK into a form of bloat (at least IMO).
SO sorry for being a bit cynical here but seeing it believing. Sun has done some pretty cool things with Java in the past, and that is a legacy which is IMO very hard to top.
Am I the only one to feel a sense of dread?
Java innovation invariably means broken apps. I'm not saying that Java shouldn't evolve, just that talk of evolution in Java usually gives me a sense of dread.
Another Date/Time API!
Its basically Joda time with new package names, its a significant improvement on the very broken and inefficient java.util.Date/Calandar
but one that works as you'd want it (assuming people can use the data required to drive it ad not have to pay some astrologers a shed load of cash) and should be the last.
“Some would say adding Lambda expressions is just to keep up with the cool kids, and there’s some truth in that”
Lambda expression should be pack & parcel of any programming language that prides itself on being relevant and high-level.
Get with the program.
Once you feel the expressiveness you won't go back.
Getting ahead of itself.
I found it a little weird that the Java 7 JVM provides the support for dynamic invocations but no way to use it until Java 8, without resorting to another language that is.
As a long time lisp enthusiast I'd welcome this in java but I have to be honest I really don;t think they would add that much, we have anonymous classes which fill the void enough (ok, with a lot of extra verbage to create) and for the times where a functional approach is really the best solution you have plenty of options to do so to mix and match languages. That is true expressiveness, being able to do what you want with the best fit mechanism.
I would rather people concentrate on improvements with more added value. To me that is modularity (please just adopt osgi), improvements around improving support for non java languages and I can wish, AOP. We are getting there. One of the major unblinkering times I ever experienced was landing a job with bunch on unix developers many many years ago having only encountered dos and windows. They had this huge wealth of components and languages right under their finger tips, it influenced my whole outlook on developing software, after a couple of months I could do things in a hour or so that would take a week or even longer, just by having an environment that worked as a whole rather than attempting to corner the market in producing silver bullets that turned into hammers.
I remember having a conversation on an ACCU conference a few weeks after MS announced .NET and teh excitement a few people expressed about being able to mix and match languages was quite high, I tried pointing out you could theoretically do that with java but it just seemed like SGML/DSSSL vs XML/XSLT, i.e. not as trendy. At least java survived better that SGML.
Multiple language development is the future, we are already getting there with , need to apply some dynamic business rules, use a functional language, need to do some text processing as part of that, write it in (j)awk, do some database stuff with the results? then use groovy/gorm and of course java to bind them all into an EE. (BTW, im referring to multiple languages within the same component/app rather than a system with different to physically separate units in different languages)
Sorry, gone completely off the track there.
Try VisualStudio C# 2010
Compared to my last Eclipse "experience" (a few months ago) it appears to be much more polished and full-featured. I guess there are fewer plugins, but for run-of-the-mill programming VS2010 appears to be an excellent tool. Memory hunger has been reduced to 140 MBytes, which is OK for the year 2010 Anno Domini.
..if you want to write C#, oh, and pay (I know there's a free version, but you're not supposed to develop anything commercial on it, and it's hobbled in quite a few areas so not that suitable for developing much beyond Windows console and forms apps.
Eclipse is a dog, there's no two ways about it. As an IDE I prefer Netbeans (or at least did when I last used it in the dying days of Sun). Unfortunately world+dog (Spring etc.) all seem to have Eclipse based dev environments
No doubt about it, after a period of probably around 10 years of using a "real" java ide and emacs, it's only the past couple of years I've spent more time in eclipse than emacs. Machines nowadays can fullfill the promises these ides can provide (not saying emacs can't, honest, still do a lot of typing, emacs + eclipse would be amazing).
But it is bloody good if you have a decent spec machine. From my point of view the reason people are choosing eclipse it is, as a platform it provides so much more and easier, and it's early backing of OSGI was a breakthrough, I;ve been poking about the eclipse 4 releases/source, pretty sure eclipse's popularity is going to improve.
As someone who uses both MS development tools (the paid for ones, at great expense) and eclipse I know what I prefer. Were MS tools are goood is tight integration with their software stack, try introducing something not produced by them and it all goes pear shaped. But even when MS tools are good, they still fall far from the tooling support the java world has had for years.
One thing I wish oracle had kept was the trend towards dropping the use of Java X for version numbers. It's starting to bug me that everyone is doing this, e.g. chrome and mozilla, It's as everyone is getting spinal tap to come up with version numbers.