Slashdot Mirror

Grr... I don't know why java.net doesn't setup a redirect to handle that, if they want everything to go through SSL, but that
URL should actually be:

https://openqabal.dev.java.net

There are probably other FOSS projects to create a truly decentralized, federated social-networking and collaboration package, but the one I'm intimately familiar with is
OpenQabal. OQ is all about developing social-networking and collaboration software that puts users in control of their own information (including the much mentioned "social graph"), supports identity federation, and facilitates distributed conversations. Development is just getting started, but we're working off of a couple of existing code-bases to get a headstart.

Disclaimer: I'm the originator, chief architect and, so far, sole developer on the project, so everything I say may be considered biased, slanted, unreliable, or whatever else your skeptical little heart pleases.

Java? Wide uptake? Surely, you jest.

No, hardly.

It's quite rare now to see any client programs written in Java;

Not in the business world, where Swing clients are probably second only after Visual Basic. Sun is also currently putting a lot of effort into improving the JVM desktop experience.

Agreed. Visual Studio, while having its faults (Windows only [duh, right?], pretty large footprint, and some rather unhelpful "help" features) is better than Eclipse. Visual Studio is quite nice if you just want to write Windows apps. C# is excellent. .NET is great. I just have too much cognitive dissonance when I totally ignore other operating systems, though. If MSFT would just provide .NET plugins for Linux / Mac they'd help themselves out a lot (and yes I know about MONO.) The annoying start page, and the MSDN search that returns nothing of use, and the !helpful code formatting notwithstanding, Visual Studio is pretty good.

I have gone through periods of trying Eclipse. People are incredulous when I tell them I hate Eclipse, so I feel like I have to see what I am missing and install it. Every time, EVERY TIME, I uninstall it, totally disgusted with the amount of RAM it takes and how heavy it feels. It's just another example of why I hate Java apps. Java on the server is fine, but on the client it just falls flat. Clunky look and feel, slow performance, and weird UI quirks seem to be the hallmarks of Java apps (and when people start using terms like percieved slowness, I reach for my Browning). Doesn't anyone else get annoyed by how slow Eclipse feels? Anyone?

To me, an IDE shouldn't get in the way of development. It should like a good waiter -- there when you want but otherwise out of your face. That's where both Eclipse and Visual Studio fall short.

My IDE of choice is jEdit. Yes, I know it's Java based. I do hate the fact that it's Java and it looks like every other Java application: butt ugly (and even with the Look and Feel plugin that is supposed to give more of a native OS look and feel...stupid swing components). But most importantly, it's a mighty fine text editor which has IDE essentials like syntax highlighting, comprehensive searching, and lots of powerful text editing tools (like block select, which my coworkers who use Dreamweaver just drool over.)

If you use Eclipse but hate it for being so heavy, I'd recommend jEdit. It's just the right amount of IDE. Not so much IDE cruft that writing code is less writing code than it is learning how to use the IDE, yet it's got enough utility to facilitate the business of writing code. Any web development I do (ASP, PHP, Cold Fusion, html, css, javascript) is all done in jEdit and I can code a page in the time it takes Eclipse to open a file.

Indeed, it is a real shame nobody sees this sort of deception for what it is. There is someone running around defending the fact that the JRE system libraries will be memory resident regardless of whether a Java application is actually used.

They had a session about the Consumer JRE at JavaOne, and they are NOT putting it in memory. They are doing a touch of the files at startup (or possibly after login) to put the files in the disk cache. Still a cheat, if everybody did it the situation would be hopeless, but still.

the real solution is to reduce the enormous size of the JRE libraries to something more reasonable.

Which indeed is what they are doing.

The one problem for java applications is still startup time. I just don't know what can be done about that except preloading java at boot. Which is a waste if you are not running a java app that day.

Actually, Chet Haase recently blogged about the changes being done in this area. Unfortunately many of these quickstart "cheats" are for Windows only, when questioned about this at JavaOne they said they didn't have enough engineering hours to do this for other operating systems but would welcome community contributions to this with open arms.

Linux and other users WILL still benefit from the Java Kernel work by Ethan Nichola's team though, this will be backported to Java 6 as part of the Consumer JRE project.

The one problem for java applications is still startup time. I just don't know what can be done about that except preloading java at boot. Which is a waste if you are not running a java app that day.

Actually, Chet Haase recently blogged about the changes being done in this area. Unfortunately many of these quickstart "cheats" are for Windows only, when questioned about this at JavaOne they said they didn't have enough engineering hours to do this for other operating systems but would welcome community contributions to this with open arms.

Linux and other users WILL still benefit from the Java Kernel work by Ethan Nichola's team though, this will be backported to Java 6 as part of the Consumer JRE project.

The one problem for java applications is still startup time. I just don't know what can be done about that except preloading java at boot. Which is a waste if you are not running a java app that day.

Actually, Chet Haase recently blogged about the changes being done in this area. Unfortunately many of these quickstart "cheats" are for Windows only, when questioned about this at JavaOne they said they didn't have enough engineering hours to do this for other operating systems but would welcome community contributions to this with open arms.

Linux and other users WILL still benefit from the Java Kernel work by Ethan Nichola's team though, this will be backported to Java 6 as part of the Consumer JRE project.

Can Swing please be replaced with something that doesn't suck in terms of performance?

I think it is doing quite well these days. Have been working quite hard on the OpenGL and Direct3D pipelines. See this recent blog on OpenGL shaders for instance.

Can it also look halfway decent?

Now that they have open sourced it... let's hope. ;)

I'd rather hear a rumor that IBM will soon be dropping AIX (ack! pthhhht!) and only offering Solaris and Linux going forward. Up until recently, IBM's big advantage over Sun was that their boxes scaled to more processors (the Power-based ones, anyway). Now that Sun's finally shipping the Niagara-based stuff, IBM's advantage has shrunk, if not evaporated.

Hmmmm, let's just check something...well, whaddya know — there's a version of WebSphere for Solaris. I guess the Global Solutions guys won't have to learn Glassfish after all....

I'd much rather prefer:

SGPJ - Solaris, Glassfish, PostgreSQL, Java

How are they going to do it? Java? Flash? ActiveX? DHTML and JavaScript? AAlib and a text box? Something like Java and JOGL seems most likely, but I have a hard time imagining it rendering at any decent framerate.

At any rate, I'd guess they're targeting the casual gamers. I have a brother-in-law that spends at least two hours a day playing Flash games. Carmack must be going after those people.

No, not at all. Maybe my original comment had insufficient information - I teach them the same theory (concurrency is concurrency, after all, and an earlier exercise in the curriculum is just that - the development of a generic resource pool (which could host threads, connections, whatever)), but in many respects you have to do a great deal more work using the traditional Java (and most other languages') threading model to accomplish the same thing. This is not because some "shitty implementation" hides details, but rather because of elegant design decisions, such as the separation of business logic from execution/concurrency logic (which I guess you like to be all intertwined in your code).

With Java 5, you can write your tasks (logic) first, and then decide, as an architectural decision, whether you'd like them executed sequentially, concurrently, with various types of thread pools, etc.

Also, the traditional limitations, such as always blocking when you want to acquire the lock for an object when it's not available, are significantly complex to work around, whereas the Java 5 concurrency utilities offer several different types of locks, and ways of acquiring them.

And don't get me started on Conditions - the traditional model of Object.wait() and Object.notify() is simultaneously too simplistic, and introduces far too great complexity when developing complex things (think the controller for an aircraft's controls that has to deal with inputs from two pilots and an autopilot, all of which affect shared resources). Again, Java 5 Conditions provide an elegant platform within which to solve such a problem.

Using something that "does the work for you" is not "shitty" - it prevents you from having to sit all day and debug multi-threaded code. In either case, you have to understand what you are doing, but let's face it, multi-threading is actually very simple on it's own - it's the implementation, and how you use it to accomplish difficult things, that are exceedingly hard. I'm glad you think you can write a better thread pool than the maintainers of the Java programming language. You should join up at http://dev.java.net/, they can use you! (it's open source now through GPL, remember?)

In reply to myself above, to satisfy my own curiosity, I just spent an hour researching this company. Seems they began as EJB Solutions in 1998 with one Rod Cope as co-founder. 2003 finds them headquartered in Highlands Ranch, Colorado. They were specialized in Java, J2EE, XML, and web service technologies, and without doubt the "EJB" in their name refers to Enterprise Java Beans which were all the rage back then. EJB Solutions changed their name to Open Logic sometime around June 21, 2004. They released the BlueGlue 3.2 Open Source Stack July 13, 2005. They launched Indemnification for their Certified Library of Open Source Products on October 17, 2006, but at least one blogger was not too impressed with that

Rod Cope of OpenLogic explains what this is all about in a nice video at JavaOne 2007 where he really seems to be a nice guy who perhaps deserves our support as they have been involved in Open Source and promoting it for a long time. Its just this thing about their Indemnification offer that has touched off a raw nerve here. I hereby invite Rod Cope to contact me and I will ask him about this, do a little interview with him, and report back.

I seem to remember that sun was working on an opensource DRM based on Java called Dream

There's official bindings to OpenGL with Java. That's called JSR-231, better known as JOGL. There's not software implementation of OGL in Java that I'm aware of.

At the InfoViz conference last year there was a company displaying a PCI-Express card that had 4 or 8 cell processors on it and a library to play with it. They had Java bindings to that, so you could something like what you want. They had an extremely impressive realtime mesh-extraction from volumetric data demo that they were running. Unfortunately I don't remember the name of the company right now, but that card had very serious potential to be a vector supercomputer in a box for the high-end sciviz crowd.

Could have fooled me.

I've always been under the Impression that Java can't use hardware rendering, wouldn't that restrict usage for more intense applications?

I've always been under the Impression that Java can't use hardware rendering, wouldn't that restrict usage for more intense applications?

While I'd agree that SUN doesn't have a stellar reputation where GUI's are concerned, they didn't invent this stuff yesterday... they bought most of it. SavaJe had been working on the VM for almost eight years now, Linux on mobile platforms has been around just as long and JavaFX is just a scripting language, that has also been under development for a while, that sits on top of existing Swing libraries.

They have some very slick demos that compare well with Flash.

Of course, I'll be the first to admit that it'll take a while before they have it all integrated and have some professional level tools out there, but it's an encouraging, solid, first step. Also, on this particular platform, it's already most of the way there regardless of how long it takes them to solve the consumer VM issues on other platforms.

> Yes, because it is my call.

Well, If I'm advocating monoculture, then this advocates close-mindedness. You are missing my point entirely. You can't possibly have evaluated every single Java Web Framework. I know you can't because it is not possible. You are depending on the community to do that for you. Personally, I prefer to find out for myself. You jump all over my case because I think there is too much choice to make an informed decision. You are insisting that you have made an informed decision, but I'll bet that you haven't evaluated (read used) more than one or two major frameworks.

As for your assumption that I work at a big company with rigid requirements. Not right now... I own a small consulting firm and I work at the family business. Both less than 50 employees. Before that I worked at a large firm, but it was not a rigid environment. In fact, we were considered CMMI level 1. My opinions come from evaluating the situation and seeing what works well and what doesn't. My genuine case for "too many" cases is this... You suggest MyFaces + Tomahawk. Now, I am going to ask have you tried RAJAX? https://rajax.dev.java.net/

Each time you respond, I am going to either point out that you didn't give the suggested framework it's proper attention, or I will suggest another framework to look at. My guess is that you will eventually give up. Now, as far as big companies with rigid requirements... You strike me as the sort of guy that has trouble working on a team, quote - "Yes, because it is my call." Would you really have a problem if I were your team lead and we would have the following argument - "We are going to use Struts2 for this project," you - "I think we should look at other frameworks like MyFaces + Tomahawk," me - "I appreciate your enthusiasm, but right now there are just too many frameworks and I would rather move forward than spend the next few weeks arguing over which framework we'll use." Are you the type of person that would become unproductive and spend hours complaining to other team members about the team lead not listening to you? It sounds like what CMMI people call the "hero". It's not meant as an insult, heroes are considered very productive at most organizations, but managers with experience running teams of developers are beginning to work out practices to eliminate that sort of behavior.

Also, when I mentioned that the decision of framework should be made by an architect... Well, if your projects are so small that you want to bang them out and get them out the door, why are you using Java? Why not Ruby on Rails? Seriously, Java is great, but for quick projects there is a lot more overhead than just about every other platform out there. Now, if you are working on decent-sized projects that require regular maintenance and service thousands of people per day, then someone should be helping to manage the technology to keep projects somewhat in line so that if SystemA at your company needs to talk to SystemB, then it can happen with little effort. Whether that person's title is architect or he is just one of the older fellas who has been at the company a long time. I have personally been a part of projects where two vastly different technologies were required to work together, in one case, we had to make a Java web-app talk to a custom, non-standard server (http://www.aolserver.com/). Bad architecture leads to more work. If you're in one room deciding that MyFaces + Tomahawk is best for your company, but in another room someone is deciding that Apache + mod_perl and perl's CGI module are the best thing for your company, then I'm glad I'm not you when management decides that your projects need to be integrated.

So, anyways... Tell me what you think of RAJAX, then when you're done take a look at SEAM, and if you have time left when you're done with those two, read my website for information on Struts2 (dig, dig, wink

Java 6 and Java 7 are relatively minor changes.

What!? Java 6 I'll give you was not as big a change as Java 5, but Java 6 was mainly a stability, manageability and performance oriented release, but even that came with some pretty big changes - support for scripting languages with the Rhino JavaScript engine included for instance. Also, the Consumer JRE changes will be backported to it, with the Kernel JVM, the Quickstarter and Direct3D/DirectX pipeline improvements in Windows to catch up with the OpenGL pipeline.

Java 7 might be one of the biggest changes to the Java language and platform ever, take a look at this list for instance. Note that not all of these might be included in Java 7, or indeed ever. Danny Coward who I believe is in charge of the Java 7 release together with Mark Reinhold talks a bit about the difficult balance between giving people the new features they want and keeping it simple.

Although still in it's infancy, JavaFX kinda sounds interesting. I've played around with it a little bit, and it's definitely fun. There are several examples out there showing you how it's done and they get the point across pretty well. It's supposed to be an alternative to Flash and Silverlight and although showing promise, it definitely falls short in it's multimedia capabilities(sound and video). Multimedia would perhaps be the killer feature to add to Java 7. Anyway, they're supposedly optimizing Java 7 to handle JavaFX and JavaFX Script and this is perhaps one of the features that "might" encourage people to upgrade.

https://openjfx.dev.java.net/

Btw, JavaFX was previously known as F3 (Form Follows Function?) for those that may be looking up more details and examples of it.

you can already get the source for OpenJDK 7 here. it's just the JDK, though, as far as i know. i'm also not sure if OpenJDK is technically the same thing as the regular JDK.

A computer, by common definition since 1980s means something you can install programs, there is some payware or free SDK open to development .

I am speaking about these:
http://www.symbian.com/developer/index.html
http://trolltech.no/developer
http://msdn2.microsoft.com/en-us/embedded/default. aspx
http://www.java.net/

You know, Symbian, Linux, WinCE handheld devices are "handheld computers". Lets say, the mail client doesn't fit your needs? Go to sourceforge or handango and get a better client for your needs and install. You need a specific function such as GPS? You buy add on device, install its software and you got it.

A $60 J2ME (Java) supporting regular phone is more close to computer than iPhone is.

I may have lost my little remaining trust/respect to Apple centric media after iPhone. That is a big deal since I use Macs whenever possible. I won't be checking Apple hating gray beard sites either. We don't deserve this kind of media...

GPLed opensource product that institute DRM. It went something like a ssha encryption of the binary content and the provider generated an encrypted key based on your key which was based on your account information(from the provider). Then whatever player you were using needed a plugin that used another program to decrypt the media and stream it into the player.

Just to follow on your example, in such a system, the plugin would probably have to be a closed-source binary blob, or else you could just modify it to intercept and spit out the decryption key as it was being received from the provider. (I'm not trying to personally attack you -- what you created there was as good a DRM system as most of the real ones on the market, but it's running into the fundamental limitation of DRM.) It's all smoke and mirrors.

Anyway, after doing a little Googling, I think the "open source DRM" thing a while back was related to someone on the Gstreamer project discussing adding support for DRMed formats -- but it's still not clear how they'd accomplish that. Some people have pointed towards Sun's drm-opera project as one possible avenue, but AFAIK that's nothing but vaporware, and it too was widely criticized as being impossible when Jonathan Schwartz announced it. According to this article there have been two past attempts to create "open source DRM": one was OpenIPMP in 2002, another was Media-S, more recently.

OpenIPMP has a SourceForge project page, although the latest update was a year ago. Apparently there's some code that can be downloaded, but aside from that they are cagey on how it works, and heavy on buzzwords. Nothing about it makes me suspect that they have really discovered anything huge (and a DRM system that didn't rely on obscurity would be pretty huge). If anyone is familiar with the project and wants to comment, I'm genuinely curious.

Media-S apparently evolved out of an effort to make a "Secure OGG" format. They at least have an FAQ. Basically, they're going for the straightforward 'binary blob' route:

If Media-S is open source, how can the encryption be secure?

If a company wishes to use Media-S to protect their content, SideSpace Solutions highly recommends purchasing a binary distribution license. Under this license, any modifications to Media-S (such as a change of encryption engine or pri

Sun's DReaM. https://dream.dev.java.net/

Doesn't RH own JBoss?

JBoss has teamed up with MS...

http://today.java.net/pub/n/3397

I kind of liked the https://paperairplane.dev.java.net/ aproach to the problem.

Trick maybe the wrong word - changing the L&F is by design with code such as:

try
{
    UIManager.setLookAndFeel(UIManager.getSystemLookAn dFeelClassName());
}
catch (UnsupportedLookAndFeelException ex) {}

While not perfect, it is pretty good. There are screen shots and some of the deficiencies laid out here: http://today.java.net/pub/a/today/2003/12/08/swing .html

That article is misleading. It has nothing to do with AJAX,
and nothing to do with the Web.

It's basically JavaFX Script, a language for defining GUI's.
https://openjfx.dev.java.net/JavaFX_Programming_La nguage.html

These are better articles:
http://arstechnica.com/news.ars/post/20070509-firs t-impressions-suns-javafx-platform-for-rich-applic ation-development.html
http://weblogs.macromedia.com/jd/archives/2007/05/ sun_javafx.cfm

-- Mike

https://jinux.dev.java.net/

I don't know how useful it is, but I haven't heard of anyone actually using it either.

Who actually cares? JIT+VM-in-kernel has been available for many years in the form of Inferno, and nobody cared then either.

Besides, we already have reasonably light-weight runtimes for Python, Ruby, etc. that perform on embedded devices, with infinitely more expressive power than Java. Even Sun appreciates this and is integrating expressive, dynamic languages into future JDK releases, but the typical JVM runtime is still way too heavy to be useful for embedded work and the stripped ones still take up too much space to fit a dynamic language platform on top, let alone execute it fast enough to be useful.

I've been studying FX since the announcement yesterday, and I think that Sun is overhyping it to the extreme. As it turns out, all JavaFX is is a new scripting language formerly known as F3. The purpose of this language was to offer control over the Java2D and Swing APIs in a manner that is easy to use and fast to develop. Because of the control provided, developers are able to create richer GUIs.

Somewhere along the way, the concept got derailed. Sun must have seen the iPhone and started worrying about what would happen to J2ME should it take off. So they yanked F3 off the shelf to show how similarly impressive GUIs could be created for cell phones. But before they could announce it, Microsoft jumped in the fray with their Silverlight announcement. (Silverlight being a powerful multimedia technology solution in search of a problem.) Not to be outdone, Sun somehow managed to convince the press that if you throw F3 (nay, JavaFX!) scripts into an Applet, you have a strong competitor to Silverlight. A rather incredible claim, IMHO, as JavaFX is lacking in the streaming video department. Even more telling is the fact that none of the JavaFX examples are actually applets!

Thankfully, Sun seems to be hedging their bets. None of the pages on the JavaFX site even mention Silverlight, almost making it look like the entire idea was a press invention. Sun's pages make a few passing references about running the technology in an Applet, but nothing firm.

My verdict? I think that F3/JavaFX is the GUI layout technology that Swing developers have been waiting for. With any luck, the technology will create a new market for Java Desktop Applications. The rest of Sun's claims can be safely ignored.

the "correct" url is now:

http://openjdk.java.net/

(the old one still has old content)

So please tell me where you're getting these figures from.

There is also the matter of Java Kernel being developed for Java7 which reduces the size even further: http://weblogs.java.net/blog/enicholas/archive/200 6/09/java_browser_ed.html

As of yesterday (8 mai 2007) sun released the JDK under the GPL2.
Once the rush is over, go to https://openjdk.dev.java.net/ and have a look around for a free implementation.
AFAIK one of the main devels of KAFFEE is now one off five people looking over openjdk.

see http://java.sun.com/javaone/sf/2007/articles/openj dk_sands.jsp for an interview

Two words: Java Kernel

http://weblogs.java.net/blog/enicholas/archive/200 6/09/index.html

This is *the* technology that will bring Java to the desktop. The community should let Sun know if they value this technology and now that the JDK is open-source you can help make it a reality!

Their tutorial on this page https://openjfx.dev.java.net/Getting_Started_With_ JavaFX.html is based on Netbeans. No thanks. If Sun is serious about this technology becoming popular, they'll make an eclipse tutorial. I don't want to drink the Koolaid and abandon Eclipse, which I use for everything. I would like to run a simple tutorial with Eclipse. I suspect that a lot of people won't even bother with JavaFX if they can't just stuff it into another Eclipse project to try it out.

For those who actually want to know about JavaFX instead of saying - Java is slow or applets suck or even worse, JavaScript is insecure or something else irrelevant...

Here is the project FAQ
It is going to be some kind of open source and going to be developed for both desktops and mobile phones as well as for the Blu-Ray and HD-DVD stuff which already has Java built-in to the standards.
Also from what I understand it is build off of a project called F3 and work from a recent purchase by Sun called savaje that makes Java software for cell phones. So it's uniting a lot of efforts that have been in the works. To say that this is just a rehash of applets of the '90s is fairly naive.

for demos, go to: https://openjfx.dev.java.net/#demos

Not to mention JavaFX. It is even to convert Flash files to it.

*cough* http://openjdk.java.net/ *cough*

The http://openjdk.dev.java.net site is defunct.

The important bits are all there, sans part of the Java2D and JavaSound implementation libs.

*cough* http://openjdk.java.net/ *cough*

The http://openjdk.dev.java.net site is defunct.

The important bits are all there, sans part of the Java2D and JavaSound implementation libs.