Will Oracle Keep Funding Sun's Pet Java Projects?

gkunene writes "Oracle expects Sun to contribute to its operating profit right away. To make that happen, Oracle may pull funding and staff from projects such as JavaFX, Project Looking Glass, and Project GlassFish."

Comments

Glassfish is a Must-Have for Oracle

[AKAImBatman]

IMHO, JavaFX has been a solution looking for a problem. Applets aren't coming back (thank God), so stop trying to create an ideal Applet platform. HTML5 is meeting that need well enough, thanks' much. Pulling funding from the JavaFX project would hardly even be noticable.

Project Looking Glass is one of those things I'd hate to see go, but Sun hasn't exactly done much with it. Oracle needs to decide that they'll support it full hog as a core product or just leave the project to the OSS community. This noncommittal attitude has been leaving the project in limbo.

Now Project Glassfish, that's a whole other ball of wax. Oracle screwed up Orion (the BEST J2EE server back in the day) to insane levels of uselessness under the guise of Oracle Application Server. (Hey look! Oracle is almost as good at naming as Sun!) Glassfish (aka Sun Java System Application Server) is modern, scalable, easy to use, and absolutely wicked when deployed. Oracle would do well to give up on OAS and just let Sun keep doing what they're doing with SJSAS/GlassFish.

Re:Glassfish is a Must-Have for Oracle

[pohl]

I agree. Pulling funding for glassfish would be a horrible move.

Re:Glassfish is a Must-Have for Oracle

[BabyDave]

Wasn't that supposedly one of the main reasons that they bought BEA - to get Weblogic to merge with/replace OAS?

Re:Glassfish is a Must-Have for Oracle

[MrEricSir]

Applets are still used quite extensively, actually. And now we have webstart, which is more or less the same candy in a different wrapper.

Re:Glassfish is a Must-Have for Oracle

[AKAImBatman]

Ever since Microsoft got away with a slap on the wrist, Oracle has been buying their way to a monopoly. They give excuses for purchasing competitors (some of which might even be true), but their core aim is to be the big fish in the pond.

Oracle may get some benefit out of BEA's product line or they might trash it. Doesn't matter either way. Oracle eliminated a competitor, bought a market, and is looking to reap the rewards of that maneuver. The tech is secondary.

That being said, the Sun purchase is slightly different. Oracle and Sun have been a strong pairing on the high end of database deployments. Oracle needs Sun and their hardware to survive. It doesn't hurt that owning the business gives Oracle enough tools to hit IBM where it hurts...

(I'll have to visit IBM sometime and see how many bloody stains I can find on the walls. There has got to be some serious head banging going on over there. ;-))

Re:Glassfish is a Must-Have for Oracle

[pembo13]

I'd take a WebStart client over an AJAX client any day.

Re:Glassfish is a Must-Have for Oracle

[RightSaidFred99]

Yeah, I don't know how Glassfish got lumped in with the "pet projects" moniker. I think whoever wrote the summary doesn't know what it is.

Re:Glassfish is a Must-Have for Oracle

[arthurp]

I'd be really sad to see JavaFX die. I know people hate applets and although I don't agree with them I can't really blame them. Applets have done some serious sucking over the years. But I think times have changed a lot. And especially with all the new JVM languages popping up I'd be really sad to see Flash continue to be the goto technology for interactive graphical web apps. This is partly because I hate flash though.

Re:Glassfish is a Must-Have for Oracle

[msobkow]

More to the point, Glassfish has been sold with support to a number of companies. Unlike JavaFX (which has virtually no market share), a significant number of paying customers have bought into the Sun Application Server/Glassfish.

In other words, Sun has contractual obligations to continue with Glassfish, and Oracle has inherited those obligations. They can't just drop support.

Re:Glassfish is a Must-Have for Oracle

[MyLongNickName]

How is something +4 Informative when no reasoning is given behind the thinking?

Re:Glassfish is a Must-Have for Oracle

[mrtom852]

JavaFX is aimed at emerging markets - mobile and set top box. This is going to be around for a while but has been very slow off the mark (notably lack of Linux support - I'm not talking desktop here!) and needs a kick up the backside.

I'm not sure if Glassfish will survive in the long run. It has some great components, like grizzly, that will survive but at the other end of the scale it has parts that should be put down, like MQ.

Re:Glassfish is a Must-Have for Oracle

[larry bagina]

a lot of moderators were wondering if pembo13 prefers ajax or java.

Re:Glassfish is a Must-Have for Oracle

[H0p313ss]

How is something +4 Informative when no reasoning is given behind the thinking?

Can you imagine the kind of twisted reasoning that led to that conclusion? Are we perhaps better off not knowing?

Re:Glassfish is a Must-Have for Oracle

[hubert.lepicki]

Just wondering how big this number is... I am into web applications business for quite some time now, but can't remember of any company using Glassfish for deployment of Java apps. People use JBoss for Java, Passenger for Rails, and ASP for .net stuff, but I didn't came across a single Glassfish production environment...

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

HTML5 is fine for crappy little MySpace style web pages and stuff, but in the professional world you want to maintain separation of concerns. HTML5 goes against all the good work that's been done over the last decade towards improving that and hence improving maintainability, portability and accessibility with the XHTML standards.

HTML5 is not fit for large scale business use, it's a kids standard taking us back to the days of old. If we want to make sure everyone can publish then we need to use web applications like YouTube, Wordpress and so on, bastardising the webs most important standard so it supports different styles of syntax meaning there's ambiguity when it comes to making use of it.

Leave the HTML markup to professionals who are competent enough to use standards like XHTML and make good use of separation of concerns, let them build web applications that the masses can publish easily with. Don't break the web with a shitty standard that effectively takes us back to the 90s where HTML was just a damn mess and we had the painful to look at AOL and Geocities homepages of old.

Re:Glassfish is a Must-Have for Oracle

[rbanffy]

I am astonished to hear Project Looking Glass is still around.

Re:Glassfish is a Must-Have for Oracle

[rbanffy]

He was merely informing he would take a WebStart client over an AJAX client any day.

Well... I wouldn't.

Now, someone mod me um +5 informative.

Re:Glassfish is a Must-Have for Oracle

[Chyeld]

In other words, Sun has contractual obligations to continue with Glassfish, and Oracle has inherited those obligations. They can't just drop support.

Wanna bet?

Dear Valued Customer,

As you may already know, Sun has recently been acquired by Oracle. As part of this process, Oracle also inherited the service contracts you held with Sun. However, due to the current economic climate, some of Sun's less profitable product lines need to be discontinued or consolidated into existing product lines Oracle already has.

As such, Oracle is declaring that the following products: X, Y, and Z are being end of lifted as of December 31st 2009 and will no longer be supported Jan 1st 2010 and onward.

If you held a service contract for any of these products, you have the following options:

Accept a pro-rated refund on the remaining length of your service contract along with a nominal cancellation fee.

Contact your Oracle sales rep for a one time offer: a heavily discounted conversion from your current product and contract to Oracle's X, Y, and Z Squared server and a corresponding service contract covering the remainder of the length of your old one.

We do apologize for this inconvenience, but are confident that once you use our XYZ server, you will not look back.

Sincerely,
Bob

Re:Glassfish is a Must-Have for Oracle

[tcopeland]

> JavaFX has been a solution looking for a problem

There certainly was a lot of energy poured into it, though. I feel bad for folks like Josh Marinacci, who has been working on it for a while. Seems like every other Java Posse episode had some mention of JavaFX and its progress (or lack thereof).

Re:Glassfish is a Must-Have for Oracle

[D Ninja]

They give excuses for purchasing competitors (some of which might even be true), but their core aim is to be the big fish in the pond.

I think you mean the big GLASSfish in the pond.

What?

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

It's free as in freedom, or at least it's going to be once Sun gets it's act together.

The Linux version is likely to be released this summer (according to a high-end technologist from Sun's swedish section, whose lecture i attended last week). This is, obviously, a nice thing. The Flash IDE is closed-source, expensive and does not run on Linux.

Other than that, i've been told JavaFX performs very well, but i cannot confirm this. This would make sense, as the Java VM is quite fast after all.

Re:Glassfish is a Must-Have for Oracle

[Red Flayer]

Ever since Microsoft got away with a slap on the wrist, Oracle has been buying their way to a monopoly. They give excuses for purchasing competitors (some of which might even be true), but their core aim is to be the big fish in the pond.

Buying their way to a monopoly is very different from buying their way to being a big fish in a little pond.

Note that buying BEA still makes then only the second-biggest middleware firm (SAP still being larger in that market).

I agree that Oracle wants to be the dominant competitor in each of the markets it competes in, BUT that is not the same as having monopoly position.

Truth be told, aside from the Sun acquisition, most of Oracle's acquisitions in the past few years have been about horizontal growth -- getting Oracle middleware products into markets where they had little presence (finance & banking, insurance, etc). There has not been so much of them buying competitors in markets they already have a big presence in, which is where the monopoly fears should come.

Re:Glassfish is a Must-Have for Oracle

[rackserverdeals]

According to this blog post, Oracle is the biggest user of applets in their Oracle Forms. The author thinks that moving oracle forms to javafx could be a big plus for Oracle and I think there are merits to what he says.

The popular alternative is Adobe Flash. It's been a while since I tried programming in actionscript but last time I did, I had lots of bumps and bruises from knocking my head against the wall.

The last release of Project Looking Glass was in January of 2007. I don't think there's been much going on with that. Since then, I think Sun has been contributing more to Compiz to get it working with Soliars.

I don't know if Orion was ever "the BEST" J2EE server, but it claims to have been the first fully J2EE compliant one.

Oracle didn't buy Orion, they licensed it and used it as the base of their Oracle Application Server. Don't know how that screwed it up since Orion still exists on it's own and isn't owned by Oracle.

Glassfish is the J2EE server reference implementation. It's important to keep it as a reference implementation and is becoming very popular.

I remember before Tomcat became the reference implementation for the Servlet spec, it was popular to use but hard to find hosting for. After it became the reference implementation (and performance improved) more hosts started offering it instead of (or in addition to) resin.

In the opensource J2EE server space, Glassfish is well behind JBoss, but more popular than Apache Geronimo and it is gaining. If they need to keep a reference implentation for the J2EE spec, it would likely be an open source project. JBoss I think is out because they can't get enough control of it, Geronimo is heavily backed by IBM, (who Oracle is competing with on the DB and middleware front) so that only leaves Glassfish.

Re:Glassfish is a Must-Have for Oracle

[rackserverdeals]

Dude! I hope you didn't reply to that message and give any personal information. It's clearly a phishing attempt, and a poor one at that!

First, the Sun/Oracle deal hasn't been finalized.
Second, you can see their form message software screwed up and didn't replace the variables X, Y and Z with real products. I see that happening a lot in spam lately.
Third, they didn't sign it with a full name, title, and contact information.

Re:Glassfish is a Must-Have for Oracle

[fm6]

Applets are still used quite extensively, actually.

Used? Yes. Extensively? No. There are too many competing technology for embedding applications in a web browser, and Java's applet API is the least powerful among them.

And now we have webstart, which is more or less the same candy in a different wrapper.

No it's not. JWS applications don't use the applet API and do not run in a browser window. They're just like non-embedded applications, except that instead of typing "java main class" or executing a JAR file, you execute an XML file that tells your JRE the URLs it needs to download. You might use a web browser to obtain the XML file. But you can also have a local XML file that you can run without firing up a web browser.

Even at Sun, people mostly agree that applets aren't really useful. They're probably be around as long as Java is, but if Oracle has any sense, they'll stay in maintenance mode with no more API development.
 

Re:Glassfish is a Must-Have for Oracle

[CrashandDie]

but [Oracle's] core aim is to be the big fish in the pond

Oracle probably has quite close to a 100% market share with Fortune 100 companies, and well over 90% with Fortune 500. They are the big fish in the pond.

Re:Glassfish is a Must-Have for Oracle

[thetoadwarrior]

Applets are no worse than anything else. The only difference is the price tag on Flash stops more people (at least those with a conscience) from making any old piece of shit and tossing it up on the web. Professionally done applets are excellent and the best solution to web apps so far.

http://www.wordle.net/ is rather popular despite using an applet but that's because it's done right. It loads nearly instantly and it's not some awful collection of graphical effects tutorials thrown together.

JavaFx is used for more than Applets. It's used in mobile phones too. My main reason for ignoring it is it's basically Java but trying to be all trendy and written different.

I'll stick with normal Java until I have a reason to try JFX. It doesn't help that it's a Windows/mobile only thing.

Re:Glassfish is a Must-Have for Oracle

[thetoadwarrior]

Web start apps work perfectly offline without having to download extra functionality that differs between companies, is always up to date, works like a real application on and off line.

Those are 3 good enough reasons.

Re:Glassfish is a Must-Have for Oracle

[Sir_Real]

They'll can Glassfish because they're committed to Weblogic, and it would be very not good to release 3 different J2ee servers in 5 years. OAS is an absolute rotting turd. Weblogic is probably going to suffer the same fate unless they manage to write some awesome tooling. There's a ton of potential to create a painless j2ee development stack with Oracle and Weblogic eclipse plugins. They've never managed to do this. Eclipse + OAS is a friggin nightmare. OAS + Oracle DB works fine if you don't do anything fancy like HA JMS Queues (ActiveMQ) or use the data-guard debacle against an oracle RAC db. You'd think they'd get database connection failover right, seeing as they own both ends of the stack. As of 10.1.3 R2, you'd be wrong. They needed Sun badly. I don't know if they'll be able to fix things, or if they even need to (for all it's flaws, it's still the hottest corporate stack going right now).

Re:Glassfish is a Must-Have for Oracle

[afabbro]

Oracle needs Sun and their hardware to survive.

I'm genuinely curious why you would think that. Oracle is a software company. I suspect that Sun's hardware business will be the first thing Oracle jettisons.

Oracle has a high-end database machine which is made by HP. I really don't see what owning SPARC servers (not a growing market), x86 servers (high effort, low return), or tape systems (StorageTek) gives them.

Re:Glassfish is a Must-Have for Oracle

[glwtta]

Glassfish (aka Sun Java System Application Server) is modern, scalable, easy to use, and absolutely wicked when deployed.

Shit, I must've been using some other Glassfish, because I found it to be an overcomplicated, cumbersome beast to set up and administer.

Every time I start thinking that I need Glassfish instead of Tomcat, I slap myself really hard in the face and go back to the drawing board - it's actually done wonders for keeping "Enterprise-iness" at bay in my projects.

Re:Glassfish is a Must-Have for Oracle

[rackserverdeals]

I'm genuinely curious why you would think that. Oracle is a software company. I suspect that Sun's hardware business will be the first thing Oracle jettisons.

Oracle is mostly deployed on Solaris/SPARC with Linux coming in second, despite Oracle's Linux push.

The cost of Sun Hardware is a drop in the bucket compared to Oracle licensing costs and if people are willing to pay that much they want a solution that will give them peace of mind on mission critical systems. Solaris/SPARC still has some benefits over Linux/x86.

Oracle has a high-end database machine which is made by HP

HP and Oracle came out with the Exadata storage server. That's not a "high-end database machine". Oracle seems particularly interested in Sun's Open Storage systems. Their other storage solutions, should also be of interest and you have to wonder what will come of Exadata now that Oracle will be able to manufacture their own with ZFS.

I really don't see what owning SPARC servers (not a growing market), x86 servers (high effort, low return), or tape systems (StorageTek) gives them.

While UltraSPARC servers have been declining, the CoolThreads servers have been growing. In a previous story, someone commented on getting an Oracle branded server and the core counts made it sound like it was a rebranded Sun CoolThreads server.

If you don't see what owning the most popular platform used to deploy their software, x64 servers for edge servers and smaller deployments and storage and backup solutions might mean to Oracle, don't know what I can tell you.

Re:Glassfish is a Must-Have for Oracle

[rackserverdeals]

Glassfish isn't so bad compared to other app servers and the integration between NetBeans and Glassfish has really helped a lot.

Tomcat is a lot easier but it's not a full J2EE app server, it's only the servlet container.

A lot of the times you don't need the full J2EE stack. Tomcat is sufficient.

Re:Glassfish is a Must-Have for Oracle

[afabbro]

HP and Oracle came out with the Exadata storage server. That's not a "high-end database machine".

No, but this is.

Re:Glassfish is a Must-Have for Oracle

[CynicTheHedgehog]

Glassfish may continue to live on as the reference implementation for the JEE stack. It's open source so it's not just going to vanish. It already uses TopLink (from Oracle) as its EJB3 persistence implementation and already plays very nicely with Eclipse and especially NetBeans (which they now also own). If Oracle was smart, they would build an enterprise Java development stack around NetBeans, Glassfish, ADF Faces, and Oracle DB and reap the benefits of a massive installed user base and the contributions of the open source community. It would be healthy competition for RedHat's JBoss+Eclipse+Postgres stack (although I'd take the latter any day.) Heck, all its missing really is an enterprise DB and a big name to provide support.

Not sure where WebLogic is with EJB3, ESB, and other emerging APIs and technologies, but Glassfish is already there, so you would think it would be a simply cost/benefit analysis in terms of the cost of bringing one up to the level of the other.

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

They already have BEA Weblogic, still one of the best JEE-Server with a huge market share - so they don't exactly NEED glassfish

Re:Glassfish is a Must-Have for Oracle

[rackserverdeals]

That's just some exadata storage servers thrown in a rack with some DL360s.

When Oracle owns Sun, they will own all the hardware necessary to replicate that in house. Which would bring down consumer costs and increase profits for Oracle.

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

Oracle bought BEA Systems 2 years ago... BEA's obviously most valuable product was WebLogic, but if you look deep enough, their whole line of products where competiting against Oracle's own (editors, VM, application server, platforms). In other words: just as in BEA's case, and many, many other Oracle's former competitors, "resistence is futile" and "you will be assimilated": cancel whatever project doesn't fit your (finantial) strategy, integrate and resell whatever does fit. Glassfish can't keep running against Oracle's own self-competing two or three other server platforms, Java FX is irrelevant in Oracle's business and PLG (ha ha ha ha !!).
Java development as we know it (sometis open source, sometimes not) is oficially dead now, move on. The end of an era. I will look for a job somewhere else.... as a DB admin maybe.

Re:Glassfish is a Must-Have for Oracle

[greg1104]

They can't just drop support.

Sure they can, the only question is how long it will take before they can do it cost-effectively. Sun's normal sales terms talk about how they can pull out of their side of a support agreement in section 5.5. Typically Sun's contracts are executed for some number of years at a time, with even their Perpetual Entitlement Contracts having a renewal date in them. So Oracle could announce an end-of-life date effective a couple of years out, probably 3 years, and start tapering down the support work they do and the free resources available immediately (try and find good Oracle support info without a paid Metalink account).

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

IMHO, JavaFX has been a solution looking for a problem.

If that.

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

Used? Yes. Extensively? No. There are too many competing technology for embedding applications in a web browser, and Java's applet API is the least powerful among them.

Really, what are they? All the "apps embedded inside a browser" that I know of are applets. What is more powerful than Java - in a browser, are we including Active X controls in this? You and your moderation is really confusing me, there must be something out there you all had in mind.. but what on Earth is it? Wishful thinking? I really hope you were all thinking Active X, because if it was flash, we need to sit down and have a long talk about what "application" means. FTR, here's an extremely common applet, EMC's NaviSphere. Here's another one, hold onto your pants, WebEx (both ActiveX and Java). I bet you'll find Java applets in just about any medium sized or larger business. I have a hard time using the word extensively myself, but in the small world of "application in a browser", you're only talking about Java and ActiveX, they're both everywhere, regardless what you all think/wish.

And now we have webstart, which is more or less the same candy in a different wrapper.

No it's not. JWS applications don't use the applet API and do not run in a browser window.

Are you unfamiliar with candy or the wrappers they come in? They're both Java. One is inside a browser, one not. One has a more restricted security policy. Do you really know what an applet is?

Even at Sun, people mostly agree that applets aren't really useful.

I agree, because there is very little overlap between something that must be Java, and must be inside a browser. The browser doesn't provide a whole lot of value to that combination, just a route to get to the application. JNLP is an answer to the 'how do I get from a browser to my Java app' concern.

Re:Glassfish is a Must-Have for Oracle

[petermgreen]

There isn't much applet specific API really, the applet class allows the applet to request a few things from the host but mostly it's just standard java APIs with security restrictions applied.

Afaict there are two types of applet, "untrusted applets" that just run and "trusted applets" that pop up a security warning with signature information asking if the user wants to run them or not.

Afaict the latter category can do pretty much whatever they want.

Similar things apply to web start apps afaict, there are both "trusted" web start apps that can do whatever they like and "untrusted" ones that can't.

One annoying thing is that sun created an api (under javax.jnlp) to allow users to give untrusted code the ability to (among other things) read or write a specific file without giving them full access to the system. But for some reason they only made it availible for web start apps not applets.

Re:Glassfish is a Must-Have for Oracle

[Hognoxious]

Applets aren't coming back (thank God)

Hasn't flash (and/or flex) effectively become the same thing?

Re:Glassfish is a Must-Have for Oracle

[Hognoxious]

I second that. Sun don't even own X.

Re:Glassfish is a Must-Have for Oracle

[Ashcrow]

From a technical point of view I'm sure your right but remember, Oracle is not really about the technology. For Oracle it's about the cash. Think of Oracle like a rapper and Sun like, well, a nerd. The nerd is smart but but in the end he will get slaughtered by the rapper, yo!

Really though, technically better doesn't mean a company will back it esp when it means backtracking on previous 'we are better than they are' comments.

Re:Glassfish is a Must-Have for Oracle

[fm6]

There isn't much applet specific API really...

The issue isn't how big the applet API is, the question is whether JWS applications use it. And they don't.

Look, an applet is a particular kind of Java application, and JWS applications just aren't the same. Unless you're going to insist that all Java GUI applications count as a applets, in which case I hope you and Humpty Dumpty will be very happy together.

Re:Glassfish is a Must-Have for Oracle

[MichaelSmith]

Java applets give more security to the client than flash applets. Developers prefer flash. Clients would prefer java if they understood the difference. Developers get to choose the technology which gets used on web pages.

Re:Glassfish is a Must-Have for Oracle

[petermgreen]

Look, an applet is a particular kind of Java application
Actually it's more a type of AWT custom control than a type of application. All that really means though is you don't have to write a "public static void main" to create your main window since it's already provided by the environment creating your applet.

and you don't need a browser to run an applet, the JDK has come with an "appletviewer" right from the start afaict (the oldest version I could easilly find was 1.1.6 and that had it).

and JWS applications just aren't the same
There are a few little bits of API that applets get that web start apps don't (mainly stuff for interacting with the browser) and a few little bits of API that only web start apps get (the stuff that allows untrusted web start apps to load/save from/to files the user selects) but the vast majority of the code in an applet or jws app is standard java/awt (and if you want it swing).

Trusted applets and trusted web start apps can both do whatever they like . Untrusted jws apps can do a few things that untrusted applets can't (which pisses me off because I see no good reason for it).

Re:Glassfish is a Must-Have for Oracle

[digiti]

You obviously never tried deploying a web start application. If it works then great its a wonderful thing, problem is that the simple idea is botched horribly on many machines due to multiple VM versions with incompatible web start revisions etc. There are some cases where you just can't figure out the problem. I defended and advocated JWS for years but Sun just can't get it right even with the huge investment for Java 6u10 JWS is still broken for many applications (not to mention breaking the native look and feel so badly that all sorted tables no longer work!).

BTW This is largely FX's fault which caused Sun to essentially move all non-fx development to maintenance mode.

Re:Glassfish is a Must-Have for Oracle

[lanc]

sparc servers as a not growing market, huh? You, Sir, seriously need to have a look at the T2+ and the Sparc64 based servers. Sure it took them a while to get those to the market - but see them rock (;)) now.

Re:Glassfish is a Must-Have for Oracle

[Dewin Cymraeg]

I strongly disagree that HTML5 is meeting the needs of the RIA world. If that were the case, why would Adobe Flex be doing so well?

As someone who has worked on web development, desktop development (in the good old days of Delphi) and used Flex, I can categorically say that any HTML-based web development is painful and slow compared to using Flex or any homogenised development system.

HTML+JavaScript+AJAX+CSS+WS+Java+Spring Web (or Tapestry or Struts or....)

OR

Flex (CSS+MXML+ActionScript) + Java on the server

OR

JavaFX (CSS+JavaFX Script) + Java on the server

I am personally fed up with trying to use HTML as a platform for applications. It isn't fit for purpose and it was never intended to be. It takes good old fashioned ideas of sane development and uses them as to clean its rear end.

NEWS: INTERNET BROWSERS ARE A RUBBISH PLATFORM

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

IMO, JavaFX is a brilliant idea, just don't compare it to Flash. It takes ideas from flash (the UI) and integrates it with the Java platform. It is not a head-to-head competitor to Flash. Flash is used in a different market. If you're a ad developer or you develop some other shiny apps, where design is most important, use Flash. But for serious software (tm) for which Java is used mostly, JavaFX must be seen as Swing 2. I think it's a nice combination of using statically compiled Java for the back-end business logic and a clean scripting language for the UI side. Whats wrong with that? Now you can develop Java desktop (or web whatever), i.e. "Rich" apps easily which dont look like something from the previous century.

Re:Glassfish is a Must-Have for Oracle

[fm6]

We can argue all day about which is better. I'm not really an expert. The simple fact is that Java applets are drastically less popular than all the alternatives. That must suck if you think Java is the better technology, but it's the way things are.

Re:Glassfish is a Must-Have for Oracle

[fm6]

All that really means though is you don't have to write a "public static void main" to create your main window since it's already provided by the environment creating your applet.

Dude, that's a fundamental difference in the structure of the software. It means the whole thing is using a major variant of the runtime. Calling that a small change is like saying an a change from "printf" to "shutdown" is a small change.

Do you need this environment to run a JWS application? No you don't. If you don't need the applet environment, it's not an applet.

Re:Glassfish is a Must-Have for Oracle

[Da VinMan]

Oracle's acquisition of BEA probably means that WebLogic is going to rule the Oracle JEE roost for the foreseeable future. Orion/OAS/OC4J is likely dead within 2 years; just as soon as they can incent customers onto WebLogic (which may well be renamed into the OAS brand). I don't know what they'll do with Glassfish; it's clearly redundant. However, as FOSS, it will surely live on perhaps as a fork.

If Applets are dead then so is client-side RIA. And it's not dead; just look at Adobe Flex then AIR. BTW - JavaFX deserves a place at the table no less than Adobe. Client-side RIA is at least as viable as HTML5 going forward; especially since client-side RIA is hand in glove architecturally speaking, which is something HTML can not achieve without a re-design.

Hopefully, someday it will just be common practice to stop abusing document oriented architectures for complex application creation. I saw it done with Lotus Notes, Microsoft Office automation, and on the web with HTML, etc. and I'm sick of these architectural abominations.

Re:Glassfish is a Must-Have for Oracle

[Anonymous Coward]

If that were the case, why would Adobe Flex be doing so well?

If pavement is really so great for cars to drive on, then why are car sales so good? Huh? Huh? Huh?

That had to be the stupidest apples to oranges comparison. Ever.

I am personally fed up with trying to use HTML as a platform for applications.

Maybe that's because a) You don't know what you're doing b) you're looking in the wrong place. If you don't know Javascript inside and out, you are an ineffective web developer. As a bonus, you're also an ineffective Flash developer, but you think throwing enough animations at the problem will solve it.

NEWS: INTERNET BROWSERS ARE A RUBBISH PLATFORM

News Flash! Dewin Cymraeg just announced his incompetence! Does anyone care? No? Bueller? Bueller?

Re:Glassfish is a Must-Have for Oracle

[rackserverdeals]

sparc servers as a not growing market, huh? You, Sir, seriously need to have a look at the T2+ and the Sparc64 based servers.

In sun's quarterly report they said that the SPARC enterprise line of server revenues were declining while SPARC CMT line and x86 server revenues were both growing but not enough to make up for the declines in SPARC Enterprise.

So if you take x86 out of the picture, the gains in SPARC CMT covers the los of revenues in non CMT SPARC servers even less.

Re:Glassfish is a Must-Have for Oracle

[thetoadwarrior]

I have to admit I have little experience as I've only recently got into it but so far my tests amongst my friends and myself had been successful.

I just asked NetBeans to enable web start, it generated a default jnlp, I tweaked it and it works. Luckily my work has a load of old machines which I can easily test on which I'll do when I get closer to having a decent program.

Re:Glassfish is a Must-Have for Oracle

[digiti]

The main issues with FX are the same as those people have with ActionScript. The language is reasonable as a scripting language to tie things together but it feels very odd for most Java developers. Unlike action script you don't get the advantage of the amazing GUI tools Adobe offers.

Web start seems simple when creating hello world, NetBeans gets you started with a self signed JWS version. For production you will need a proper certificate which is harder but doable. Then there is the issue of pack.gz with its odd special case servlet (don't get me started on that one). Then you have the issues of HTTPs deployments not working for some people and the multiple incompatibilities between JWS versions installed which occur when someone has a corporate PC with a specific version of Java installed.
There are lots of issues some summed up here for older versions:
http://www.jroller.com/vprise/entry/should_you_use_jws

Newer versions of JWS (Java 6u10) broke different things but didn't really fix lots of the long standing issues. Most people using JWS just got burned one time too many.

Re:Glassfish is a Must-Have for Oracle

[petermgreen]

Professionally done applets are excellent and the best solution to web apps so far.

The one issue I have with applets is that afaict (and please do correct me if i'm wrong) the security model is all or nothing. Either you run in an extremely restricted (no way to copy things to the clipboard, no way to save files, no way to access servers other than those your applet was downloaded from) sandbox or you have a scary security popup and then the applet can do whatever it likes .

Java web start provides a way to ask the user for permission to do a number of things the sandbox would not normally allow. Why they didn't add similar functionality to applets I do not know.

Re:Glassfish is a Must-Have for Oracle

[afabbro]

I'm sorry, you were saying?

NO

[youngdev]

We need javafx to be ported to linux and wait until gf v3 is ready for release.

Re:NO

[multipartmixed]

Um, gf3 was released _years_ ago, and nobody liked it because it was buggy as hell. Here is sample from one of the support forums, you'll see what I mean:

Dear Technical Support:

I'm currently running the latest version of GirlFriend and I've been having some problems lately. I've been running the same version of DrinkingBuddies 1.0 forever as my primary application, and all the GirlFriend releases I've tried have always conflicted with it. I hear that DrinkingBuddies won't crash if GirlFriend is run in background mode and the sound is turned off. But I'm embarrassed to say I can't find the switch to turn the sound off. I just run them separately, and it works okay.

Girlfriend also seems to have a problem co-existing with my Golf program, often trying to abort Golf with some sort of timing incompatibility. I probably should have stayed with GirlFriend 1.0, but I thought I might see better performance from GirlFriend 2.0. After months of conflicts and other problems, I consulted a friend who has had experience with GirlFriend 2.0. He said I probably didn't have enough cache to run GirlFriend 2.0, and eventually it would require a Token Ring to run properly. He was right -- as soon as I purged my cache, it uninstalled itself.

Shortly after that, I installed GirlFriend 3.0 beta. All the bugs were supposed to be gone, but the first time I used it, it gave me a virus anyway. I had to clean out my whole system and shut down for a while.

I very cautiously upgraded to GirlFriend 4.0. This time I used a SCSI probe first and also installed a virus protection program. It worked okay for a while until I discovered that GirlFriend 1.0 was still in my system. I tried running GirlFriend 1.0 again with GirlFriend 4.0 still installed, but GirlFriend 4.0 has a feature I didn't know about that automatically senses the presence of any other version of GirlFriend and communicates with it in some way, which results in the immediate removal of both versions.

The version I have now works pretty well, but there are still some problems. Like all versions of GirlFriend, it is written in some obscure language I can't understand, much less reprogram. Frankly I think there is too much attention paid to the look and feel rather than the desired functionality. Also, to get the best connections with your hardware, you usually have to use gold-plated contacts.

And I've never liked how GirlFriend is totally "object-oriented." A year ago, a friend of mine upgraded his version of GirlFriend to GirlFriendPlus 1.0, which is a Terminate and Stay Resident version of GirlFriend. He discovered that GirlFriendPlus 1.0 expires within a year if you don't upgrade to Fiancee 1.0. So he did, but soon after that, he had to upgrade to Wife 1.0, which he describes as a huge resource hog. It has taken up all his space, so he can't load anything else. One of the primary reasons he decided to go withWife 1.0 was because it came bundled with FreeSexPlus. Well, it turns out the resource allocation module of Wife 1.0 sometimes prohibits access to FreeSexPlus, particularly the new Plug-Ins he wanted to try. On top of that, Wife 1.0 must be running on a well warmed-up system before he can do anything. Although he did not ask for it, Wife 1.0 came with MotherInLaw which has an automatic pop-up feature he can't turn off.

I told him to try installing Mistress 1.0, but he said he heard if you try to run it without first uninstalling Wife 1.0, Wife 1.0 will delete MSMoney files before doing the uninstall itself. Then Mistress 1.0 won't install anyway because of insufficient resources.

Looking Glass

[pete-classic]

They're spending money on Looking Glass? I just went to the web site and they're still featuring the five-year-old demo video.

-Peter

Re:Looking Glass

[Unending]

It's worse than you think.
I worked on the project 3 years ago and it was a horrible mess.
They don't have any sort of 3D desktop concept all they have is a 2D desktop with 3D windows.
The underlying 3D system is impossibly complex and non-nonsensical.
Mouse clicks go through so many layers of checks that response time is ridiculous.
They are using Java3D, which is incredibly slow anyway.
To top all this off it doesn't look like they have changed anything in the last three years.
I might have a slightly tainted view and I haven't looked at the code in three years, but I'm still highly unimpressed.

Re:Looking Glass

[diegocgteleline.es]

As far as I know, that proyect is dead. The last time I read something about it, it was about some X.org extensions that were added to the core.

Re:Looking Glass

[fm6]

1998 called. It wants its Java cliches back.

Performance issues used to be a big problem with Java. That's long since been solved. The conventional wisdom was that these caused by Java being an "interpreted" language. That hasn't been true for a long time, and even when it was, it was only a minor factor in Java's performance issues.

Aside from the big overhead in firing up the runtime (still a problem, but not an issue for a service application, like Looking Glass) the biggest impact on Java application performance was bad source code compilers. Sun was in such a hurry to get the thing to market that all the early compilers were hastily adapted from C++ compilers, and created code that was inefficient and full of memory leaks — this on a platform that was specifically designed to make memory leaks impossible!

The Oracle acquisition is like a big second chance for Java, and a lot of other Sun technologies. Finally, they're under the control of a management hierarchy that doesn't consistently shoot itself in the foot!

Re:Looking Glass

[afabbro]

Finally, they're under the control of a management hierarchy that doesn't consistently shoot itself in the foot!

True, Oracle has better aim...they generally shoot their customers in the foot.

Re:Looking Glass

[not already in use]

The Oracle acquisition is like a big second chance for Java

Agreed. It would be nice to see Java evolve. Instead it's remained stagnant and is likely to become the next COBOL, something people will dread writing in for the next 30 years while other languages continue to evolve. C# started where Java left off, and is now leaps and bounds beyond Java in nearly every aspect. I think it was pretty negligent on Sun's part to let it get to this point.

Re:Looking Glass

[fm6]

Say what you will, but I run clean, mean Java applications every day. None of them are widely distributed (Sun really was too optimistic about Java's potential for standard desktop applications) so I can't give you an example you can easily run yourself. But if you ever have access to a Sun server built in the last couple years, check out the ILOM Java Remote Console feature. As robust and powerful as any native code terminal software I've ever seen.

Re:Looking Glass

[shutdown -p now]

Performance issues used to be a big problem with Java. That's long since been solved.

Some were, some weren't. The lack of stack allocation of small helper objects is still a hindrance, and, yes, the JIT can do escape analysis, but it's still very primitive in practice (and you really need copy semantics on language level to get the full benefits of that).

Aside from the big overhead in firing up the runtime (still a problem, but not an issue for a service application, like Looking Glass) the biggest impact on Java application performance was bad source code compilers. Sun was in such a hurry to get the thing to market that all the early compilers were hastily adapted from C++ compilers, and created code that was inefficient and full of memory leaks this on a platform that was specifically designed to make memory leaks impossible!

That's nonsense. It's pretty damn hard to make a bad Java compiler, given how limited the JVM instruction set is, and how many optimizations are actually done by the JIT. And how, exactly, can a bad Java compiler "create code ... full of memory leaks"? I'm not aware of any way to do that in JVM bytecode, unless you go out of your way to do so deliberately. Explain, please.

The bit about adapting from C++ compilers also sounds extremely dubious to me. Java is not at all like C++ semantically, and has only superficial resemblance syntactically. Meanwhile, a decent optimizing compiler is a very complicated beast, due to the complexity of the language, on all levels starting from the parser. It simply doesn't make any sense to try to "adapt" a C++ compiler for Java.

Re:Looking Glass

[cubiczee]

Did you know that java is often faster than C++ today? Did you know google uses java extensively for server applications?

Re:Looking Glass

[fm6]

I'm no expert on compilers. But my source certainly was: David Griswold, the original lead for HotSpot project. Now, this conversation happened more than a decade ago, so it's possible I misremember it. And of course it's always possible I misunderstood it. But I don't think I did.

Re:Looking Glass

[RAMMS EIN]

``Did you know that java is often faster than C++ today? Did you know google uses java extensively for server applications?''

No, actually I didn't. Do you have any references for those claims? They sound entirely plausible to me, but that doesn't make them true.

Re:Looking Glass

[Anonymous Coward]

1998 called again. It wants it's language back too.

Re:Looking Glass

[fm6]

I don't have any references either. But it was a design goal for the HotSpot virtual machine. The idea is that HotSpot optimizes code after the application has started running based on what code is being heavily used. Hence the name: VM identifies and optimizes execution "hot spots".

Does this work as well in practice as in theory? You got me. But in the case of C++ native code, you can't even do it in theory.

Re:Looking Glass

[binarylarry]

Really? Whats GCJ all about then?

They basically treat Java as a superset of C++, in the same way C++ is treated as a superset of C.

Re:Looking Glass

[RAMMS EIN]

``I don't have any references either. But it was a design goal for the HotSpot virtual machine.''

Was the design goal to yield faster programs than contemporary C++ compilers, or was the design goal simply to find a good compromise between compiling everything before running (long startup time but fast execution) and not compiling anything at run time (fast startup but slow execution)?

``Does this work as well in practice as in theory? You got me. But in the case of C++ native code, you can't even do it in theory.''

I have heard from several people that "Java is faster than C++". So I guess there must be some benchmarks on which some Java implementation has scored better than some C++ implementation. I would like to see those.

My comment wasn't about benchmarks, though, but about applications. And I really have yet to see an application written in Java that doesn't feel sluggish and a memory hog to me. I recognize that this is purely anecdotal and doesn't actually say anything at all about Java. It could just be that all Java applications I have seen are badly written.

As for C++ not being able to optimize hot spots in theory, I think you are quite wrong about that. As far as the theory goes, I think there is nothing that prevents a conforming implementation of C++ from doing so. In fact, you could do the same thing Sun's Java implementation does: compile to some non-native code, then do you best to write the fastest interpreter for that code, including run-time profiling, just-in-time compilation, and what have you.

In practice, I think it is more common to use profiling before creating the final, shipped binary. You compile your program, then you run it in a profiler (identifying the hot spots), and then you re-compile the program, using the profiling information to create a more efficient executable. This way, you get some of the benefits of HotSpot's run-time analysis and compilation, but with no run-time cost. The drawback is, of course, that you will do your optimizations based on how you _think_ your users will use the software, instead of based on measurements of how they are actually using it during the current run (as HotSpot does). Which approach yields the better results is an open question.

Re:Looking Glass

[Anonymous Coward]

The real issue with performance is actually AWT/Swing. It's single threaded nature makes it one big bottleneck, it was even so bad that in Java6 Sun admitted defeat and added SwingWorker to Java itself.

You need to do a lot of nasty hacks to make a decent performing GUI-app.

Re:Looking Glass

[SomeStupidNickName12]

Have you actually done any real development in Java or do you enjoy talking out of your ass?

Re:Looking Glass

[not already in use]

I've developed in java longer than I have C#. And yourself?

Re:Looking Glass

[fm6]

Was the design goal to yield faster programs than contemporary C++ compilers...

Yes. That was explicitly stated in early white papers and presentations.

Re:Looking Glass

[shutdown -p now]

I'm no expert on compilers. But my source certainly was: David Griswold, the original lead for HotSpot project. Now, this conversation happened more than a decade ago, so it's possible I misremember it. And of course it's always possible I misunderstood it. But I don't think I did.

I think you did, since HotSpot is not a source code compiler in the first place - it's a JIT. I can definitely imagine bits of C++ optimizer being used in that.

However, it's worth keeping in mind that HotSpot only appeared in Java 1.2. Before that, it was pure interpreted bytecode.

Re:Looking Glass

[shutdown -p now]

They basically treat Java as a superset of C++, in the same way C++ is treated as a superset of C.

They don't. They can map Java constructs to C++ constructs, which is possible to a certain extent (but definitely not 100% - think of GC, reflection, proxy objects...). This doesn't mean that Java has much in common with C++ - e.g. the object model is very different, generics are entirely different, C++ has nothing analogous to "abstract" for classes (not for methods!) or "final" for both.

Java designers themselves have repeatedly stated that Java roots lie in Smalltalk much more than in C++. C'ish syntax is only there to make the existing masses of C/C++ developers feel more familiar. Even then, if you know what it takes to parse C++, it's pretty obvious that a C++ parser is useless for any other purpose (hint: think about why things such as "typename" and "template" on function calls are needed in C++). The compiler would be pretty useless too, since 1) you don't need most of those fancy C++ optimizations (such as RVO or inlining) when compiling Java source code to bytecode, and 2) bytecode, which the JIT has to compile to native code, doesn't look anything like C++.

And, of course, Java is not a superset of C++ the same way C++ is a superset of C. It's trivial to point out numerous C++ features that do not have any Java counterparts: MI, templates, pointers and references, unsigned types...

Re:Looking Glass

[binarylarry]

That's good point.

Someone should fix the official GCJ FAQ:

Given that, and the idea that languages implemented using Gcc should be compatible where it makes sense, it follows that the Java calling convention should be as similar as practical to that used for other languages, especially C++, since we can think of Java as a subset of C++.

http://gcc.gnu.org/java/faq.html

Those nutty GCC developers, always putting up misinformation about the compilers they develop before consulting Slashdot.

Re:Looking Glass

[fm6]

HotSpot is not a JIT. A JIT is something that compiles an entire class when it's loaded. HotSpot only compiles hot spots.

And it is a compiler. Not a source code compiler, obviously, but the people who built it (I worked with them as a tech writer for several months) still considered it a compiler. Of course, parsing byte codes is not nearly as complex as parsing source code (you could even say it's trivial) but parsing is only part of what a compiler does. Just as important, and a lot more difficult, is the compiler's job of generating optimized object code.

The conversation about memory leaks occurred while Griswold was doing a "brain dump" of HotSpot's innards for the benefit of a video camera and some interested engineers. When he started talking about how HotSpot cleaned up orphaned memory, I asked him why there was any orphaned memory to clean up, since the Java memory model doesn't allow for programmer allocation of memory. He blamed the fact that current Java source compilers had been hastily adapted from C++. I assume that's long since been fixed.

Does that make sense in terms of compiler design? I wouldn't know. But that's what I remember.

However, it's worth keeping in mind that HotSpot only appeared in Java 1.2. Before that, it was pure interpreted bytecode.

Yes and no and that's beside the point. Yes, the original JVM was a pure interpreter. But no, the Symantec JIT was a standard part of the JRE by the time I started working at Sun in 1997. And beside the point, because the usual assumption that an interpreted program is always slower than a native code program is too simplistic.

A software application is a complicated beast, and code execution is only one of many potential bottlenecks. The one I've seen most often is a shortage of RAM, which causes a lot of virtual memory "thrashing". I used to see this a lot when I did SOHO consulting, and had the same argument with every single client: no, a faster CPU won't make Word run faster, try installing more RAM.

And of course memory leaks are a good way to bloat memory.

I actually ran into this issue when I was working on Java 1.1 documentation. (This is an example of memory bloat, but not memory leakage.) Part of my job was to generate the API documentation from the source code, using JavaDoc. When I took over this process, it took 12 hours to run on a Sun workstation. The writer who had done it before me just assumed it was the interpreter being a slow interpreter. That didn't make sense to me, but I didn't know enough about the technology to offer an alternative explanation.

Then I went to a talk by a guy who was developing big Java applications. One of his complaints about the current JVM was that it re-inspected every class file on a regular basis, which meant that every class file you loaded had to remain part of the program image throughout program execution. (He had a workaround for this that I no longer remember.) A light bulb went off in my head: JavaDoc gets class information by inspecting the class file itself (easier than parsing the source code, I guess). That meant that my JavaDoc run was loading every single class into memory! Its footprint must be huge! The workstation I was using had (I think) 128 MBs of RAM, not nearly enough.

So I moved the process to a server with huge amounts of RAM. Execution time went from 12 hours to 20 minutes. A problem that was being blamed on the fact that the code was interpreted was in fact sloppy class file handling leading to memory bloat. I think it very likely that most of early JVM issues with performance had to do with memory bloat. That could be true even if I'm completely wrong about the source compiler.

Re:Looking Glass

[shutdown -p now]

Given that, and the idea that languages implemented using Gcc should be compatible where it makes sense, it follows that the Java calling convention should be as similar as practical to that used for other languages, especially C++, since we can think of Java as a subset of C++.

You know the difference between "superset" and "subset", do you?

But let's move on.

Those nutty GCC developers, always putting up misinformation about the compilers they develop before consulting Slashdot.

GCJ developers are not qualified to make statements on whether Java is a superset or subset of C++, because they're only making a compiler for one of those languages, and not the languages themselves. If you want an authoritative reply, you ask C++ and Java language designers. I've already told you what Java language designers think about it (and you can easily find all the references you might need if you google around). If you want to hear the C++ side, post a question on comp.std.c++, and you'll probably get a few people on the ISO C++ WG reply.

Getting back to GCJ - in practice, if you've actually used CNI - its Java/C++ interop capabilities (which is the context in which they make this statement) - then you'd know that it doesn't provide a truly faithful mapping either way. You obviously can't call an arbitrary C++ class from Java, but even the other way around, the mapping is not perfect - e.g. it won't correctly map Java protected or package visibility specifiers from class members, and it won't handle Java generics well, either.

Re:Looking Glass

[shutdown -p now]

HotSpot is not a JIT. A JIT is something that compiles an entire class when it's loaded. HotSpot only compiles hot spots.

I've no idea where you've got that definition of JIT, but it's certainly not a common one. A JIT is a "just in time" compiler - that is, it generally compiles code at the moment of first use. That's all there is to it. It certainly doesn't have to compile code on class load (.NET JIT doesn't do that, for example), and it doesn't have to compile it at all.

And it is a compiler. Not a source code compiler, obviously, but the people who built it (I worked with them as a tech writer for several months) still considered it a compiler.

Of course it's a compiler. It's not a source code compiler.

When he started talking about how HotSpot cleaned up orphaned memory, I asked him why there was any orphaned memory to clean up, since the Java memory model doesn't allow for programmer allocation of memory. He blamed the fact that current Java source compilers had been hastily adapted from C++. I assume that's long since been fixed.

He was probably talking about memory used by HotSpot itself for compilation, then. In that context, it may matter that it was derived from C++ - for example, they might have used some allocation techniques where you just grab memory as you need and don't release it, and then release it only at the end of compilation unit, for example. Since JIT compilation is "contiguous", they might have had problems.

Or maybe it was just good old memory leaks in C++ code.

And beside the point, because the usual assumption that an interpreted program is always slower than a native code program is too simplistic.

Of course this is wrong; however, a well-optimized interpreted program is still slower than a native code program ;)

I understand that things aren't often CPU bound. But at moments when they happen to be, the difference can be striking, especially in terms of user experience on the desktop. There's a reason why Java has moved to JIT since then, and why .NET has been designed for JIT from ground up.

Re:Looking Glass

[fm6]

I've no idea where you've got that definition of JIT, but it's certainly not a common one.

I suppose I shouldn't be so rigid about how that term is used. My attitude probably comes from working with the HotSpot people, who always made a point of distinguishing between what JITs did and what their compiler did.

Of course it's a compiler. It's not a source code compiler.

Never said it was.

He was probably talking about memory used by HotSpot itself for compilation...

That's not at all how I recall the conversation. I honestly believe you're getting this issue wrong, but I don't have the background to argue with you intelligently. So let's just agree to disagree, shall we?

a well-optimized interpreted program is still slower than a native code program

Sigh. That kind of blanket statement is precisely the kind of oversimplification I'm trying to debunk. If execution of the program code is the bottleneck in a program, then yes, interpreted code is always faster than native code. But code execution is not always a bottleneck. I very much doubt that it was a bottleneck in all those early Java programs that established Java's reputation as glacially slow. Those would have been mostly simple GUI apps that didn't have to do anything compute intensive — there wouldn't have been enough interpretation going on to impact the application so thoroughly. Even if I'm totally wrong about the source compiler, you can't possibly blame the interpreter for all the slowdown. There must have been issues elsewhere (class loading? memory allocation? linking to native code?) in order to account for these issues. Even if I'm totally wrong about the source code compiler, it just makes no sense to focus so thoroughly on the bytecode interpreter when assigning blame for performance issues.

Re:Looking Glass

[shutdown -p now]

Of course it's a compiler. It's not a source code compiler.

Never said it was.

"the biggest impact on Java application performance was bad source code compilers"

That kind of blanket statement is precisely the kind of oversimplification I'm trying to debunk. If execution of the program code is the bottleneck in a program, then yes, interpreted code is always faster than native code. But code execution is not always a bottleneck.

That kind of blanket excuse if also a simplification. Yes, code execution is not always a bottleneck. Yes, much more often, I/O is. But, nonetheless, code execution adds up to everything else.

Also, you'd think that it doesn't matter, but OOD tends to result in code with multiple layers that call into underlying layers, etc. At this point, even the performance of those calls alone starts to matter eventually, simply because there are so many of them - IIRC, the stats were that your average "properly designed" OO application spends ~20% on methods calls alone (i.e. the instructions that make up the actual call, and not the logic in the method body).

Note that I'm still not saying that Java was slow just because it was a bytecode interpreter runtime originally (it's one of the reasons, but far from the only one). It also had a pretty crappy GC early on, and the bytecode execution wasn't optimized the way HotSpot optimizes code it compiles now (aggressive inlining, escape analysis and stack allocation, etc).

Another killer of perceived Java performance was Swing, pre-1.5. You simply couldn't watch those slowly-redrawing drop-down menus in a Swing application without tears.

Re:Looking Glass

[fm6]

"the biggest impact on Java application performance was bad source code compilers"

Notice the absence of "hotspot" from that sentence. How you could even think that I was referring to hotspot just totally baffles me.

Re:Looking Glass

[shutdown -p now]

Notice the absence of "hotspot" from that sentence. How you could even think that I was referring to hotspot just totally baffles me.

Looks like we're walking around in circles...

My original reply to that message of yours that I've linked to did not mention HotSpot at all. My reply to that particular phrase pretty much boiled down to, "you can't have a Java source compiler that affects runtime performance of the code in any significant way". From there, you replied that the information comes from the lead of HotSpot project, and then I conclude that the claims in your original post might actually make sense if they had anything to do with HotSpot (or Java's JIT compiler in general), but not Java source code compiler.

If you're willing to get back to the original claim that "the biggest impact on Java application performance was bad source code compilers", ad stand by it, then I don't mind starting back from there. :) You could, for example, point out some specific ways a Java compiler could non-deliberately generate bytecode that will cause performance problems or memory leaks at run-time. I'm simply not aware of any such, given the bytecode instruction set, even for a trivial compiler implementation (i.e. the one that just does a straightforward translation of Java AST to bytecode, with no optimizations whatsoever, not even constant folding).

and the problem would be... ?

[jipn4]

I think Sun has spent way too much money and effort on Java. The result has been a lot of bloated, badly designed libraries. If Oracle cuts that back, I think Java will actually improve technically.

Better fish to fry

[bogaboga]

I think Oracle should just strike agreements with *all* PC OEMs to have Java shipped with whatever OS these OEMs are loading on PCs. I know Java is Open Source Software and that those who need it can download it free of charge.

What troubles me is the exercise of having to repeat the installation procedure on so many machines. I recently installed 47 systems, six of them Kubuntu 9.0.4 systems with KDE 4.2.2. and the rest were Windows XP systems. It was not fun.

So to Oracle..."You have better fish to fry...now do the needful." It will not hurt you (Oracle) in any way.

Re:Better fish to fry

[Anonymous Coward]

So use a slipstreaming process to integrate it into the disk. Part of the windows integration you can specify some runonce commands on first logon and have it execute the commandline for the unattended java installer on a network or even put on the CD along with everything else.

Plus you can script the installer to be full automated with a one line batch file that you just execute.

Java is stupidly simple to install, hard part was getting what flags were needed.

Re:Better fish to fry

[mR.bRiGhTsId3]

Ah, but you still won't be able to get in the *buntu machines, because the ubuntus have backed themselves into a corner with the whole, one CD at all costs thing. The JRE is just too heavy to come on the CD.

Re:Better fish to fry

[Zarluk]

I recently installed 47 systems, six of them Kubuntu 9.0.4 systems with KDE 4.2.2. and the rest were Windows XP systems. It was not fun.

If it was the other way around (six windows machines and the rest Kubuntu ones), it would be completely different, I know ;-)

You could automate the linux installation, while dealing with the "insert disk,next, next, finish, reboot, (insert disk,next, next, finish, reboot)" windows cicle of installation, and see what system installs the first.

Re:Better fish to fry

[Anonymous Coward]

And if it was all Solaris, you could create one JumpStart server, an install script and you're set.

Re:Better fish to fry

[Washii]

Interesting. The current downloads I see for the generic Linux JRE (from java.com), ranging from x32/x64 RPM/BIN are all below 20MB.

Doesn't exactly seem that space-prohibitive for me.

Re:Better fish to fry

[ieatcookies]

Create vanilla install images that include java and use those to image the rest !

Re:Better fish to fry

[turbidostato]

"And if it was all Solaris, you could create one JumpStart server, an install script and you're set."

And if it was all Debian or Ubuntu, you could create one preseed server, an install config file and you're set.

That the parent poster didn't know about it only shows that people aren't born knowledgeable but nothing else. So, what was your point again?

Re:Better fish to fry

[Anonymous Coward]

What troubles me is the exercise of having to repeat the installation procedure on so many machines. I recently installed 47 systems, six of them Kubuntu 9.0.4 systems with KDE 4.2.2. and the rest were Windows XP systems. It was not fun.

If you were installing 47 systems perhaps you should have thought about getting an automated installer solution set up. Debian/Ubuntu make this quite easy and I know there are solutions for Windows XP. I'm assuming you had to install flash manually as well and there must be tons of other stuff.

Seriously, this is not java's problem, it's just your incompetence.

Re:Better fish to fry

[Anonymous Coward]

That Debian and Ubuntu don't have the type of ISV support in the enterprise software space to make them able to compete with RedHat and Solaris on servers.

Re:Better fish to fry

[turbidostato]

What's this? A "true Scotish" falacy? It's first time now that Red Hat is mentioned. Regarding specifically Debian vs Solaris, exactly what you achieve with Jumpstart is what you get with FAI or even with preseed (being preseed much easier). About ISVs, they are well and good... if I need what they offer. That's not the case in this instance.

JRuby will surely go.

[Anonymous Coward]

Put the money into JParrot.

Re:JRuby will surely go.

[argent]

Put the money into JParrot.

I'm sure there's a great Monty JPython joke in there, somewhere.

Re:JRuby will surely go.

[Random BedHead Ed]

JParrot's passed on! It's no more! It has ceased to be! It's expired and gone to meet its maker! It's a stiff! Bereft of life, it rests in peace! If you hadn't nailed it to the perch it'd be pushing up the daisies! Its metabolic processes are now history! It's off the twig! It's kicked the bucket, it's shuffled off its mortal coil, run down the curtain and joined the bleedin' choir invisibile!! THIS IS AN EX-JPARROT!!

How much of it is Open Source?

[argent]

One would hope that if the Open Source projects like Project Looking Glass, are worthwhile... they will be picked up by people who are using them. If they can open-source others rather than just killing them at least some can stay alive without showing up on the bottom line.

Re:Such projects perhaps should die.

[ShieldW0lf]

Oracle expects Sun to contribute to its operating profit right away. To make that happen, Oracle may pull funding and staff from projects such as JavaFX, Project Looking Glass, and Project GlassFish.

Ahh, but Oracle may decide to turn their offices into an exotic nightclub and force the engineers to work overtime as erotic dancers. You never know what they might do...

Speculation for nerds, stuff that's made up

History

[UseCase]

I remember thinking the same thing when Adobe bought out Macromedia. I think there is hope for some of the larger more useful pet projects but Oracles primary is making there a new acquisition profitable. Anything not strong enough to adequately monetize will probably be Open Sourced or shelf-ed.

So what observations can be made from other companies in our industry that have acquired companies with a strong library of technologies? What has lasted and what has fallen by the wayside historically speaking?

Applets are still used for games

[pjt33]

Applets may not be great for much else, but they are great for games. I don't care about JavaFX per se but it was at least partly responsible for recent work to improve the Java2D graphics libraries, and a lot of Java games would benefit from further improvements there.

Re:Applets are still used for games

[cdgeorge]

I don't think so. Even Yahoo is pulling the plug on their games (e.g. chess) and is using Flash instead.

Re:Applets are still used for games

[dshk]

We are not Yahoo but we are using applets to serve about 1 000 000 unique users per month. E.g. for chess.

Re:Applets are still used for games

[fm6]

I play a lot of games on the web, and I've never seen a really good Java game. Flash seems to dominate this market.

Re:Applets are still used for games

[thetoadwarrior]

Runescape is one of the biggest if not the biggest MMO games around and it's Java based. The publisher, Jagex, also builds numerous other web games in Java.

There's no denying Flash is big but applets haven't disappeared and if it wasn't for their horrible beginning, I think they'd be much bigger now.

So yeah, Java on the web didn't have a good start but it's not bad. In fact Quake 2 has been ported to Java with very good benchmarks. http://www.bytonic.de/html/benchmarks.html Anyone dissing Java now is just being a code snob or doesn't have the mental capacity to do something good with it.

Re:Applets are still used for games

[k8to]

Runescape is certainly NOT the largest MMO. It is the largest free-to-play MMO, and may be the largest web-based MMO.

Re:Applets are still used for games

[thetoadwarrior]

For 2007 it was 3rd which is certainly up there in "one of the.." territory.

http://gigaom.com/2007/06/13/top-ten-most-popular-mmos/

Perhaps their numbers have gone down some but they are still one of the big ones.

Java isn't (really) open source

[Anonymous Coward]

Java isn't really open source, that's why it's a huge pain in the ass under Linux.

Yes, Sun released a version of Java under the GPL. It's the "next" version of Java, Java 7 or 1.7 or whatever they're deciding to call it.

The current version of Java, the one that everyone uses, is most definitely not open source. It's free, sure, but it's licensed in such a way that Linux distributions can't package it. (Easily - some have worked around it, but the bottom line is that installing Java 6 on Linux involves an interactive process. It can't be automated.)

The "open source" version of Java is missing large chunks of Java and is basically not at all ready.

So, Java isn't "really" open source. It's "going to be" at some point in the future - or at least it was. With Oracle in control, who knows.

(Sure, Oracle can't un-open source what was released - but since that isn't enough for a full version of Java anyway, it's not like it really matters.)

Re:Java isn't (really) open source

[Anonymous Coward]

Get user-friendly packages of the open source version of Java 6 under the GPL2 with classpath exception here:

    http://openjdk.java.net/install/

That's Java 6, not "the next version".

There was no license to accept when I installed it on my Ubuntu box using apt-get.

Mercurial trees for both Java 6 and 7:

    http://hg.openjdk.java.net/jdk6/jdk6
    http://hg.openjdk.java.net/jdk7/jdk7

Re:Java isn't (really) open source

[LDoggg_]

Wrong. Wrong on every point.

Java isn't really open source, that's why it's a huge pain in the ass under Linux.

It's in the standard repos on most distros now. It's not any different than installing python or perl or any other language/platform.

Yes, Sun released a version of Java under the GPL. It's the "next" version of Java, Java 7 or 1.7 or whatever they're deciding to call it.

They released the 1.6 JVM and libraries that it was legally allowed to.

The current version of Java, the one that everyone uses, is most definitely not open source. It's free, sure, but it's licensed in such a way that Linux distributions can't package it. (Easily - some have worked around it, but the bottom line is that installing Java 6 on Linux involves an interactive process. It can't be automated.)

Everyone should be using 1.6 because 1.7 isn't released yet. 1.6 is GPL and open source.
That "interactive" process was clicking on the EULA before it was open sourced. Not that big of a deal then, but it's not even an issue anymore.

The "open source" version of Java is missing large chunks of Java and is basically not at all ready.

Big chunks? The JVM and libs were almost complete. The small parts that couldn't be released were 3rd library implementations that Sun didn't have rights to release as GPL. The GNU Classpath project filled in the gaps almost from day.

So, Java isn't "really" open source. It's "going to be" at some point in the future - or at least it was. With Oracle in control, who knows.

It is open source. Really.

(Sure, Oracle can't un-open source what was released - but since that isn't enough for a full version of Java anyway, it's not like it really matters.)

No, it does matter. Billions of IT dollars are still being spent on projects using the java platform.
I'm sure Oracle would have rather had Sun's implemenation all to themselves, but then they should have bought Sun a couple of years ago.

Re:Java isn't (really) open source

[hubert.lepicki]

Well, I AM using openjdk for commercial development with Java... it's 99% free software. Sound already works, GUI/3d works, web start... well, with 50% chance ;). No JavaFX yet but I'm sure they'll discontinue it anyway. The point is that Java (OpenJDK) is usable environment for developing and running GUI and web apps. Despite some missing bits it fullfills my needs in 100% without any closed-source blobs. It is not like with OpenSolaris (I feared that they'll go this way), where you can't use it without kernel binaries - here you can just grab a copy, compile and use.

Re:Java isn't (really) open source

[owlstead]

"So, Java isn't "really" open source. It's "going to be" at some point in the future - or at least it was. With Oracle in control, who knows."

I've just downloaded and compiled 1.7 (which is still rather far from being completed, but I needed the cryptographic changes such as no crypto provider signing). There was indeed a binary plugin needed with: a few classes that did SNMP). That's it. The rest is fully open sourced.

For those trying the same thing:
Compiling delivered 1.3 GB by the way (the delivered JDK and JRE are much smaller of course), and is not for the faint of heart. Use a RAM-disk to write to makes it go pretty fast (some tens of minutes in all on Linux). Try and use C++ caching or your compile will crawl through the C/C++ code.

Re:Java isn't (really) open source

[RAMMS EIN]

``I'm sure Oracle would have rather had Sun's implemenation all to themselves''

I'm not even sure about that. Oracle also push Linux, and though they contribute to it, most of the work is done by others. I am sure Oracle likes it that way. Why wouldn't they want the same for Java?

Re:Java isn't (really) open source

[Anonymous Coward]

Fascinating. Maybe you should ask the Apache Harmony folks how 'open' Java has really been under Sun's stalwart leadership. They have a slightly different perspective.

Re:Such projects perhaps should die.

[zoomshorts]

We never know what SCO could have done or should have done either.

Shall I hold my breath?

RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

HTML/CSS/JavaScript is an insufficient platform for Rich Internet Applications (RIA). Why do you think Flash is still so widely used? It's not just video. It's complex charting, graphics, animations, etc.

If you think Flash and Silverlight are just going to go away, or that IE and its non-standard compliance and lack of SVG are just going to go away, you're dreaming in technicolour. Web standards will eventually hit a wall.

I don't disagree that a lot of functionality (including video) can be implemented by all browsers that implement that new web standards, but it won't enough.

Besides, JavaFX has distinct advantages over Flash and Silverlight. It integrates seamlessly with server-side Java code. It also shares the same APIs with JavaFX Mobile, which allows mobile and RIA apps to share the same code.

Besides, do you really want the rich web to turn into a battle between two proprietary frameworks? Parts of JavaFX are already open source, and Sun is planning to open source the rest.

Re:RIA's need more than HTML5/CSS/JavaScript

[AndrewNeo]

Besides, JavaFX has distinct advantages over Flash and Silverlight. It integrates seamlessly with server-side Java code.

So does Silverlight with ASP.NET code, doesn't it?

Re:RIA's need more than HTML5/CSS/JavaScript

[AKAImBatman]

HTML/CSS/JavaScript is an insufficient platform for Rich Internet Applications (RIA).

There's a reason why I specifically mentioned HTML5. Video, Canvas, Audio, SVG, Networking, Storage, multi-threading, etc. The platform meets and even exceeds the Flash and Silverlight platforms.

This ain't your grandma's HTML, boay!

Re:RIA's need more than HTML5/CSS/JavaScript

[pallmall1]

Besides, JavaFX has distinct advantages over Flash and Silverlight. It integrates seamlessly with server-side Java code. It also shares the same APIs with JavaFX Mobile, which allows mobile and RIA apps to share the same code.

The JavaFX "advantages" are promises that don't yet exist (read the comments in the link).

Re:RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

This is true. However, the point is to give Java shops the opportunity to move away from Flex. Whatever Microsoft shops choose to do is not Oracle-Sun's concern.

Re:RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

There's a reason why I specifically mentioned HTML5. Video, Canvas, Audio, SVG, Networking, Storage, multi-threading, etc. The platform meets and even exceeds the Flash and Silverlight platforms.

Which browsers have fully implemented these? And how many corporations do you think will deploy intranet webapps that specifically omit support for IE?

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

Until Microsoft abandons them for their Next Big Thing.

Re:RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

There's a reason why I specifically mentioned HTML5. Video, Canvas, Audio, SVG, Networking, Storage, multi-threading, etc. The platform meets and even exceeds the Flash and Silverlight platforms.

P.S. Do development tools exist for these features? Flash/Flex, Silverlight and JavaFX already have development tools and IDEs.

Re:RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

Besides, JavaFX has distinct advantages over Flash and Silverlight. It integrates seamlessly with server-side Java code. It also shares the same APIs with JavaFX Mobile, which allows mobile and RIA apps to share the same code.

The JavaFX "advantages" are promises that don't yet exist (read the comments in the link).

What are you talking about? JavaFX can already use the Java API and it already has a mobile component. The author of the link you sent me only said that work remains to be done for tooling and performance. There was no mention of work needing to be done to integrate with the Java API or the mobile component, which was introduced with JavaFX 1.1.

Re:RIA's need more than HTML5/CSS/JavaScript

[Glonk]

Firefox 3.5 implements the majority of those. IIRC Webkit and Safari are almost there too, and even MS is making a sprint to HTML 5 -- IE8 has started implementation there.

As for editors, yes -- they are under development.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

Try Netbeans. It has kick-ass code completion for Javascript, CSS, and HTML. Including the HTML5 features. SVG is best created in a tool like Inkscape or some other vector drawing program.

I assume you know how to create and edit standards-compliant images?

Re:RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

IE 8 has specifically omitted support for SVG. Seems Microsoft has a conflict of interest in regard to Silverlight. There's no way an RIA application will be deployed if it doesn't support the browser with a 70%+ installed base.

Re:RIA's need more than HTML5/CSS/JavaScript

[javacowboy]

As much as I love Netbeans, I have to say that their JavaScript debugger is still a little rusty. Hopefully they'll iron out the kinks for 6.7.

Re:RIA's need more than HTML5/CSS/JavaScript

[squoozer]

...and we will get good cross browser support for all those features in 2196, until then the world will be running Flash, Silverlight and JavaFX in that order (shame beause I think JavaFX actually has quite a lot to offer). Even if cross browser support isn't required getting those features into IE in a reasonable time frame is not likely.

Re:RIA's need more than HTML5/CSS/JavaScript

[OwnedByTwoCats]

Yeah, but Silverlight and ASP.NET bind you tightly to the Microsoft server platform.

Server-side Java code can run on free stacks all the way down to the operating system, as well as on proprietary stacks. Silverlight/ASP.NET, not so much.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

There's no way an RIA application will be deployed if it doesn't support the browser with a 70%+ installed base.

You mean 66% and dropping like a rock?

Microsoft made their choices. And the market is not happy.

Re:RIA's need more than HTML5/CSS/JavaScript

[sik6si2o1faan6]

Don't know about Silverlight, but the Mono people say you can use Mono to run ASP.NET: http://mono-project.com/FAQ:_ASP.NET

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

No version of IE has a 70%+ install base. All versions of IE combined only account for about 70%.

Re:RIA's need more than HTML5/CSS/JavaScript

[Jack9]

I agree, but JWS gives you deployment of proper RIAs. Even flash is not enough, that's why there's AIR. JWS is better than AIR in many ways.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

only?

70% is still the majority...and will be for another 20% at least.

Re:RIA's need more than HTML5/CSS/JavaScript

[lahvak]

Lot of RIA are used internally inside companies. We are just migrating to a new webmail software, and we were told very clearly "if you use IE, install firefox, the application currently does not support IE".

Re:RIA's need more than HTML5/CSS/JavaScript

[Hognoxious]

do you really want the rich web to turn into a battle between two proprietary frameworks?

Hell no, I want mine to win, and the sooner the better.

Parts of JavaFX are already open source, and Sun was planning to open source the rest.

Fixed that for you,

          Larry.

Re:RIA's need more than HTML5/CSS/JavaScript

[iron-kurton]

JavaFX has distinct advantages over Flash and Silverlight. It integrates seamlessly with server-side Java code.

Yes, but use of standardized protocols such as SOAP almost makes this point moot.

It also shares the same APIs with JavaFX Mobile, which allows mobile and RIA apps to share the same code.

That is pretty cool

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

Oh yes, 4% less, that changes the whole picture!

Everyone's gonna develop using tech that doesn't work for 2/3 of users (only counting IE users here), right.

It'll be a LONG ass time before people even consider using those features. We basically just stopped catering to version 4 browsers, which had a ~1% market share for a darn LONG time.

Even if IE was hemorrhaging 10% of the market share a year (not the case, and it'll likely slow down too), it'll still take a good 7 years before it suffers the same fate as v4 browsers. That means year 2016. By then, we'll be using Windows 8 or probably 9... Things can change a lot in so many years.

I don't see html 5 going anywhere soon.

Re:RIA's need more than HTML5/CSS/JavaScript

[binarylarry]

The Mono people also say you can run .NET applications on Mono.

Unfortunately, Wine has better application compatibility.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

Maybe you don't use the features, but I fired my IE users. If they can't use my sites, too damn bad. And you know what happened? My brand new site just got featured on major news sites and is currently pumping 130,000 hits a day.

Boo. Hoo.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

No one's stupid enough to get rid of 2/3 of their visitors, much less without any big incentives. It's certainly not an option for most places (e.g. can you picture out newegg doing that and losing 2/3 of their sales overnight? Right, nobody can)

I call BS on your 130k hits. Anyone on the web is a millionaire porn star, has a 12" cock, and is the CEO of $LARGE_COMPANY.

Re:RIA's need more than HTML5/CSS/JavaScript

[digiti]

Another poster who never actually seriously tried FX!

We are serious Java and Flash experts and we spent a considerable amount of time with a direct line to the FX developers... Everything we were able to produce was just terrible.
Migrating skills from Java to FX is an exercise in futility, developers who had no Java experience had an advantage there. All us Java geeks were cursing the backwards way of doing things and the oddities, the IDE is ridiculous the live preview doesn't work and there is no way to visually build anything. Sure you can import really simple SVG graphics (not the elaborate stuff) but no animations and you need to start maintaining it in FX otherwise if node name changes well...

Communicating with the server, don't make me laugh! Did you try serializing an FX object to a server?
We were better off using plain Applets or even web start both of which now suck more because Sun spent so much energy on FX and botched the quality of Java 6u10.
Just read the Java developer forums where everyone is cursing FX. Other than Sun employee/contractors or people who didn't actually try FX no one is excited by this nonsense. Did you see the flashy demos, did you actually dig into the source code to see that most of them just use bitmap graphics for all the flashy animations???

Don't get me started on how ridiculous the mobile version of FX is.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

HTML/CSS/JavaScript is an insufficient platform for Rich Internet Applications (RIA).

There are some really powerful RIA Frameworks that are based on HTML/CSS and mainly JavaScript. Have a look at http://extjs.com/ or http://qooxdoo.org/. The biggest problem with these frameworks are video and sound, but I'd say that's not part of the everyday RIA.

Re:RIA's need more than HTML5/CSS/JavaScript

[JavaIsGreat]

Nop this is not true. I have used SilverLight as front end with interactivity written in JavaScript and Servlets as backend that generates the XAML. It works

Re:RIA's need more than HTML5/CSS/JavaScript

[sgt scrub]

Sun is planning to open source the rest.

Don't you mean, "Sun WAS planning to open source the rest". I can't help but have a wait and see view point on what Oracle will do with assets they just purchased as far as open sourcing goes. I hope they do but won't hold my breath.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

No one's stupid enough to get rid of 2/3 of their visitors, much less without any big incentives.

IE users made up less than a third of my traffic before I started firing them. Most of them happily switched to another browser. The 66% numbers LIE. You know why they lie? Because corporations have IE as the "Corporate Standard". Thus work PCs are driving up the numbers to unrealistic heights. The vast majority of web users have already switched their personal computers.

I call BS on your 130k hits.

Well that's too damn bad for you. Keep wasting your time and effort with an IE friendly site. Please. Your users will become my users when I offer them a better experience.

Re:RIA's need more than HTML5/CSS/JavaScript

[Anonymous Coward]

Except that SOAP sucks donley balls.

Re:RIA's need more than HTML5/CSS/JavaScript

[iron-kurton]

What are the alternatives then? JSON is nowhere near good enough for complex transactions, and proprietary protocols such as .NET's remoting or RMI are a bad idea.

I'm just happy that there even IS a standard we can develop against, and if you can suggest another one that is supported by most major platforms, I'm all ears...

Re:Such projects perhaps should die.

[AliasMarlowe]

...force the engineers to work overtime as erotic dancers...

How to lose staff and alienate customers?
Oracle has a track record of such brilliance.

It all depends.

[Anonymous Coward]

Oracle wants to make money so a lot has to depend on the financial viability of each project.

In my opinion the following will probably happen.

1. JavaFX - This will probably continue only because Sun has put a TON of effort behind it and contrary to other post it is not another Applet. Microsoft has Silverlight, Adobe has Flash and Sun/Oracle will have an open source alternative with a free development platform called JavaFX.

2. Glassfish - This will become the reference implementation for J2EE and Oracle will kill (as best they can) development on large scalability features out of it.

3. MySQL - Oracle will try and kill it the best they can. Their focus will be to convert these users over to the free version of Oracle's own DB.

4. NetBeans - This is a tough one. I am a heavy Netbeans user, but I see it being replaced by Jdeveloper now that Jdeveloper is free. I could see some of the features of Netbeans moving in to Jdeveloper. My hope is that they could take Jdevelopers speed and move it in to Netbeans :-)

5. Hardware - I would guess that Oracle will focus the hardware to run their Application stack the best. I don't see it going away any time soon.

6. Java - Oracle will do no better or worse than Sun in managing Java. Both companies have their issues and strengths.

7. OpenOffice - Oracle will probably keep some developers on this project. It will probably only get a small amount of love because of the revenue that it brings in.

8. Solaris - I see all of the cool features of Solaris (dtrace) moved over to Oracles Linux (if possible). That Linux version will run on "standard" X86 hardware or for "full support" it will run on the Oracle hardware (Sparc). This process will probably take 5 or more years.

9. People - It has been said but I expect over 10,000 employees from Sun to be let go.

10. Suns current CEO will go down in history as one of the guys who ruined a good company.

Re:It all depends.

[Bill, Shooter of Bul]

Well, They specifically said that they were going to keep funding mysql already, while specifically saying that they weren't going to comment on the future of open office. So I think that says something.

Re:It all depends.

[rackserverdeals]

8. Solaris - I see all of the cool features of Solaris (dtrace) moved over to Oracles Linux (if possible). That Linux version will run on "standard" X86 hardware or for "full support" it will run on the Oracle hardware (Sparc). This process will probably take 5 or more years.

That contradicts your initial premise and sounds more like wishful thinking.

All the cool stuff in Solaris is already in Solaris and making money. Yes people still use Solaris on new machines. Last quarter Sun sold almost $1 billion worth of SPARC based servers.

Since Solaris/SPARC is the leading platform for Oracle deployments, it makes more financial sense for Oracle to keep Solaris competitive against Linux in addition to it being cheaper than trying to integrate things into Linux.

Now that Oracle will have ZFS, I would wonder what might happen to Btrfs, which is an Oracle project.

Re:It all depends.

[Anonymous Coward]

Okay, $10000 question; Does Oracle in the longer perspective feel like paying for research and development for an entire operating system that is partly very modern, and partly creaking with age, when they might just as well join the resource pool around Linux? I foresee this post being labeled as "Troll" by Solaris loyalists, but the question is quite real.

Re:It all depends.

[VGPowerlord]

Glassfish - This will become the reference implementation for J2EE and Oracle will kill (as best they can) development on large scalability features out of it.

Just a nitpick, but Glassfish already IS the reference implementation of J2EE.

Having said that, I don't disagree that Oracle will stop adding features to it to encourage the use of WebLogic or Application Server.

Re:It all depends.

[Anonymous Coward]

In the second post-apocalyptic age of man, three groups rose to power. The Farmer Federation was run by good people -- well, mostly good. They'd thump your head with the Neoagrari Bible every Tuesday, Thursday, and Saturday, but they wouldn't cut off your toes or burn your granary.

The Survivor Stronghold was run by some mean men with diamond-chipped shoulders. They commandered the old castles from the first post-apocalyptic age, and they'd sooner shoot you than look you in the eye.

And then there were the Solaris Loyalists.

Java *IS* open source

[javacowboy]

You don't know what you're talking about.

Java is open source. Most of the source code for Java has been released under the GPL.

They started by releasing the JDK 7 code under an open source license. They then backported this code to OpenJDK 6 by removing some of the JDK 7 features and testing it under the JDK 6 TCK (testing kit).

The latest version of OpenJDK 6 is available for installation on Ubuntu and Fedora via their respective package managers.

The only parts of the proprietary Java 6 that are missing from OpenJDK 6 are:

1) SNMP code.
2) Applet/JavaWebStart code (although they're in the process of open sourcing it.
3) Latest bugfixes since JDK 6 Update 7 but these are slowly finding their way to OpenJDK 6.

Please do some basic research before posting your misconceptions as "facts".

Re:Java *IS* open source

[Xtifr]

You don't know what you're talking about.

Actually, it sounds like he does, since most of what you said agreed with what he said (as well as the facts).

Most of the source code for Java has been released under the GPL.

Most is not all. Java 6 is not open source, even if "most" of it is. Sun seems to get upset if you package up less-than-all of Java so "most" doesn't seem good enough.

The latest version of OpenJDK 6 is available for installation on Ubuntu and Fedora

The latest version of Java 6 is in the "non-free" repository on Debian and the "multiverse" (equivalent to non-free) on Ubuntu. Which you have to enable separately. GP may have exaggerated the difficulty of installing somewhat, but it's hardly automatic. And yes, there's OpenJDK, but we've already established that that's not the full system.

The only parts of the proprietary Java 6 that are missing from OpenJDK 6 are:

The parts that make it non-free? Look, we all know it's mostly there, inches from the finish line, but until the job is done, don't stand there in a flight suit proclaiming "mission accomplished", thank you very much. And don't sit around bashing those who point out that the job isn't done, 'cause they're right. You can say (and I would have agreed with you) that it's close enough for a lot of purposes, and it's easy to exaggerate the amount of work that remains, and making a big deal out of the amount that remains is silly, and all that, yes, but the job still isn't done!

Re:Java *IS* open source

[Anonymous Coward]

Once you've had Mac [apple.com], you can't go back!

I agree. Macs will fuck you over and/or dumb you down to the point where using a Real Computer would be impossible.

(Fuck you, Apple. GET LDAP SUPPORT WORKING CORRECTLY!)

Re:Java *IS* open source

[saleenS281]

SNMP and java webstart aren't "large chunks" and don't make the product "unusable".

Re:Java *IS* open source

[javacowboy]

You don't know what you're talking about.

Actually, it sounds like he does, since most of what you said agreed with what he said (as well as the facts).

He said that Java 6 wasn't open source and that it was next to impossible to install it on Linux. That means he doesn't know what he's talking about.

Most of the source code for Java has been released under the GPL.

Most is not all. Java 6 is not open source, even if "most" of it is. Sun seems to get upset if you package up less-than-all of Java so "most" doesn't seem good enough.

OpenJDK 6 passed Sun's TCK, which is what the JCP agrees needs to be done for a runtime to be called "Java". SNMP is not considered essential for this test to pass. The browser plugin is also not considered part of "core" Java. Therefore, Java 6 has been open sourced.

The latest version of OpenJDK 6 is available for installation on Ubuntu and Fedora

The latest version of Java 6 is in the "non-free" repository on Debian and the "multiverse" (equivalent to non-free) on Ubuntu. Which you have to enable separately. GP may have exaggerated the difficulty of installing somewhat, but it's hardly automatic. And yes, there's OpenJDK, but we've already established that that's not the full system.

What's your point? It's possible to run Netbeans and Eclipse on OpenJDK 6, which are two heavy client-side apps. Since OpenJDK 6 has passed the TCK, it's a complete Java runtime. The only things missing are SNMP and the plugin. Since IcedTea has implemented the Java plugin, it's possible to have the full Java experience with OpenJDK 6, which is available on the package managers of many Linux distros.

With the latest version of Ubuntu, it's possible to install the proprietary version of Sun Java 6 without tweaking any settings.

The only parts of the proprietary Java 6 that are missing from OpenJDK 6 are:

The parts that make it non-free? Look, we all know it's mostly there, inches from the finish line, but until the job is done, don't stand there in a flight suit proclaiming "mission accomplished", thank you very much. And don't sit around bashing those who point out that the job isn't done, 'cause they're right. You can say (and I would have agreed with you) that it's close enough for a lot of purposes, and it's easy to exaggerate the amount of work that remains, and making a big deal out of the amount that remains is silly, and all that, yes, but the job still isn't done!

Since it passed the TCK, the job is done. You can also get an open source Java browser plugin. I don't know what you're trying to prove.

Re:Java *IS* open source

[Anonymous Coward]

Where did you learn to read? I hope you're not typical of Mac users. The GP was specifically asking about the open source PROJECTS not the Java LANGUAGE.

Re:Java *IS* open source

[owlstead]

"Sun seems to get upset if you package up less-than-all of Java so "most" doesn't seem good enough."

Only if you actually *call* it Java. And with good reason. That's not protecting the source, that's protecting a trade name. You see this in many OS projects, e.g. X.org, and the many Firefox derivatives.

Microsoft happily used J# for their Java derivate product on .NET . Of course, anyone thinking that they actually were going to support that got screwed. Sun didn't do a thing about that, even after the law suit won by Sun.

Re:Java *IS* open source

[Anonymous Coward]

4) ISV testing / support

That should really be number 1 though. Outside of P2P file transfer... nearly all Java apps are implemented for business reasons, for which ISV support is no joke. Other side of that is OpenJDK is only used for running P2P apps on Linux.

Mod half flamebait, half insightful please.

Re:Mod parent down

[ShieldW0lf]

Congrats on being a position whore. Post directly to the story or at least find a related post.

Ok, now I'm pissed... you mean people actually get paid to do this? I've been acting like a position slut and giving it away for free! I wants my money!

Re:Such projects perhaps should die.

[tritonman]

Consider projects like the imageio. I believe this finally made it into the JRE (not positive though), but it is extremely slow. I tried using this for processing TIFF files and it was 50 times slower than anything else I tried.

Unlikely

[Eravnrekaree]

Pulling funding from JavaFX would probably be a bad idea since JavaFX is meant to keep Java competitive with .Net and Flash platforms which are rapidly taking over web and application markets. There is a big market in content design so it seems ridiculous to cut funding to those projects and they would shoot themselves in the foot. I see it more likely that MySQl is in danger, since this is heavily overlapped with oracles own database applications.

Re:Unlikely

[DuckDodgers]

The .NET and Flash platforms represent an enormous market.

But Oracle and the JavaFX community have their work cut out for them if they expect JavaFX to gain a real foothold in this market. Flash is ridiculously dominant.

I suspect the best we can hope is that competition from Silverlight and JavaFX forces Adobe to make Flash fully open source. I wouldn't be upset if JavaFX makes significant inroads, but I honestly don't see it happening.

Re:Unlikely

[BrainInAJar]

I see it more likely that MySQl is in danger, since this is heavily overlapped with oracles own database applications.

Just like they did with BerkeleyDB ... oh, wait...

Re:Unlikely

[shutdown -p now]

Pulling funding from JavaFX would probably be a bad idea since JavaFX is meant to keep Java competitive with .Net and Flash platforms which are rapidly taking over web and application markets.

Java was never competitive with Flash when it came to Web clients, nor was it competitive with .NET when it came to desktop applications (setting aside cross-platform issues). You'd have a point if they had a foothold to preserve, but as it is, JavaFX is struggling uphill all the way, and not even against Flash, but against Silverlight for the position of the "better Flash". And, so far, it's losing.

Re:Unlikely

[Anonymous Coward]

JavaFX was a lost cause before it even started - just another lame 'me too' from Sun. Pulling funding from a money vacuum with no realistic hope of getting market traction isn't a "bad idea" - it's just good business.

Re:Unlikely

[turbidostato]

"Just like they did with BerkeleyDB ... oh, wait..."

What do you mean "oh, wait"? Any serious development on BDB since Spleepy Cat aquisition? Is there any "flagship" for BDB except for OpenLDAP which is in competition with Sun's -now Oracle, identity management offers?

I won't foresay on the future of MySQL, but certainly there's no contradiction in "further development" and "make it fail"; that's called "designed to fail" and, were this the intention of Oracle it indeed would make a good strategy: if tomorrow Oracle announced "Madams, Sirs, the show ends here" regarding MySQL, tomorrow you'd have a fork and they'd lose momentum and grip so the day after tomorrow you'd probably would still have a flourishing MySQL ecosystem, only under a different name. If they follow development, only to let MySQL stagnate over some few years, they could manage to drive users slowly to a more insteresting field (be it free offers from Oracle as a "first dose for free" or even to Postgres as an entry point for "real" RBDMs to produce higher numbers of "real Linux/Unix DBA wannabes" that would look after Oracle on due time).

Re:Unlikely

[BrainInAJar]

they could manage to drive users slowly to a more insteresting field (be it free offers from Oracle as a "first dose for free" or even to Postgres as an entry point for "real" RBDMs to produce higher numbers of "real Linux/Unix DBA wannabes" that would look after Oracle on due time).

Well... Whether or not that happens aside, would everyone moving to a real RDBMS be such a bad thing overall?

Re:Unlikely

[digiti]

Since no one is using or interested in FX what you are saying is far more unlikely.

FX has damaged Java's existing client side developers (yes there were many) by pulling funding from many important tasks such as the native look and feels which still don't work properly.

Its publicly despised by most Java developers who tried working with it and designers who compare it to Flash are just shocked that something as godawful as this would be released as a "Flash competitor".

Please actually try a technology before buying into the marketing nonsense that Sun has been spewing on FX. Our team did try this to produce demos and we were shocked at how horrible it was (I am talking production 1.0 and 1.1 versions!).

Re:Unlikely

[turbidostato]

"Well... Whether or not that happens aside, would everyone moving to a real RDBMS be such a bad thing overall?"

Surely not but don't think Oracle would seek that because "it is a good thing". It's my opinion that all big for profit software companies, specially once Microsoft showed the road go or will go into "dumbing down" their products, not only because that broadens their potential market but because the dumber the technician and decision makers in generl, the more important becomes marketing and the less the technical merits of the product (wasn't it Larry Ellison the one saying more or less that?). If you are already a big company this is a good thing for you.

Oracle's New 'Hardware' company

[mpapet]

I would argue hardware is where Oracle saw the value in Sun. IMHO, there should be a home for Solaris at Oracle simply because it's a strong, viable server OS.

History has shown Sun has terrible problems running open source projects larger than their own paid contributors. I don't see Oracle improving or even interested in this.

Most of Sun's software projects will fade into oblivion as GPL'd abandonware because nearly all of them are also-ran projects started as Sun's version of things like Flash.

Two wrongs make enterprisy things.

[Anonymous Coward]

Can't wait to see where Oracle takes Java.

Hopefully somewhere deep in the woods for murder-suicide pack.

No.

[nurb432]

They will pull funding on all of the projects.

Got your copy of the code you need yet? Better do it soon.

Project Kenai

[multipartmixed]

I just hope they don't go pulling the plug on Project Kenai.

Kenai is Sun's version of SourceForge/GitHub/Google Code. I'm hosting a project there and it works well enough, a few minor tweaks and it will be fantastic. I chose it because they had bugzilla, mercurial, forums with feeds and a rudimentary wiki with syntax I didn't hate. And a low-barrier to entry (I am more than capable of setting all that stuff up myself, but I'd rather spend the time hacking code).

Funny, though, I only just realized why I must have received that "please evangelize Kenai!" message in my inbox this morning...

Re:Project Kenai

[greg1104]

Does it make a profit? Doesn't look like it could, so expect Kenai to get canned. The discussion of the merger over at The Register has a good intro to the per-project profit focus of the very well managed Oracle.

Re:Project Kenai

[Anonymous Coward]

As the engineering director for Kenai it's nice to see it make on slashdot (in a positive mention no less...)

To your comment "a few minor tweaks and it will be fantastic" - can you let us know what tweaks you're looking for and we'll see when we can get them in (in an upcoming sprint).

Matt Thompson at sun dot com

Open Source Java

[Anonymous Coward]

Now that Java is open source, I expect Oracle will let the open source crowd access any of the pet java projects.

There will be the "Industry Java", well funded, and of course, a few little open source java implementations, free class libraries, and other small open source projects for niche markets, or those who can't afford a compiler, etc.

Oracle needs to focus on profitability. Abandoning java to the free but high-quality open source programmers will allow Oracle to focus on their core competencies.

oracle + sun = bad

[Anonymous Coward]

how long before Oracle won't certify on hardware that is not SUN? They already cut VMWare out of the loop by not certifying ESX as a viable Virtualization platform for Oracle Ebiz, etc.

No more Wonderland?!

[Anonymous Coward]

Oh no... and Wonderland 0.5 was about to come out.

OpenOffice

[crow]

It's all speculation at this point, but the project I'm concerned about is Open Office.

Exotic Nightclubs and Erotic Dancers

[Anonymous Coward]

Scott, Jonathan

I can't believe this! My plans for our premises and staff have been leaked all over the Internet already. I still think your engineering talent will make much more money through our Exotic Sunset Clubs than they ever did from Java (especially with the tips). We've just got to keep the lid on this, any more leaks and we'll be seeing it as SCO's recovery plan.

Later

Larry

Java is sun's pet project

[heroine]

That's like saying Oracle is going to pull Java. The very reason we all have to use Java for our day jobs is because Sun keeps pumping money into odd projects & promoting it. Everyone likes designing Java API's but Sun is the only one making progress on implementation. How many API's would get implemented if Sun didn't spend the money? Android still has just 1 implementation.

Makes sense..

[Anonymous Coward]

Sun got involved in many open source projects without a clear idea business model on how to monetize the projects. So I wouldn't be surprised if Oracle pulls the plug on a lot of these projects if they believe it won't contribute much to their bottom line.

Luckily Java is GPL'd

[ManWithIceCream]

I'm so happy that Sun managed to release Java under the GPL before it was bought. I can't imagined what would happen if Oracle got a hold of Java without the community being able to fork it if nessecary.

Hmmm; Lets see

[WindBourne]

Oracle has been pushing Oracle on Linux for more than several years. They have made it their preferred OS. Oracle's stated goal is not just to own the server, but the desktop space. In server space, Linux is now more prevelant than is Soliar. In addition, it has more in-roads into the desktop than Solaris.

But you think that Oracle bought Sun for its hardware and will now switch to Solaris, and drop Linux? Is that correct?

Re:Hmmm; Lets see

[Anonymous Coward]

In server space, Linux is now more prevelant than is Soliar.

Not in mission critical environments. You know, where it counts.

But you think that Oracle bought Sun for its hardware and will now switch to Solaris, and drop Linux? Is that correct?

If they're smart they'll support both. Linux on the low end, Solaris in the mid to high end.

You know, your little linux fanboi shrieking does not change the fact that Solaris on Sun hardware is light years ahead of anything linux has to offer.

As a datacenter professional of over twenty years, I've ripped an assload of Microsoft and Redhat out and replaced it with Sun/Solaris and IBM/AIX. They just work.

Re:Hmmm; Lets see

[mpapet]

Oracle's stated goal is not just to own the server, but the desktop space.

Sun doesn't have a more useful desktop in any way. Who knows though, there have been plenty of dumb decisions in corporate America, thinking Sun's desktop is viable may be one of them.

But you think that Oracle bought Sun for its hardware and will now switch to Solaris, and drop Linux? Is that correct?
It's not that simple. Solaris has more features for big iron. For commodity hardware, Linux absolutely rocks. IMHO, they can easily support both Linux and Solaris on the Sales/Service side and not confuse anyone or cannibalize either customer base. But they desperately need to do a better job at running OSS-like projects if they want Solaris to stay vaguely relevant.

Re:Hmmm; Lets see

[rackserverdeals]

They day they announced the merger, Oracle CEO Larry Ellison said that Solaris/SPARC was still the biggest place where Oracle was being deployed.

With all the pushing that Oracle has been doing towards Linux, their customers don't seem to be following. The question must have been raised whether Oracle should be leading their customers in OS choice, or following.

Oracle won't drop Linux, but they'll probably do more what IBM does. They'd rather sell you on AIX but if you want Linux you can have it.

Re:Hmmm; Lets see

[rackserverdeals]

Sun's desktop strategy has moved from creating a desktop OS, to creating an OS that developers would prefer and their thin client and application delivery system.

SunStudio and DTrace for developers and Sun Ray for other people's desktops. In addition to remote desktops to thin clients they have technology that allows you to run linux and windows applications servered remotely within you Solaris session.

Desktops in the corporate world are more than just the physical machines and OS's these days. Sun's Identity Management software is very good too. Oracle has done a bit in this area and combining the work of both companies could really help out.

Re:Hmmm; Lets see

[WindBourne]

I think that you are a lot closer now than you were the other day. Back then you were indicating that Oracle would drop Linux. I give ZERO chance of that. I also think that there is ZERO chance that they will drop Solaris. So as I said before, I think that Oracle will order the sales ppl to start supporting Linux. Sun's current crop HATE Linux and only support as about to lose the sale. In my humble swag, I think that Oracle may favor Linux, but, I think that they will offer up both willingly.

Re:Hmmm; Lets see

[rackserverdeals]

I think that you are a lot closer now than you were the other day. Back then you were indicating that Oracle would drop Linux

I never said that what I said was "They will support Linux but will likely push Solaris". Being their second most popular OS, it wouldn't make sense to drop Linux

I think they might decide to switch their development platform back to Solaris though.

Right now for Oracle, it seems to be Solaris/SPARC and Linux/X86 but I think it makes more sense now to support Solaris/x86 more than they have in the past as a platform for Oracle and I hope they keep Solaris/x86 releases of Oracle up to date.

In the last thread, you said the Sun reps gave you a hard time about deploying Linux on Sun hardware. Were you going to do it on SPARC or on their x86 servers? If it was the SPARC kit, I can understand their reluctance, while Linux on SPARC is possible, it's not as well supported and there doesn't seem to be much ISV support. If it was on x86 hardware, I can see them wanting to push their own OS.

If you go to HP or IBM and want to get their Unix hardware, they're going to push HPUX or AIX respectively, I don't see that as any different.

ZFS

[axehind]

I wonder if they will finally release ZFS under the GPL?

By way of example

[pjt33]

One really good Java game.

Re:By way of example

[fm6]

I'm underwealmed. Very slow startup (somebody doesn't know how to manage resources effectively), very crude graphics, and it took me only a few minutes to get bored. It starts out trying to teach you how to cast "fireballs", which amounts to a cutsey implementation of the that lame old computer game artillery. The only way to figure out the right trajectory is trial-and-error, and after a dozen errors I lost interest.

If this is the best Java gaming has to offer, then I rest my case.

Seemed inevitable to me

[mpfife]

I hate to say it, but when Sun went ga-ga over Java a few years back - I knew it was a bell for this very day. Exactly how were they going to make a revenue stream on such a software platform of applets? They did get big, they did fill a need, but they also never lived up to real profitabilty. They saw their market and revenue shrink until this day when Sun is bought by a database company.

Sun made is fortunes on vertically integrated stacks of hardware and software. But it all started to slip away. One might argue that's no longer a viable business plan - but IBM seems to be doing ok. Sun still makes great hardware, but Intel's bitten off more and more of their server market for years as desktop hardware is getting powerful enough. Witness the rise of low-cost PC solution in Google data centers.

As to the software, yes, Java was a great solution and created a great, easy-to-program vertical stack that bridged awful heterogeneous hardware and databases. But it was slow. Horrible slow when you have mission critical servers that need to handle really big, popular websites or god forbid high-speed transactional services. The applet model just didn't catch on as they probably expected.

Say also what you will (flame on) to base your revenue model on open-source just smelled of death from day one. Yes, it *seemed* to fit with Sun's vertical marketing. They poured tons of money into Java that at first seemed really well spent, seemed to fit the virtualization model things were going towards, their vertically integrated solution, and they designed a great language. But commercial open-source only survives on a service model - and when everyone can use Java - why pay Sun the premium? They just didn't come up with enough compelling reasons to pay them after the platforms were in place to stay afloat. They did a great service to the world buy funding and designing Java, but it also seems to be the very thing that lead to their demise. Sure, they probably would have lasted far less long if they'd stayed proprietary - but I just never saw how they could keep their business afloat going open-source.

As a final point, I noticed that in the late 90's early 2000's, every college taught all their classes in Java. I went back to my alma mater (Purdue) and many other top-tere uni's are now pretty exclusively all back to C++/C for their classes. We no longer hire Java coders here where I work either. You need to know C/C++.

Maybe this is just another story of SGI - the business model moved on, and it was time to go. SGI tried to stay proprietary, Sun tried to go Open-source - but both didn't work in the end because the computing evironment had changed so much that the business model itself just wasn't viable.

Re:Seemed inevitable to me

[rackserverdeals]

You had so many things wrong in just your first paragraph, I couldn't read the rest.

Re:Seemed inevitable to me

[Parker Lewis]

"But it was slow. Horrible slow when you have mission critical servers that need to handle really big, popular websites or god forbid high-speed transactional services. The applet model just didn't catch on as they probably expected"

Maybe you should learn more to talk about Java on web. JSP and Servlets is the core of a lot of serious system, running 24 x 7. Probably, your bank is using Java on the website. And Java is just GREAT when running critical missions. Or a bank will not use it. This is why Java becomes the industry standard. Maybe your company still uses C++ for some tasks, I can imagine, as a lot of languages, has a niche. But you really missed the point about Java.

Natural monopolies

[pseudorand]

I think that the entire computer industry is a Natural Monopoly. Think about it. Intel, IBM and Sun basically duplicate a heck of a lot of work trying to do the same thing: make silicon add faster. Microsoft and the Linux community likewise do the same thing with operating systems. Flash/Silverlight/JavaFX/GWT: redundant. PHP/Rails/J2EE/.Net: redundant. And it's not just the companies that duplicate efforts. All their users and those who develop for these platforms duplicate efforts as well.

These technologies aren't identical. They each have some nice features that appeal to some small subset of the user and developer bases, but the majority of uses can be implemented on whatever technology stack is available. The differences either force developers to learn multiple ways of doing things and learn about multiple sets of quirks of each tech stack, or get tied to one particular stack and risk having major parts of their skill die if that platform looses support.

That said, I think Oracle should probably dump as much of Sun's crap as possible. A platform either needs 110% support, needs to implement lots and lots of features, needs to fix bugs as if the earth itself depended on it, needs to be cheap and easy, or needs to not waste everyone's time in the first place.

Think about the PC. Microsoft and Intel got the 10% right that did 99% of what 99% of potential customers needed. It was affordable, it worked well enough, and it was easy to use. Since then, they used their economies of scale to take off, and now, for all the academic imperfection of i386, it's a really good, really stable platform. Even Windows (save some Visa graphics drivers) is a pretty stable platform by this point.

True, competition from AMD made Intel get of their ass with x86_64, multi-core, and virtualization, but AMD did it all while maintaining 100% backwards compatibility with x86. i386 is like an open standard that anyone can try to do better than Intel. If they do, they make money and Intel plays catch-up. If they don't KEEP doing it better... well, they they can just get some dumbass CEO to buy a shitty GPU maker.

True also that perceived competition from Linux has probably made Microsoft work a bit harder and cleaning things up. But in reality, Linux just proved that the *nix market really IS a natural monopoly and killed the other *nixes. Linux hasn't truly provided competition to M$ yet because it's not trivial to port an app from Windows to Linux. Wouldn't it be nice if there was just one OS to worry about (or better yet, one browser).

The true problems with technology today are: 1) Providing fast, reliable (read: clustered/redundant) full-featured (no limited SQL or proprietary database access methods) access to server-side data without compromising security on the server and without bumping into firewalls while forcing the end-user application developer to have to learn as little as possible about how that works. No specifying port numbers. No knowledge of http caching. Transparent (to the developer) statefullness. Minimal, simple interface for handling failed requests that uses the programming language's own error-handling mechanism (i.e. exceptions).

2) Full-featured client-side programs (read: 3D, video, multiple windows, local filesystem with appropriate security and space limitations) that can be written in any language, work on any client platform, don't compromise security on the client even though you're running untrusted code and don't require deployment of a browser plug-in or other runtime environment.

Oracle should aim big and do that, and keep it simple and easy instead of trying to tie developers to their language (Java) and/or platform (.NET).

Re:Natural monopolies

[RAMMS EIN]

``Intel, IBM and Sun basically duplicate a heck of a lot of work trying to do the same thing: make silicon add faster.''

But the competition between Intel and AMD is what has been driving CPU speeds up.

``Microsoft and the Linux community likewise do the same thing with operating systems.''

Again, there is a lot of progress being made because of competition between the various camps. Good features are being cloned, new features are being added, and performance and security are being improved.

``Flash/Silverlight/JavaFX/GWT: redundant.''

Perhaps. Or maybe we'll see some good come from it yet.

``PHP/Rails/J2EE/.Net: redundant.''

Rails gave a major boost to the web development landscape. Competition from .NET has caused Java to improve massively.

``And it's not just the companies that duplicate efforts. All their users and those who develop for these platforms duplicate efforts as well.''

This is true, and a lot of time and effort has been wasted. On the other hand, one has to wonder if we would really have been better off if that hadn't happened.

The trick is that having multiple projects that cater to the same market allows evolution to occur. Each project can move in its own direction, and the world will vote with its feet and cause the projects that make the best choices to thrive. This functionality can then be incorporated by the competing projects, or not, as they see fit. I do believe we end up with better results this way than if we wanted to always avoid duplication.

Re:Natural monopolies

[Anonymous Coward]

Current options are so lame (ie java, net, flash, etc) that we are currently using delphi/c++/realbasic plus dcu/activex (and lots of controlers to standar interfaces) and bundler (ie molebox... kind of a lite thinstall) to keep it: easy to develop, flexible and fast to grow in unexpected directions. As allways, the secret its to have the levels of abstraction right in some kind of architecture. Its not ideal, but gets "on time", "on butget" and "easy to develop"

Re:Natural monopolies

[Anonymous Coward]

Interesting post, but I would argue that precisely because there's a high redundance in the suplying of those big markets that there is no monopoly... Just an 'unstable equilibrium'. There is basically two paths for Oracle, now that they have almost reach their monopoly goal on the 'applications development stacked platforms' market: work for the evil, as previous ATT, IBM and Microsoft once were, or work for the good, like Apple, Microsoft (in the 80s, against evil IBM), and Google were.

Given previous history, there's no reason why once you conquer a so much desired monopoly status, you'll abandon it. In fact, given IBM's previous attempt, Sun's buyout seems more of a "secure position" kind of move rather than a "aggressive attack and kill" one.

Changes nothing

[curtix7]

If sun was hurting but wanted to stay on their own they would have to cut back on less profitable projects anyways. (I might be wrong but I think they were already doing so recently)

They'll die

[extremescholar]

Oracle will stop funding all "non-value added" work. If it doesn't contribute to the bottom line, it's gone. Look at their recent acquisition of Primavera. they used to have good support, great customer service, and a reasonable Annual User conference. They're short changing support and more or less cancelled the user conference. It's all about the dollars; and those project don't generate any.

and?

[KZigurs]

Hallelujah?

Re:Java really is open source, but not for long

[Anonymous Coward]

Even if your whole argumentation is 100 % true, it seems to me that you don't get previous poster's points. An application developer in mainly, after all, an user of late Sun's products. Just as your own users: it doesnt matter if 'JVM and libs were almost complete': either they are 100% complete, or they are not. You can't miss, for a clean install of Eclipse as an example, 20 lines of code. And it's not the same at all to say "Billions of IT dollars are still being spent on projects" rather than "Billions of IT dollars are currently being spent on _open source_ Java projects".
The point is: Oracle will terminate Java's openness as soon as it realizes it is not financially profitable. For them, the sooner, the better: Sun had everything to win from an open source platform, but what's there for Oracle ?

It'll be like this: Oracle retires funding from open source JDK 7, which might be released or not. It stops maintaining Sun's decades old JVM, but releases it under a new name, say "Oracle Security Proof VM", which is exactly the same as Sun's, but with a few "extensions" that make it easy to program securely for their OC4J certified stack. Because lack of funding, the whole professional developers community massivly migrates to OSPVM (there's a certified certification course that you can buy at OTN's, which comes with a certified discount coupon when you order a certified database with it), and there's no point in running current Open JDK 8.1.0.23rt applications on it (they are probably incompatible, even if Oracle says they are not - they just aren't certified). And so, Java is dead: first as a profitable product, then as an open source platform, and finally, as a second class discount offer. I don't know who killed it 'mostly' - was it Sun, or it's users, or the FSF ? Oracle gave the fatal shot.
Don't believe me ? Some words for ya: PeopleSoft, BEA's WebLogic, Orion web server, Red Hat Linux.

I'm telling you all: Java is dead, I've already started looking for a (mediocre) DBA course - I already had the feeling that DBAs earned more than I did, with less than half of a developer's effort and competences.

How about MIDP3?

[edxwelch]

Maybe MIDP3 would be one of the projects pulled as well? MIDP3 is the new version of J2ME (java for mobile phones). The commitee has being working on it for 6 years, with still no sign of ever seeing the light of day. Meanwhile, J2ME increasingly becomes less popular as the API has become dated in comparision with new platforms. It would make sense to put it out of it's misery at this stage.

Re:Applets

[presidenteloco]

The only thing wrong with applets is that they were hatched in 1995 with two serious limitations.

1. They had no effective way of caching code or data on the client machine (in a safe sandbox) and so had inherent performance limitations and interminable initial download waits for anything serious.

2. Microsoft successfully killed them by refusing to properly support (and default-install in the OS) a Java runtime environment.

If those two facts had been different, we would be living in a world of powerful rich-client web apps with Java on server-side and client-side.

Applets are here to stay

[Billly Gates]

They are now AJAX and flash driven.

Java was not ready while flash and shockwave were. I am hoping javafx takes off. The problem is that many new computers sold do not have a jre installed but come with flash.

Javafx should have come out back in the late 1990s when flash was starting to gain attention.

I am hoping Java will come back as Microsoft is slowly taking over java in the server room and adobe flash/air is all the rage. Silver light may take off in a few years after it becomes standard with new computers.

I am still bugged MS bought their way to depreciate the (applet) tag which led to Java's demise.

Applets could have been something if sun would ever allow the applets to run native and not be re-compiled into byte code on spot while the user waited on their 486 with a 14.4 modem. That was truly ugly and nearly killed it on the desktop.

My guess is silverlightFX or something will come out next and will be AJAX like and support HD video and other things and so will AIR in the next 5 years. Lets hope JavaFX does make a little comeback. It does not suck that much today with core2duos and broadband with 3d cards.

At least Microsoft didn't buy them out

[Billly Gates]

I think web developers and /.ers would take up their pitchforks and the future of the internet could be at stake itself.

Java does have potential and I think someone at Sun should explain to Oracle that raising hte costs and putting up barriers to make higher RIOs would actual hamper Java as many hobbiests and college studnets will learn c#.net instead of java.

Flash?

[Parker Lewis]

JavaFx is to compete with Silverlight, not .Net. And which kind of killer application on Flash are we talking about? Youtube videos? How about the next generation browser with embbeded video players? Flash was a "must have" in the 2k year, in the .com bubble era. With the "web 2.0" (I hate this term) using Ajax, Flash has not real advantages in a rich web application.

mixed feelings

[gt3]

At least we'd still have Adobe Flex, which is far more advanced and future proof than JavaFX would ever likely be and i doubt Microsoft would ever kill Silverlight. If JavaFX dies, after the applet thing failed, well, that'd just be sad.

Its Virtualbox that I will miss

[Chrisq]

A great way of running multiple OSs, much easier than the competition.

ORCL share holder, I want them to cut funding to J

[Anonymous Coward]

As an ORCL share holder, I want them to cut funding to non-profitable projects once it is clear there will never be a profit from them that makes up for the prior costs.

BTW, I hate java. It is just as it was in 1994-ish when I first saw it. Java is slow, uses too much memory and the average java programmer knows almost nothing about the OS. Writing a cross platform C++ program isn't very hard. We were doing it in 1993 for 8 platforms including win32 and MacOS. By 1997 it was fairly easy to be cross platform with C++. I don't code anymore, but it must be extremely easy now, unless MS screwed it up further - which is likely. The Visual C++ compiler always was "different" when compared to others, but at least it wasn't as strict as the HP-UX aCC compiler.

Now that Java is OSS, there's nothing Oracle can do to kill it, which is good. Also, Oracle enterprise software is tied to Java, so it won't go away quickly enough.

Anyway, ORCL should kill off non-profitable efforts.

Re:Java really is open source, but not for long

[Anonymous Coward]

Oh, God, not another Python freak. Get back in your cage or we'll beat you with your leash!

To the adults who have read the parent post: Java's not dead. The very WORST that will happen is that Java will fork, and development will be taken over by the IcedTea people at Red Hat (if for no other reason that annoying Ellison is good for your Karma). There are far too many interested parties for Java to die. And it's GPL, whether this ding-bat understands that or not.

You can't unring a bell.

Hey! You! Python Freak! What did I tell you! Back! (CRACK!) BACK! (CRACK!) BACK IN YOUR CAGE! (WHACK, CRACK, SMACK!)