Netbeans 6.0 beta and Glassfish V2 released today

Discussions

News: Netbeans 6.0 beta and Glassfish V2 released today

  1. Sun has announced the release of Glassfish V2, their open-source application server (and the reference implementation of Java EE), along with the availability of the NetBeans 6 beta. However, as this newspost is being written, Your Humble Editor notes that the websites for both applications show no changes from last Friday, when he refreshed both applications locally. This news post will change slightly once the new pages go live. [They're up now.] Glassfish V2 adds clustering and a good bit of performance to prior releases; it's done very well on the SpecJAPPServer benchmarks. It's supported by Sun, as the Sun Java Application Server 9.1 release. Glassfish' adoption has been climbing quite well; Eduardo Pellagri-Llopart characterized it as providing WebLogic's power at JBoss' price, which your Humble Editor found to be a very interesting statement, for a few reasons. For one thing, in advertising, comparisons mean very specific things. Saying "Foo is as good as Bar" is actually an oblique compliment of Bar; it means that Bar is good enough that comparisons to it elevate Foo. Therefore, Sun's complimenting WebLogic. This isn't a bad thing, it's just interesting, considering how many "market leaders" the space has. Likewise, the JBoss' price point comparison is also interesting, and indicates how important JBoss has been, historically, to the space. NetBeans 6 adds multiple language support (including Ruby), a much faster editor, local history of file changes, integrated support for Subversion, and a built-in profiler (instead of mandating that the profiler be downloaded as a separate product.) NetBeans' Ruby support has been gaining mindshare in the Ruby community, according to Sun - meaning a Java editor is being used to write Ruby. They did not evoke a comparison to Eclipse when saying this. :) Ruby in NetBeans also provides an easy integration path for Ruby in a Java application server. Glassfish V2 is provided as part of the NetBeans download (as is JRuby); one can write a Rails application using NetBeans, then deploy it in a Java application server, and add value to it through the use of Java APIs as desired.

    Threaded Messages (22)

  2. I think this issue is rather important. I would guess it would break other ejb2 apps than ejbca? https://glassfish.dev.java.net/issues/show_bug.cgi?id=2992
  3. That may well be, but it seems to me that it's not like they've ignored the bug. Looks like they're on top of it, but realized that by "fixing" your bug, they may introduce others. Looks like some sensitive code there. I've had a couple of issues pushed to P4, to be fixed after the release. Perhaps not as showstopping as your issue, but they certainly aren't minor issues. But, I would certainly admit that while I play in that sandbox every day, these issues are edge cases doing things outside of the mainstream. If I've found anything, these guys are on top of the bug lists, and have been very responsive. I'm just glad they've released. Looking at how fast the RC's have been coming it out, it's nice for them to get the thing out the door. Because the best thing about it now is that if you really really want, you can get paid support for it. Before it was beta, so there was no "support". But now it's released, bugs and all, and if you want your bug potentially fast tracked, then you can get a support contract. I think paying customers get more votes on the bug database than non-paying ones. And that's not some grand back room scheme or conspiracy. It's not a bait and switch thing, it's just reality. Now, we get to move on to the v3, speed it up some more, and make it more pluggable.
  4. You are completely right, I would be the first to admit that they are really responsive and on top of reported issues! All credits to the Glassfish team for this. I'm just putting a little pressure to get my bug fixed :-) Cheers, Tomas
  5. I'm just putting a little pressure to get my bug fixed :-)
    Thanks for reporting the issue. An update release for GlassFish V2 is being planned. See http://wiki.glassfish.java.net/Wiki.jsp?page=PlanForGlassFishV2UR You may send your requests to USERS at glassfish dot dev dot java dot net and DEV at glassfish dot dev dot java dot net. The following blog also explains GlassFish roadmap: http://blogs.sun.com/pelegri/entry/future_glassfish_releases
  6. I think this issue is rather important. I would guess it would break other ejb2 apps than ejbca?

    https://glassfish.dev.java.net/issues/show_bug.cgi?id=2992
    Eek! That is surprisingly bad! I am really surprised that they would release a product that loses env-entries on an EJB like that. I guess maybe this is a wait until the first SP release. That's ok, though, I am glad to see v2 out there anyway.
  7. I think this issue is rather important. I would guess it would break other ejb2 apps than ejbca?

    https://glassfish.dev.java.net/issues/show_bug.cgi?id=2992
    Are you sure that your application is still broken? A fix was put in so that the env entries don't disappear anymore. As far as I'm aware, the only reason that the bug is open is because the existing code introduced a performance regression that we'd like to fix in an upcoming release. But the env-entry semantics should be working.
  8. It's done[ Go to top ]

    They just released them! :) http://www.netbeans.org/community/releases/60/index.html https://glassfish.dev.java.net/downloads/v2-b58g.html
  9. Woohoo! NetBeans is my favorite IDE, Glassfish is a great app server. Great news!
  10. Re: Glassfish[ Go to top ]

    I am going to, for a brief moment, ignore Joe's request for no self-promotion and refer to my blog on the momentous release of Glassfish v.2: http://douglasdooley-astrosolutions.blogspot.com/ This is a huge day for Sun and a big day for the rest of the application server contenders, as I believe portability will lift all those committed to choice, diversity, and Java... douglas dooley
  11. I like NetBeans. I build TestMaker on it. I've been waiting for Subversion support. Where are the Glassfish V2 performance benchmarks posted? -Frank Cohen http://www.pushtotest.com
  12. Where are the Glassfish V2 performance benchmarks posted?
    See http://blogs.sun.com/theaquarium/entry/883_66_jops Here are few GlassFish V2 launch blogs: http://blogs.sun.com/theaquarium/tags/glassfishv2
  13. Congrats on the release! I actually managed to find some time to test the clustering. Setup was as easy as they come, and the whole things seemed to be a really clean implementation instead of an after thought. And the docs were superb -- just the high availability section was 230 pages, and very well written. About the "WebLogic's power at JBoss' price" claim ... Glassfish V2 is a surprisingly strong showing, and if i were BEA i'd feel somewhat worried. However, what sets WebLogic apart is not necessarily features, but the fact that it's battle proven. Only time will show how robust Glassfish will be in production.. Still, well done guys, really impressive! /Henri Karapuu
  14. However, what sets WebLogic apart is not necessarily features, but the fact that it's battle proven. Only time will show how robust Glassfish will be in production..

    Still, well done guys, really impressive!
    Thanks for the kind words. FYI about GlassFish code base... It is fairly old. Perhaps "V2" is misleading. GlassFish V2 corresponds to version 9.1 of Sun's Application Server. GlassFish project was started on June 6, 2005 with Sun's Application Server 8.x. The original 7.x (Sun ONE) project started in 2000. We are collecting GlassFish adoption stories in a group blog called stories. Sun who initially donated the code base for GlassFish takes the sources and productize it under the name Java Enterprise System with other middleware software from Sun. It is a yearly subscription-based services model. Currently, there are about 1.7+ million subscribers (see slide 17) of Java ES with many mission critical deployments. Now, you may also buy subscription-based support for just the Application Server (GlassFish V2). See http://blogs.sun.com/theaquarium/entry/global_on_the_web_purchasing
  15. 2 Mistakes[ Go to top ]

    I am a fan and past consulting service vendor to the Sun application server team, so I don't mean to write this with any malice to the team. It seems to me that Sun is making 2 mistakes in the Glassfish v2 announcement on TSS: 1) While it is great that Sun improved performance, the real news of this release is that the basis of Sun's enterprise services is now ready for production. I see very few customers doing a straight application server to app server comparisons anymore. Instead they are looking for a stack that gives them application development, SOA, database connectivity, Spring support, Web service interfaces, process management, and workflow. They want and need the whole solution. Sun's achilles heal until the v2 release was the poorly performing app server. v2 puts Sun in a much better position to talk about the high value things they deliver. 2) Talking about GlassFish as a mature product is a losing battle. Sun has made so many changes to the product name (way back to Netscape Enterprise Server, and anyone remember iPlanet) that it has no credibility to talk about GlassFish as a mature product, even if it is true. Sun has delivered an excellent, well performing application server with an excellent IDE. The news needs to be "It's time to take a fresh look at Glassfish" and stay away from the "project started in 2000" lines. -Frank Cohen http://www.pushtotest.com
  16. Time to take a fresh look at GlassFish[ Go to top ]

    "It's time to take a fresh look at Glassfish"
    Agreed, thanks for the corrections and kind words. BTW: This blog provides some good context for GlassFish V2. And, here is the official message
  17. Re: disagree[ Go to top ]

    Frank, Your message is more convulted than Sun's, in this case. Yefim and Gartner have been trying to convince us of a stack, starting with Portals and EAI and continuing with SOA, but the bottom-line is that app servers are core to operations...app server v. app server comparisons serve everyone except the WebLogic product-team, for it will demonstrate that pricing is the only core difference between the two...I'll agree with you that the support contracts are truly the major news of this announcement, but I fail to see where they fell down in promoting this... As for heritage of an app server, this is mute, but worthy of a comment from the Sun product team, for they have been hard at work for taking back a market that is naturally theirs since 2000...Netscape, iPlanet, Sun ONE, Java Enterprise System, and now Glassfish may confuse the MarCom folks, but it is fairly clear that they have succeeded in becoming relevant again even as the vast majority wrote them off for dead back in the go-go days and continuing up to modern times...its worth pointing this out... I understand that the Sun product team may not want any vitriolity on the basis of their announcement, but that is part of the reason I don't represent them, I'll just call out detractors that the Sun app server program is the JEE 5 leader, by a long-shot, IMO...not that you are a detractor, but I fail to see your points on performance, stacks, and heritage...this launch was executed succinctly, and it would appear that everyone else is playing catch-up because of it...
  18. Re: diagree[ Go to top ]

    Hi Douglas: My view of the world: developers have moved on from straight comparisons of app server to app server. They now want to evaluate development and deployment stacks for building apps, services, etc. That makes me want Sun to concentrate on announcing the whole stack and not the individual parts. That makes me want Sun to give me a method to evaluate the whole stack and not the individual parts. For instance, what is best built on Sun's stack? And how do I evaluate it for performance and scalability? Less convoluted? -Frank
  19. and dev prod[ Go to top ]

    ... and how do I evaluate it for developer productivity? -Frank
  20. Re: Sun's app server business[ Go to top ]

    Frank, While I can concede that the developer ranks want developer productivity, that is basically encompassed in JEE 5 and NetBeans messaging...also, Sun announced the 'stack' with Metro and openESB, and I believe the SDK is shipping with an option for openSSO... The real story here and for Sun going forward is deployment, and utilizing the decade-long value proposition of relatively portable apps to move off of proprietary app servers...here, I feel like they have hit the mark dead-on with low barriers to entry and volume pricing... I understand where we are having this convo., on TSS, so there is some consideration for start-up time (which is improved and pretty good from what I read), Spring and Seam support, and benchmarks...however, that is not what should be emphasized in this launch.... Sun needs to move Sun customers off of other app servers (read: WebLogic) and on to Glassfish/9.1 to maintain account control, and grow Java OS-like implementations, and not be so damn focused on Solaris...the Glassfish team hugely accomplished this through messaging and mechanisms to the extent that the current conventional wisdom on app servers no longer applies....
  21. NetBeans 6.0 and refactoring-move[ Go to top ]

    Hard to believe. Even NetBeans 6.0 doesn't provide refactoring for moving class members! I'd really like to know what the justification is for not supporting such a basic refactoring functionality.
  22. BEA SpecJAppServer2004 results[ Go to top ]

    Would someone with Sun pleease confirm the SPECJAppServer2004 results for BEA WebLogic Server 10.0? I see that WLS has the top 3 high scores. It seems that they retain bragging rights for performance on HP, sun, UltraSparc, AMD, and Itanium machines. The other issue I am wondering about is scaling to larger systems with SPECjAppServer2004. Blake Connell at BEA tells me you can't take a single server score and extrapolate to a larger system. I would be interested to learn what TSS community members are running their app servers on? Is it Sun's JVM 1.5, Sun JVM 1.6, JRocket? The scores for Glassfish and WLS appear to be on different VMs. Last question is about the SPEC submissions. Why did Sun not submit a GlassFish v2 SEPC submission (instead of their own commercial implementation of it)? -Frank Cohen http://www.pushtotest.com
  23. Re: BEA SpecJAppServer2004 results[ Go to top ]

    Would someone with Sun pleease confirm the SPECJAppServer2004 results for BEA WebLogic Server 10.0? I see that WLS has the top 3 high scores. It seems that they retain bragging rights for performance on HP, sun, UltraSparc, AMD, and Itanium machines.
    It depends how and what you look at, but BEA does have those three good scores. SJSAS continues to have the highest score on a single T2000, and Oracle has some categories in which they claim the highest score. SPECjAppServer is a system-level benchmark, so the system is what's important. If you want to compare appservers, you can only compare similar configurations. Saying that you have the best AMD score (or whatever) means only that the particular hardware, database, and everything else you used as a whole is better than the whole system anyone else used. It says nothing in particular about whether the appserver software, or the hardware, or the database, or anything else from one result is better than or worse than another result with a different configuration.
    The other issue I am wondering about is scaling to larger systems with SPECjAppServer2004. Blake Connell at BEA tells me you can't take a single server score and extrapolate to a larger system.
    That's correct, you cannot. Many things change as you add load on the systems -- in particular, when you add lots of appserver load, you still have a single database, which must scale accordingly. And the data sizes are all completely different, which affects the appserver.
    Last question is about the SPEC submissions. Why did Sun not submit a GlassFish v2 SEPC submission (instead of their own commercial implementation of it)?
    Because SPEC submissions can only be done on commercial products; the software used for the submission must have a 24x7 4-hour response time support option. And strictly speaking, the Sun Java Systems Application Server is not an "implementation" of glassfish V2 -- it's a packaging of it. Not that I want to cut hairs...