Roumen's Weblog

You say you agree with most of Jacek's points...care to elaborate on the points you don't agree?

Posted by Marc on březen 16, 2007 at 03:32 odp. CET #

... and I agree with most of the criticism, especially about the editor in 5.5 (which is going to change for 6.0 :)

Posted by Roumen on březen 16, 2007 at 04:19 odp. CET #

Interesting article indeed. Some things stuck out for me.

1. Platform neutral installers are good and enable consistency. Tailoring installation to an operating systems packaging is a nice to have but not a large feature that has ever decided the IDE choice for me. You can argue it makes it available to install easier. And what would you choose for linux: .deb, .rpm (3, 4, etc.), .tgz, etc., all?


2. Plugins for NetBeans I find to be more consistent in quality and delivery. There is less of a plugin maze to navigate (you could say that is because there are fewer plugins, but that is rapidly changing). There are some movements underway to avoid the plugin nightmare you can get into when an IDE community grows rapidly. (See Too many plugins too many places). Though the community is not as large yet, the point of NetBeans is that you don't need a massive amount of plugins to get started with all types of development. IMHO, Eclipse needs a large community and/or commercial entity like MyEclipse to approach the same feature availability out of the box.


3. The point of a foundation is not that important if you can participate openly in the development and direction of the technology in question. Essentially, as long as you act like a foundation. With the release of the JDK as GPL and the consistent delivery of NetBeans/GlassFish under open licenses such as CDDL, it would quite easy to say NetBeans/Sun is starting to act like a foundation. Sure some formalization and expansion on the concept can be done, but I don't think the community is currently missing out.


4. Enterprise support. I completely disagree with the Enterprise support statement in the article. IntelliJ, though an excellent development environment falls short here. Enterprise Java means to many Java EE, Web Services, effective UML, SOA, etc, not just EJB. A consistent approach to this large area of focus is what NetBeans + Packs provides. (Making all of the packs available from the update center would be nice).


5. Broad plugin support for open source libraries as a option for your core features like JSF development (MyFaces for example). True this is a nice feature of environments like IntelliJ. However I think it is feature/approach that is needed when you have a commericial IDE fighting for mindshare with Eclipse and NetBeans. You need to support as much as possible to continue to have a draw for developers.

I'm looking forward to the IDE landscape getting shook up once again as NetBeans 6.0 hits maturity. Competition can be good.

Posted by Chainreaction on březen 17, 2007 at 03:05 odp. CET #

We indeed had some discussions about this topic. There are both advantages and disadvantages to creating such a foundation - and currently we see NetBeans community growing even though the project is mainly backed by Sun. We might reopen this discussion in the future, but I think current model works quite well for both developer community and Sun. We will be doing other things to encourage larger participation - dual licensing is one of the options we are considering. We also created a new plug-in portal on netbeans.org which will make it easier to both contribute and find plug-ins. Finally there's the NetBeans Dream Team which is all about reaching out the larger developer community.

We will see, I think the question would become important if we wouldn't see NetBeans community growing, but in last 2-3 years it has been growing steadily and the community is creating support for non-JCP & non-Sun technologies (Spring & Hibernate plug-ins as examples - or various web frameworks whose support is developed on java.net).

I personally think what really matters is the number of users of NetBeans, if this number is really large (it's in hundreds of thousands now so I'm talking about millions), it will pay of to everyone to create plug-ins for NetBeans. Some companies are realizing it today, but for the largest ones it will take some time, when the size of NetBeans community reaches some critical value. One thing we really don't want NetBeans to become is another Eclipse - we want to be different (mainly in the "works out-of-the-box" and "well integrated" stories).

Posted by Roumen on březen 17, 2007 at 07:22 odp. CET #

An interesting article indeed. I agree with about all that is stated in it. I do certainly agree with the fact that Eclipse seems to have more plug-ins available for the various de-facto frameworks like Spring and Hibernate. But that is its weakness as well, which is also stated in the article, the varying degree of quality of the plug-ins. The out-of-the-box experience of NB is excellent and many of my Eclipse-aficionado friends have moved to NB because of this. Unfortunately, as stated by somebody else in this thread, only JPA and EJB are taken into account. The SOA hype is getting huge and support in this area is important to make a decision on which IDE one should decide. This is also true for MyFaces support and the lack of it in NB. MyFaces is huge and with Oracle's contribution of ADF Faces to Apache it is one of the most feature rich component libraries of JSF. Sun, even with Woodstock, can't get close to that. They need a lot of work to do in this department. Unfortunately Collaboration has been taken out of the equation as well. Although this is a separate plug-in to be downloaded from the AUC. It would be nice to have a download of NB 6 with all feature packs (I know the new installer and that is great) but also with the plug-ins that are otherwise to be downloaded from the main AUC, those of which you pretty much know that they're of a very high quality. Iwan

Posted by Iwan Eising on březen 19, 2007 at 12:49 odp. CET #