BT

Java EE Guardians Unite to Save Java EE

| by Monica Beckwith Follow 641 Followers on Jun 17, 2016. Estimated reading time: 3 minutes |

In the wake of last year’s downsizing in Oracle’s Java evangelism as well as their earlier announcement that they would be suspending future major releases of GlassFish server and limiting support, a group of Java standard bearers calling themselves the Java EE Guardians formulated a charter declaring their intent to come to the rescue of Java EE.

The Java EE Guardians are a veritable who’s who of Java luminaries, including “Father of Java” James Gosling, former evangelist Reza Rahman and many other Java technorati.

Group founder, former Java EE/GlassFish evangelist Reza Rahman says:

We now have close to 400 members in our Google Group and close to 2000 followers on Twitter.

While employed at Oracle, these foot soldiers endeavored to keep Java moving forward. However their statistics point to a serious downturn in Java EE development. The site currently includes a selection of charts illustrating the consistent down-trending of resolved issues and revision control commits.

In Oracle’s defense, the company seems to be paving a road to their vision of a cloud-based world of the future.

Although the Java EE Guardians agree that the cloud should be pursued, the site points out:

Despite all this there is growing evidence that Oracle is conspicuously neglecting Java EE, weakening a very broad ecosystem that depends on strong Java EE development. Unless things change soon Java EE 8 won’t be delivered in anywhere near the time when it was initially promised if it is delivered at all.

It is very difficult to determine why this neglect from Oracle is occurring or how long it will last. Oracle has not shared it’s motivations even with it’s closest commercial partners let alone the community. A very troubling possibility is that it is being done because Oracle is backing away from an open standards based collaborative development approach and is instead pursuing a highly proprietary, unilateral path.

The Java EE Guardians' charter seek a more “rational strategy”, and wants support from the industry to “solving this problem”:

As committed as we are we still need to effect change in a company with the size, scope and resources of Oracle. Persuading Oracle to adapt to the legitimate interests of people outside of itself – even its own customers – has proven challenging in the past. In all likelihood it may not be easy this time either, though there must always remain plentiful room for reasoned optimism. The Java EE community needs your help. We need the support of your voice and perhaps your volunteer time if you can afford it to ensure we succeed. These are the many ways you could help.

From Oracle’s point of view, it would seem that these and other Java technologies may just not be worth supporting. JavaFX still hasn't caught on as expected and NetBeans IDE still only enjoys a 10% market share. InfoQ asked Rahman about the value of investing in technologies that seem to have trouble attaining escape velocity:

I can tell you Netbeans is a nice polished IDE. In the right hands, NetBeans can be a success.

At inception, JavaFX was not a polished API, and continues to be not as polished as it could be. But it is not so bad that it should be abandoned. Your car’s dashboard could be running on JavaFX; also your browser. Java needs a front-end technology, and JavaFX with a little TLC addresses that void.

Mobile is the same, it has its problems, but it is worth moving forward, and there should be a set of companies willing to move that forward because there is a problem space for it.

Rate this Article

Adoption Stage
Style

Hello stranger!

You need to Register an InfoQ account or or login to post comments. But there's so much more behind being registered.

Get the most out of the InfoQ experience.

Tell us what you think

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Email me replies to any of my messages in this thread
Community comments

My observation exactly by Richard Richter

I was checking JPA 2.2 development and their Jira and it is literally (I mean literally literally) dead. This is a big shame. Even with Spring and others around, Java EE still kinda bound EE things together (sure, we're not talking about EJB 2.x).

You Can Help by Reza Rahman

As our website clearly points out, there are many things that can and are being done about this. One of the easiest things anyone can do is simply sign our petition to Oracle executives on change.org: www.change.org/p/larry-ellison-tell-oracle-to-m.... We already have 1200+ signatures and growing. This will send Oracle a clear signal that this is an issue that needs to be resolved quickly.


It is also important to note that while we would all like to resolve this via the JCP with Oracle, major vendors are already talking about vendor-neutral options outside the JCP without Oracle if needed.

Is JEE still relevant? by William Smith

I’ll admit to not being terribly current - I’ve not done JEE development in some time so genuine question - is JEE still relevant?

I can’t help feeling that as more and more software moves towards microservices the old style app server model that JEE was built around just isn’t that important any more. Spring kind of disrupted itself with Boot and Cloud Foundry, but I don’t really see an equivalent of that on the JEE side. And I realise that Microserices aren’t right for everyone or every app but I just don’t see what JEE fits in or needs more development effort.

Re: Is JEE still relevant? by Reza Rahman

We knew this question would come up and that's why we addressed it at the very top of the Java EE Guardians site. Java EE (including all of it's APIs) has an enormous impact on the entire ecosystem. Indeed the EE APIs collectively really have the largest footprint on the server-side Java ecosystem. The easiest way to see this is this wide ranging, very well-run survey from DZone.


In my view Java EE is still the most important, most easy-to-use and only multi-vendor technology stack. As you can see from the numerous comments in our change.org petition that has already exceeded 1200+ signatures I am hardly alone in this view. I think Java EE is very well suited for both microservices and the cloud. If you are looking for a Spring Boot equivalent in the Java EE realm, there is plenty going on including WildFly Swarm, Payara Micro and KumuluzEE. On the cloud front EE centric options are even richer including OpenShift and BlueMix.


It is also important to note that Spring Boot itself actually supports pretty much every Java EE API except for CDI and EJB 3. Indeed, Boot is essentially built on top of the Servlet specification. CloudFoundry has even better support for vanilla Java EE applications.


There really is very little logical sense in what Oracle appears to be doing at the moment. It's not even in their own best interest let alone the industry.

Re: Is JEE still relevant? by William Smith

Thanks for taking the time to reply to this.

Why is it Oracle’s responsibility though really? I mean if they don’t see JEE as being in their commercial interest anywhere shouldn’t other vendors who have made large amounts of money out of the spec over the years - IBM, Red Hat etc - step up to the plate.

Re: Is JEE still relevant? by Reza Rahman

No problems at all. Having rational discussions is what all of this is about.

Firstly it's important to understand that executives are human too. They can make mistakes, be poorly advised and miss important perspectives. The reality is that Java EE is in fact very much in Oracle's commercial interests because it is such a critical and pervasive part of IT now and in the future. That's the feedback that we are in effect in the process of providing Oracle, very publicly. The petition is easily the most prominent part of that feedback.

Secondly, all of the JCP executive committee members have indeed stepped up and are actively seeking solutions from all angles. I encourage you to follow and participate in this thread on our Google Group pertinent to the matter.

It is important to understand though that at the end of the day Oracle is the Java steward. While it shouldn't be solely responsible for Java EE, Oracle is where we need to look first as the leader in moving Java forward. In the least they need to tell us all they are willing to cede at least part of their responsibility as the steward (which by the way is fine by pretty much everyone).

Re: My observation exactly by Lukas Eder

In the case of JPA, that's not too worrying. The industry trend is moving towards using SQL along with functional programming idioms, not stateful, object oriented ORMs. JPA is around 10 years behind.

Re: My observation exactly by Reza Rahman

For one objective, data-centric perspective, here is some more DZone survey data that shows the usage patterns for persistent tools. JPA is far and away predominant at almost 70% usage. Taking ORM usage in general into account, the usage is the neighborhood of 90%. I have not seen any concrete data-centric evidence beyond sporadic subjective opinion to suggest the overwhelming predominance of JPA/ORM will abate any time in the near future. I think this excellent analysis piece by by Martin Fowler explains well both the predominance of ORM adoption as well as the long continued pockets of ORM resistance.

P.S.: For the sake of full disclosure, Lukas is the owner of the company behind jOOQ. jOOQ positions itself as a rather vocal alternative to JPA/ORM. In the survey, jOOQ adoption sits at 1.5%.

Thomas Kurion shooting from the hip by Hank Miller

I've heard from insiders that Thomas Kurian is mercurial, changing his mind like Imelda Marcos changes her shoes. He's reputedly irresponsible and more concerned with holding onto his job than producing anything of value.

Re: Thomas Kurion shooting from the hip by Reza Rahman

You'll notice that Thomas Kurian is one of the Oracle executives cited on the petition...

Re: My observation exactly by Lukas Eder

Thanks Reza. You've been linking to that single survey for a couple of times now, and I still wonder about its significant flaws:

- The sums don't add up to 100%
- There's no "other" bar to show how many people use something entirely different
- There's no indication as to how many people combine APIs (e.g. they do a lot with jOOQ + Hibernate)

In any case, have you tried adding up JDBC + JDBCTemplate + MyBatis + jOOQ + Spring Data (+ that missing "other" bar)? And add to that the potential of using native SQL via JPA and Hibernate? All of these people use SQL, and even in this interesting survey that you keep pulling out, those are more than JPA users. But hey, the only statistics that I can trust are those that I falsified myself, right? :) You got the biggest bar in this particular statistics.

I'm not so sure why you try to pull the "jOOQ" card here. Yes, jOOQ reaches 1.5% in this survey. I've seen other surveys where it reaches 8%. In any case, it is close to the order of magnitude of other APIs you're offering here, so perhaps jOOQ/SQL does matter? Anyway, I was only advocating SQL, not jOOQ with my comment, but thank you for advertising it. Much appreciated!

Re: My observation exactly by Reza Rahman

I think the data, graph and analysis linked is pretty self-explanatory for most people. Beyond that I don't have much else to say. I'd love to cite other surveys but I honestly don't know of any other major ones run by a reliable source like DZone. For surveys the reliability of the surveyor really does matter.

That said, there are other ways to get some decent indicators. Here is a Google Trends graph that shows essentially the same patterns in play. As much as some people dislike ORM, unfortunately most dislike SQL even more as well as boilerplate mapping code.

That's all I'm honestly going to say on this. This thread is no place for bickering.

Nice initiative, but... by Andrea Del Bene

I really appreciate the effort of those Java iconic people to protect and revitalize JEE. But at the same time I don't understand the attempt to defend projects such as JFX or GlassFish which have clearly failed in meeting developers needs

Re: Nice initiative, but... by Reza Rahman

The Java EE Guardians are focused on securing Java EE. By extension that means GlassFish as the Java EE RI.

I do personally see the clear value in JavaFX as well. I don't think any of these things can be objectively demonstrated to have "failed developer needs". Indeed GlassFish fork Paraya continues to have strong commercial success. Similarly the JavaFX community continues to strengthen and broaden despite clear underdevelopment from Oracle. To see why you should follow this Twitter account. Indeed JavaFX remains the only option for cross-platform Java client UI development. That's why under-investment by Oracle in JavaFX is so worrisome for the Java ecosystem.

All these actions result in weakening Java in key areas in the long and short term. It needs to change for all of our sake.

Re: My observation exactly by Lukas Eder

Here's another Google Trends graph :)

Best of luck with your endeavours. Cheers, Lukas

Re: My observation exactly by Richard Richter

I originally didn't want to answer first Lukas' point - and we know about each other, thanks for discussions under my angry JPA related blogs. I believe that comparing Java centric technologies makes more sense than (probably tongue in cheek) comparing SQL to JPA. I share the sentiment for more SQL centric view. I'd love to see JPA update that allows that - actually if JPA 2.2 had joins on root entities (like EclipseLink does, very good idea for JPQL) it would help tremendously if someone wants to tune down JPA (sure, to make it less ORMish, so what) and work with FKs explicitly avoiding the problam of eager One/ManyToOne relations (JPA lazy is mere hint there).

I believe that the push from jOOQ and other similar stacks (like Querydsl over SQL directly) is healthy. I don't believe JPA is 10 years behind. I don't believe JPA is good everywhere where it's used. I believe ORM is Vietnam of Computer Science (by Ted Neward). But I don't belive it is absolute evil either.

Re: Nice initiative, but... by Daniel Campagnoli

Having working on a project recently with JavaFX I think that is a nice addition to the Java ecosystem. Especially with using some of the Kotlin JFX libraries you can write some nice UI code. The only alternatives are Swing and SWT, so its brining something new to the table.
However I fully agree with the resources spent on Netbeans and Glassfish are duplicated compared to the other options out there. As nice as NetBeans can be its doesnt add enough compared to IntelliJ and Eclipse. Likewise with GlassFish, there's already plenty of JEE container options.

Re: Nice initiative, but... by Geertjan Wielenga

NetBeans IDE is heavily used throughout the industry. It is free and open source (unlike IntelliJ) and has a worldwide community, from organizations like Boeing and NASA, through schools and universities around the world. People at organizations like Boeing and NASA don't normally fill out surveys and students at schools and universities don't either. So, you're always going to get completely inaccurate statistics from surveys -- you're always going to get the impression that NetBeans is hardly used at all, while in fact its usage is huge.

Re: Nice initiative, but... by Reza Rahman

I've no objective basis to claim high usage for NetBeans (I think about 15-20% is generous) but I do think it's adoption is nothing to sneeze at either. Indeed my general personal observation is that it is used almost as often as IntelliJ. Google Trends at least seems to agree with that view.

As stated in the article I do think it's a very nice IDE. It's definitely much lighter weight than Eclipse, faster to add important Java centric features and far less buggy. I would say IntelliJ is about on par with NetBeans but is effectively commercial.

On GlassFish, while there are indeed many Java EE runtimes, good solid OSS options are far fewer. In all these cases (including jOOQ) having options is a good thing for all of us. That is what is a fundamental and unique strength in Java.

Re: Nice initiative, but... by Will Hartung

I can't speak to JFX.

Glassfish has only failed to meet developer needs because it was abandoned by Oracle and they stopped supporting it.

Up to that point, GF was an outstanding, leading edge Java EE application server.

Their support of GF is indicative of their support for java EE, since it is the reference implementation.

But this discussion is beyond simply GF, it's more to the JCP and Java EE as a child of that process, a process that has been dominated by Oracles (lack of) leadership.

Re: Nice initiative, but... by Victor Grazi

We frequently do IDE straw polls at the NY JavaSIG meetings. IntelliJ now gets over half, Eclipse the rest, almost nothing for Netbeans. Admittedly that is a very specialized group, but it tells you where the Java teckies are doing their coding

Re: Nice initiative, but... by Geertjan Wielenga

Pretty sure Adam Bien is a "Java teckie" too, as well as James Gosling, as well as heaps and heaps of others... and they're using NetBeans... Happy to introduce your group to NetBeans anytime.

Re: Nice initiative, but... by Reza Rahman

I am a proud NetBeans user myself. I am happy to do a session and I am pretty close to NYC. If NetBeans doesn't have due adoption in the NYC SIG perhaps that should change for the betterment of the JUG?

Re: Nice initiative, but... by Geertjan Wielenga

Good plan.

This does not give me confidence by Luis Espinal

That all of these fine folks have to band themselves to save JEE makes me question the prospects of JEE in the future.

To expound, a viable platform would not need such support to ensure its future. It's viability and appeal should be enough to ensure its usage in the Java ecosystem.

Re: This does not give me confidence by Reza Rahman

I think quite the opposite is actually true. Both the caliber and number people on the site and public petition speak for themselves. The same is true of the objective Java EE usage data that our site starts with (most notably the enviable revenue that Java EE based products generate). These things wouldn't be in place if Java EE isn't as important as it is. A lesser technology and open standard would have simply faded away quietly like so many others.

In fact the only party in this case that is behaving irrationally is Oracle. Everybody else understands what's at stake here and is trying hard to solve this problem. Oracle is the only one making this situation necessary.

Re: This does not give me confidence by Will Hartung

The problem isn't that the platform isn't viable. The problem is that the leadership from Oracle isn't viable. That's what this is all about. To try and form up some better leadership. Not just for Java EE, but Java in general.

Oracle isn't interested in the "industry" outside of what in can latch on to for its own purposes.

With Sun, when the industry grew, ostensibly, Sun would grow since Sun was an infrastructure company. Expounding standards that empower enterprise software solutions implies folks are buying hardware and services to support those solutions.

Oracle is not in that business. Since Oracle is "higher up the stack", being more an applications company now, they have little interest in empowering the enterprise as a whole.

Sun was interested in expanding the broad market. As that market grows, Sun would grow, even if it wasn't the market leader.

Oracle's goal is to maintain it's legacy database business and expand it's applications base. Neither of these are congruent with Java at all, much less evolving Java standards. Beyond ensuring whatever current Java runs their existing applications well, Oracle doesn't have much of a call to do much more with it.

Re: This does not give me confidence by Reza Rahman

I really hate to make these analogies because I don't want anyone to misinterpret our group's current focus. At this moment we want to work with Oracle and the JCP. We are not interested right now in anything that looks like a fork.

That said, let's remember this is not the first time Oracle has made a strategic blunder that has subsequently been proven to be a blunder by the community. Some examples are Hudson vs. Jenkins, OpenOffice vs. LibreOffice, MySQL vs. MariaDB, GlassFish vs. Payara and so on. Oracle really has a pretty poor track record when managing technologies that are not in it's highly proprietary and commercial comfort zone...

Re: Is JEE still relevant? by Pavel Pscheidl

There is more to JEE and Spring than application servers/servlet containers. They represent common, established functionality. Made available via a server. There are JEE specifications even Spring relies on. Without JEE, there is no Spring. In fact, all the "micro" platforms only inject exactly the parts of an application server/servlet container that are required. The server functionality did not disappear magically.

Have a look at Payara Micro or WildFly Swarm. JEE has more possibilities than Spring not only in this area, people just do not know about it, because those are not heavily advertised.

And yes, there is still a place for AS. There is more to an AS than just wrapping functionality around servlets. I'd like to say that both approaches are valid. I personally use both of them now - and there are situations where one or the other way is irreplaceable.

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Email me replies to any of my messages in this thread

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Email me replies to any of my messages in this thread

29 Discuss
BT