Apache loses Java showdown vote to Oracle

http://www.theregister.co.uk/2010/12/07/apache_google_vote_no_oracle_roadmap/

by nradov • 16 years ago

108 125 16 years ago

theregister.co.uk

kls8• 16 years ago

As a long time Java developer (I downloaded the first beta that was released). I would have to say that this pretty much did it for Java. Sure it was showing it's age but until now the pain of leaving it behind was too great for many. I think now there are a lot of people wondering what the future pains will be brought if one decides to stay in the Java world. Oracle has thrown the gauntlet down and pretty much said, Java is for paying customers. So my question now is who is the likely successor? I have been playing the guessing game for a while now as to who I think the most likely champion is and I keep drawing blanks. I think the JVM languages are too close to Oracles reach for comfort now, so I think Clojure, Scala, JRuby and the rest are out. Python is nice and has a pretty decent eco-system but there is a die hard no interpreted languages camp (I know, I know) but it exists. I don't know enough about Go to form an opinion of it's worthiness. I don't think we are going back to C or C++. Objective-C is perceived to be too Apple, .NET too Microsoft. Perl had it's moment in the sun. Ruby too controlled by a select few. My point is, I would love to know what other people think on the subject. I think the writing is on the wall, I just can't seem to read what name the writing spells.

windsurfer• 16 years ago

Language shouldn't matter, it's the tools around it. Now instead of having one massive amount of tools for a single language, the way of the future (that's already here, actually) is fragmentation and cooperation. Ever heard of LAMP? It's four very different tools and multiple languages that make up only a part of a larger software stack that is only for a particular purpose. This works surprisingly well, and it's easy to swap out any piece you want with another. And this wasn't by design of some massive company like Sun or Oracle, but merely cooperation and good old software interfaces.

RodgerTheGreat2• 16 years ago

I still hold out some hope for the Java language, but I share many of your feelings. My main issue in searching for another language is that I don't see many imperative languages with the same kind of methodical approach to object-oriented design and comparable language support for concurrency. The C++ build environment feels like going back to the stone age after using a compiler like javac, and while D seems interesting it still appears to be a rather severe step backwards. Many of Python's design decisions (significant whitespace, idiomatic exceptions-as-flow-control, duck-typing) really rub me the wrong way. I've been tinkering with Lua and FORTH to keep myself sharp, but I can't imagine treating either as a replacement bread-and-butter language like Java.

Luyt• 16 years ago

significant whitespace

Many years ago, this delayed me looking seriously into Python. But when its powerful data structures finally convinced me to start using it, I quickly became accustomed to significant whitespace: I always indented my code correctly anyway, whatever language I'd use. So it was just a case of dropping the { and }. After a few hours I was accustomed to it.

The funny thing is, I still program a lot in C++ and PHP (and Java sometimes). When I write these languages, I fall back on using { and } like I did before, without even noticing it.

Getting proficient in a new language doesn't make you forget your old languages.

gaustin• 16 years ago

As much as I dislike .NET in general (there are some excellent parts) and doing large enterprisey development, I think C# is an excellent language that is only getting better. C# started out as a clone of Java, but it seems to be what Java could have been if it had evolved more rapidly.

If you're concerned about concurrency support, there are new features in .NET 4.0 that I find pretty impressive and nice to use.

That said, I don't know if the story for developing C# applications on Mono is as solid as Visual Studio.

AndrejM• 16 years ago

Interesting.. In what way does D seem like a severe step backwards to you? Is that assumption based on the library support or the language itself?

flatline• 16 years ago

I think the answer is, at least in large part, right there in your post. There is pretty broad knowledge in the industry of how to build cross-platform C++ applications, practices that weren't mature when Java took off, and this includes targeting IOS. Python and Ruby are great for automation and web development. Is there a need for One Language to Rule Them All? The JVM is the best part of Java IMO and I'm not really clear on where that stands in this current battle, but I know there are JVM implementations out there other than HotSpot.

rbanffy16• 16 years ago

You can't really compare Java with C++ directly. All they have in common is the syntax style. Java was designed to be C++ programmer friendly, but it shares a lot more with Smalltalk than with C++.

badmash695• 16 years ago

I fail to comprehend your objection. The licensing terms are what is at dispute here and not the technical merits of the language or the JVM platform. These were the same terms ( more or less ) when you downloaded the first JDK. Apache and Google are disputing the licensing terms . Oracle has nowhere said Java is only for paying customers. Just chillax !!!

kls8• 16 years ago

"Oracle has nowhere said Java is only for paying customers"

They killed OpenSSO, they killed Open Office, the killed Open Solaris, They where contractually bound to not kill MySQL so they just raised the price through the roof to send the message. Pretty much they killed anything that you did not have to pay for. The worst part is they did not just withdraw paid positions from working on it. They alienated the volunteer members of the community until they resigned in protest.

Sun agreed to license the TCK to Apache under non-discriminatory terms an Agreement that Oracle is not, has not and will not living up to, because they don't want JDKs to exist that did not pay Oracle or more importantly they want them to licence it under the restrictive provising that they require e.g no use on servers and no use on Mobile. How many more items do they have to destroy until people start seeing the serial and systematic behavior that is going on.

do you not find it funny that Apple gave the JVM back to Oracle and Oracle is doing everything in it's power to keep non-sanctioned VM's off of Mobile devices. That is what the Apache dispute is really about, Oracle does not want a non-oracle JVM in the mobile world and I can bet you dime to dollar that it has to do with a back room deal between Oracle and Apple. Why do they want to keep the non-sanctioned JVM's off of Mobile devices, well because a closed ecosystem is more lucrative for Apple and Oracle than one in which Apache and Google are giving stuff away for free.

There is a fundamental difference in world views between Apple / Oracle and Google / Apache. Apple and Oracle are vertical stack providers they want to provide the whole stack soup to nuts and don't want anyone playing in their ecosystem, it just so happens that Apple wants the consumer market and Oracle wants the Enterprise market so they make good bed fellows. They are about the only two tech companies left that are not at war with each other.

Google and Apache represent a threat to their world vision and they have declared war to stop them. Oracle is all about every piece of their vertical being a piece that is monetized you can rest assured that Java is no exception.

Like it or not Java and the JVM are pawns in a big game of who is going to own the mobile market. It is the future market and winner takes all, if Oracle backs Apple in wining the market, then Apple can ensure that Oracle is position to be the Enterprise infrastructure for the new Mobile market. Like it or not, Java is a sacrificial pawn in a bigger game now.

The developer in me does not care, quite honestly the return to verticals means the return to easy money for writing software for verticals but the consumer in me shudders at the though of the verticals winning. You think Apples platform is restrictive, now wait until they own 90% of the market. They will make the early 90's Microsoft look like a benevolent democracy and well Oracle will just become a technical tax to anyone wanting to do business.

callahad2• 16 years ago

I think .Net is poised to capture much of Java's current market. The C# developers I know really do seem to enjoy the language and associated tooling, and it's very well established in the enterprise. I can't think of anything else that has significant mindshare and is directly addressing that space.

At the same time, alternative languages (Python, Ruby, etc.) are becoming more and more capable, and are slowly eroding the class of problems that need "enterprise-grade" solutions. Furthermore, virtualization, as well as tools like Chef, Puppet, Capistrano, Virtualenv, and RVM are making great strides at simplifying the deployment and management of "native" solutions.

Taken together, I think you'll see huge gains by smaller languages in the areas traditionally served by Java and .Net.

But if I had to pick a dark horse or two, I'd look to LLVM and JavaScript. There's huge amount of momentum and money behind LLVM, and I could easily see it becoming the next major target platform for experimental languages. Similarly, the recent JavaScript arms race has left us with at least four extremely high-performance JavaScript runtimes, and it's quickly becoming the lingua franca of the web. There have been many attempts to harness that talent and apply it to more general application development: Apple's Dashboard, Palm's WebOS, Node.JS, Appcelerator Titanium. Others initiatives are trying to pull general application development into the web dev world: Sproutcore, 280 North's Cappuccino, etc. Eventually, someone is going to strike the right balance, and JavaScript's importance and influence will skyrocket.

Other things that may spawn something really cool: Parrot, Go.

kls8• 16 years ago

While I love JavaScript, I shudder at the though of using it as a systems integration language. For the verry reason that it shines as a UI development language (flexibility), it makes me cringe to think of using it for system integration work where items like guaranteed delivery and transactions are far more important that flexibility. As well, I just don't see .NET doing it, it has very little adoption amount the unix camps and the next language will most assuredly will come with the blessing of that camp just as Java did. I mean it was the unix guys that adopted it because they could offer there server offerings on their platform of choice while not ignoring the Microsoft juggernaut.

As for LLVM, I have had the same thought and I think you are on to something, I keep thinking whatever it is, it will crop up on LLVM.

eitland• 16 years ago

>I think .Net is poised to capture much of Java's current market.

Anyone leaving Java because of Oracle is not going to .Net.

> The C# developers I know really do seem to enjoy the language and associated tooling, and it's very well established in the enterprise.

Doesn't matter. If you are leaving Java because of Oracle, then you don't care about the tooling. Moving to .Net will just move you to another vendor those guys have spendt the last years moving away from.

InclinedPlane• 16 years ago

More than that, for typical distributed "enterprise-y" type stuff (a crap-ton of services that call each other) MS's stuff is generally superior. Creating a typical WSDL based service on Java using Spring requires a lot more work than with .Net, and all of the nice little features (like tools for manually calling a web-service API) are built-in.

Unless you are a big enough company to be able to roll your own infrastructure tools to make use of linux's advantages or have very significant scale requirements the cost difference is unlikely to be huge (much of the cost will be operations labor anyway) and the service and support backing of MS is a significant boon in the enterprise world.

LLVM and Javascript are contenders but they aren't platforms yet, although I think they will see a ton of use in web apps and web services, especially outside the enterprise.

megaman821• 16 years ago

I pretty much agree.

Java is under a two-pronged attack. LLVM is quickly becoming the preferred target platform. The harder Oracle is to work with the harder it will be for new languages to target the JVM.

Also Java's niche is getting squeezed from both ends. Python, Ruby, and Javascript can do more of what Java does with less hassle. Languages like Go and Rust are bridging the divide from C/C++ to Java also. Both languages have a nice mix of good performance, predictable performance, high-level semantics, and concurrency.

badmash695• 16 years ago

please cite the link for this statement "Java is for paying customers." 'Cause without this link , it seems you are merely spreading FUD.

chc2• 16 years ago

I'm not sure what you're looking for. You want a link to an attitude or pattern of behavior?

Nelson692• 16 years ago

Well that's pretty simple. There isn't really a suitable successor that does what Java really does. It was low-enough level to have decent performance and high enough level to be used to quickly do some complex stuff with some runtime protection.

I'm curious, the JCP didn't exist until quite a while after Java got going. There are tons of documents in various states of completion and with various levels of support. To me, JCP has sort of given java a terrible look to the non-java coder. It looks fat, bloated and incredibly Byzantine. I've actually heard developers have conversations about JCP<number>, that just seems like a bad thing to me. Meanwhile SpringSource has done some great and wonderful things. Even newer JCP specs, like JPA (if that's 'newer') were sort of done outside of it all and then backported in to the JCP process.

I don't know, C has been pretty much "dead" for decades, there is very measured and limited change in that language. Putting the Java language in a similar place, adding some dynamic support to the JVM and maybe focusing on cleaning up the class libraries and interop with higher level languages might be the best thing for it.

I don't know, it all looks bad, that's for sure but i can't say the JCP was the end all be all either.

loup-vaillant3• 16 years ago

> There isn't really a suitable successor that does what Java really does. It was low-enough level to have decent performance and high enough level to be used to quickly do some complex stuff with some runtime protection.

There was such a successor before Java even existed: ML. It is safer, and current free implementations are at least as fast (and they have a C FFI). The catch is, programmers need to recognize that "OO" is little more than a brand, then see that despite it's "FP" branding, ML implements the underlying principles just as well (if not better).

wmf4• 16 years ago

Oracle has pretty much said, Java is for paying customers.

Actually, they said Java on phones is for paying customers and Java on servers will always be free. http://www.theregister.co.uk/2010/11/06/oracle_dueling_jvms/

bad_user2• 16 years ago

> Java on servers will always be free

Great, so I'm free to take a server implementation and port it to mobile phones, right?

Free as in what?

rbanffy16• 16 years ago

> JVM languages are too close to Oracles reach for comfort

Not at all. All of them run on the OpenJDK and the GPL makes it Ellison-proof. He can scream an threaten all he wants.

OpenJDK is not as cool as all the Sun-invented niceties in their own VM implementations, but it's good.

kaffiene5• 16 years ago

I don't believe that the GPL gives you patent protection. (GPL 3 does, but this is the GPL 2 I think)

dkarl5• 16 years ago

Do you think it's implausible that Apache could fork Java and Oracle could be eventually left with nothing but a trademark and some loyal corporate customers? With so many prominent Java folks leaving Oracle, I'm not sure what kind of leadership or moral authority they could muster to convince people that they're the most more credible center for the Java community. The community would suffer from a fork, no question, and it would be severely weakened, but I think Java users would prefer to go through such a tribulation rather than write off Java completely.

Personally, I think Apache should fork Java and kick it off with a renaming competition, just to get us past the most traumatic part.

kqr2• 16 years ago

I thought the main problem with forking without getting the Java(TM) seal of approval is you may be liable for patent infringement.

rmc• 16 years ago

Do you think it's implausible that Apache could fork Java

I thought Apache already did that, that's how the Apache Harmony project started. This was all done before Java was open source and one needed a 'free as in speech' java.

kaffiene5• 16 years ago

Oracle has been quite clear that Java will continue as it always has for normal, non-paying customers. Only the JRockit stuff which they were already charging for will be pay-for in the future.

But whatever, FUD away.

kls8• 16 years ago

Oracle has been quite clear that Java will continue as it always has for normal, non-paying customers

Not on mobile that have not, in fact they have stated quite the opposite. With mobile projected to surpass desktops in 18 months that is cause for concern. Make no mistake about it, the wranglings with Apache and the patent suit with Google are intricately related. This is all about ownership of mobile.

pietrofmaggi• 16 years ago

When Java was released in '95 it replaces Smalltalk in the Enterprises (mainly banks). And during this decade Smalltalk was down to little or no usage till useful stuffs (Seaside and Pharo comes to mind) bring it back to life.

Can Smalltalk be the new (old) stuff?

protomyth• 16 years ago

In order to truly replace Java, any language really, really needs to think about the ecosystem around the language. Smalltalk vendors really do not have a good deployment strategy, certainly nothing as comprehensive as Java.

The other problem is that Smalltalk is pretty alien to a Java programmer (image vs. text, block based control flow vs. traditional control flow). Garnett and Forester's recommendations might be key.

Although, I still would argue the premise and say most of these institutions know Oracle and understand dealing with them. I don't see people giving up their Oracle databases and having bundled, recommended Java might make Java more of a lock.

bergie2• 16 years ago

There is GNUstep for Apple-free Objective-C (http://gnustep.org/), and Vala is another interesting modern language (http://live.gnome.org/Vala)

SkyMarshal• 16 years ago

I can't help but wonder why Google is apparently willing to spend $6B on Groupon, but wasn't willing to spend roughly the same to secure a significant portion of the intellectual property assets that Google is heavily dependent upon.

Is Groupon really as or more valuable than Sun to Google?

felipe• 16 years ago

Good observation. Yes, it makes sense, because Google is a media company, and not a technology company like many people think.

varikin• 16 years ago

I would say Google is an advertising company instead of media company, but the sentiment is the same.

axiak• 16 years ago

It seems clear that one large part of the equation for Oracle was being able to sue Google to recoup the cost. This obviously doesn't work well for Google. In addition, Google doesn't have a stellar history of integrating acquisitions (it has had a tough time with youtube), and I can't imagine they thought it prudent to try to integrate Sun.

Still, it is a nice dream, isn't it?

joshu• 16 years ago

youtube was explicitly allowed to remain separate.

dmak• 16 years ago

What does all this mean? How did this even happen? Can someone give me the summary? I haven't really kept up with this from the beginning.

wmf4• 16 years ago

This has been going on for something like eight years so it's not easy to summarize. But you might start here: http://www.jroller.com/scolebourne/entry/sun_apache_ip_in_pi...

elblanco4• 16 years ago

Since it seems that the concept of compiling code to an intermediate bytecode run in a JIT-VM seems to have proven itself pretty well (and the concept of compiling multiple languages to said VM a la Scala), perhaps it's time for the community to look at building their own from the ground up VM with associated language support?

Perhaps https://www.parrot.org/ ?

harryh• 16 years ago

http://llvm.org/

dkarl5• 16 years ago

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies. Despite its name, LLVM has little to do with traditional virtual machines, though it does provide helpful libraries that can be used to build them.

If LLVM is not a VM, do you have a particular LLVM-based VM in mind?

orangecat2• 16 years ago

LLVM as noted. But the problem is that if Oracle's lawsuit against Google is successful, then any VM that's remotely similar to Java's is threatened.

chadj• 16 years ago

So what does the HN community think of Vala (http://live.gnome.org/Vala)? C# like syntax. Compiles down to C. Reference counting for memory management. Appears to work with clang as well.

More example syntax here: http://live.gnome.org/Vala/Tutorial

zaph0d• 16 years ago

Thanks, but no thanks.

rbanffy16• 16 years ago

Looks interesting for low-level stuff. I assume memory management will become more sophisticated with time.

kaffiene5• 16 years ago

You can't do sophisticated GC in C. The memory model of the JVM is important - it knows what is a pointer and what is data, which is important for efficient GC.

bergie2• 16 years ago

We use it on the server side with http://midgard2.org/

Check out https://github.com/midgardproject/midgard-core/tree/master/s...

binaryfinery2• 16 years ago

The power of Java is not the VM, or even the system runtime, but the open source libraries available for it. It is time to create a "new language" whose compiler will accept Java, and whose VM implementation is unencumbered by legitimate patents. Hm. If Dalvik isn't it, then something similar will be created. I, for one, would like to help. Anyone else up for it?

prodigal_erik• 16 years ago

Java is the prototypical Blub. It's saddled with a type system that wasn't even state of the art in 1991, motivating gigantic enterprisey libraries of RequestSpecificProcessorFactoryFactory crud. The VM is the reason we grudgingly tolerate it where I work.

dkarl5• 16 years ago

It may be "prototypical Blub," but it has attracted a lot of bright and effective programmers, including several top-ten finishers in the Google AI Challenge and many, many people writing oodles of useful, practical code. I think it's fair to say from history that if you want to get things done in a cool, nicely designed language, it helps if your cool, nicely designed language runs on a platform that is dominated by a popular Blub language.

Proof: If there's no Blub on your platform, you'll exclude the vast majority of people who might write useful code you can use. If there is a Blub on your platform, its popularity will ensure that the vast majority of useful code on your platform will be written in Blub, and first-class language status on your platform will be defined by ability to interoperate with Blub. Therefore, the best case scenario is for your favorite language to be a minor player on a Blub platform.

kaffiene5• 16 years ago

Not true. The VM is very important. The consistent & unified memory model is why Java can do synchronising, threading, garbage collection efficiently, cross platform and consistently.

Also, HotSpot is really important. Java is an acceptable systems language at 80-110% the efficiency of C. It's not at 50% or less.

binaryfinery2• 16 years ago

As can many other languages. However, when I use .NET, I find myself wishing for maven and all the libraries immediately available. Without the VM, there would be no point in having the source. However, even with a great VM, other languages fall short of Java, IMHO. YMMV.

There are also areas where the Java VM totally sucks, frankly. How is it that we only have decent IO after more than 10 years? Sun and now Oracle hold the Java community hostage with the VM and runtime. Time to move on.

anttipoi• 16 years ago

Results are out. http://jcp.org/en/jsr/results?id=5111

Quite a few of those who voted yes added comments about the yes being on technical merit, not licensing.

JoachimSchipper5• 16 years ago

Interesting to see that e.g. Eclipse and Red Hat voted yes.

cturner• 16 years ago

Could Harmony just rebrand and go its own way? i.e. people would be "Harmony developers" instead of "Java developers" but it would be the same thing. Or are there patent or licence obstacles to this? A platform run by Google and Apache would have a core of great engineering and the pick of the best libraries.

I appreciate that ASL is in reality an impossible boat to turn because it relies on the voluntary contribution of dozens of key developers, each running private projects with investments in software based on the existing VM tradition.

BonoboBoner• 16 years ago

While the situation is far from perfect, I am at least happy that actual progress is made, as we have a release date for JDK7 and respective JSRs at last.

Regarding the "premium JDK", I dont mind Oracle charging companies like investment banks extra bucks for additional JRockit-features (Monitoring, etc) that the open source community never actually had free access to anyway. There are so many Oracle products that heavily depend on Java, that it does make no sense for them to hurt Java. If they destroy Java, they destroy part of their income, which I just dont see happening. Although the geekish OSS-nature of Sun's stewardship is gone, things could be by far worse for the Java ecosystem.

Speaking from our situation, I dont see us switching to .NET any time soon. Although it is a great platform especially with the recent developments on C#, for us it is just not cost-effective to buy new Windows Server, SQL Server, etc licenses for every new non-trivial project. It is just far easier and cheaper to build a new Debian-machine on our VMWare cloud and apt-get whatever we need.

__david__• 16 years ago

Can't you "apt-get install mono"?

badmash695• 16 years ago

The features in JDK 7 and 8 are what we the Java developer community have been asking for. Heck, even RedHat held its nose and voted yes for the technical merits. Java is only getting beter and the JVM -- thanks to Clojure , Scala and JRuby -- has a vibrant community that is extending it beyond the limitations of Java.

But please don't let these facts get in the way of Java fear mongering. Good to see the Java haters' FUD factory is back in action. The old " java is too slow" campaign was losing its appeal. Good to see a fresh new FUD.

buster• 16 years ago

Would it be possible to "fork" Java? Alternatives could probably be Parrot or LLVM?

peapicker• 16 years ago

The article is a little misleading; the refusal by Oracle is to not let them run the certification tools _for_free_; every other JDK developer has paid for certification. Apache won't pay; so they don't get certified.

Apache loses Java showdown vote to Oracle

Related Stories