Some Early Tests Of The Eclipse OpenJ9 Java Virtual Machine

[bug77]

Hmm, I wonder how IBM justified developing J9 for so long, given how much worse than Hotspot its throughput is. Maybe it's a GC or jitter thing?

Maybe J9 has better performance on POWER.

Support. If an issue comes up, it can be fixed internally, bypassing open source upstreaming drama. At least that's how they justify implementing everything in-house.

What about memory consumption? Machines with light-load Java applications are mostly limited by amount of available memory, not by CPU computing power.

It comes from IBM, you don't need benchmarks to know it's a memory hog.

[cyrix]

Hmm, I wonder how IBM justified developing J9 for so long, given how much worse than Hotspot its throughput is. Maybe it's a GC or jitter thing?

Maybe J9 has better performance on POWER.

Or maybe it's just a matter of configuration and tweaking. Something you don't typically learn in two days after a project like this gets released.

[microcode]

Or maybe it's just a matter of configuration and tweaking. Something you don't typically learn in two days after a project like this gets released.

No such tweaking is being done for the OpenJDK runs, so I fail to see your point. Each VM has a fair shake, OpenJDK/Hotspot seems to require less fiddling to get around twice as much throughput.

[microcode]

Support. If an issue comes up, it can be fixed internally, bypassing open source upstreaming drama.

You could just maintain a downstream...

[schmidtbag]

Java is anything but dead. It may have waned in popularity because of languages like Scala and Clojure which make FP and its "resurrection" easier due to cheaper RAM and faster CPUs.... but it's far from dead. Not to be confused with Java, the JVM on the other hand is here to stay for about as long as Oracle and its customers are around as well as the billions of devices that run it...Java 8 also added more tools to make FP more tenable too... and 9 is around the corner too.

Java to me is like a 30-year-old person: their body is overall dying faster than it is rebuilding, but the rate is so slow that the person is still able-bodied enough to do mostly whatever he/she wants without immediately regretting it, or without anyone being surprised.

The interesting thing to me is programming languages rarely die out because of obsolescence (after all, look how long C and Fortran have lived). What makes Java unique also makes it still relevant today, but, it's dying because Oracle makes for a bad parent.

[bug77]

Java to me is like a 30-year-old person: their body is overall dying faster than it is rebuilding, but the rate is so slow that the person is still able-bodied enough to do mostly whatever he/she wants without immediately regretting it, or without anyone being surprised.

The interesting thing to me is programming languages rarely die out because of obsolescence (after all, look how long C and Fortran have lived). What makes Java unique also makes it still relevant today, but, it's dying because Oracle makes for a bad parent.

Whether Java (or any other language) is dying is a sterile discussion. They're tools that come and go. And when they go, it's because better tools are replacing them. Unless you're fresh out of (or still in) college and trying to figure out where to start your carrier, what's the language of the day makes little difference.

[schmidtbag]

Whether Java (or any other language) is dying is a sterile discussion. They're tools that come and go. And when they go, it's because better tools are replacing them. Unless you're fresh out of (or still in) college and trying to figure out where to start your carrier, what's the language of the day makes little difference.

I see your point, but I beg to differ, for 3 reasons:
1. Java doesn't need to die - it has potential (as this article has proven) but Oracle's negligence is getting in the way.
2. Continuing to use something dead, dying, or obsoleted only prolongs the difficulties of transitioning away from it. People hate change, and if it isn't hated it is often very expensive (of time, resources, or money). Java is widely implemented, and if support for it dwindles (or is dropped entirely) then that can trap people in a bad situation. This is worth noting.
3. A college graduate should never devote too much time working on something where its obsolescence is on the horizon. Though we don't know when Java will lose relevance, the fact of the matter is it is losing relevance. But again, it doesn't have to; something could be done about it.

All that being said, this discussion is important, because it helps people realize how they can either save Java from dying, or, make an informed decision on which language they should be devoting their time toward. Of course, you don't have to be fluent in just 1 programming language (on any given month, I might be using up to 3 different languages, though throughout the year I might use 7) but I don't see a purpose in learning a language that is or definitively will be obsoleted by another.

[DanglingPointer]

Java to me is like a 30-year-old person: their body is overall dying faster than it is rebuilding, but the rate is so slow that the person is still able-bodied enough to do mostly whatever he/she wants without immediately regretting it, or without anyone being surprised.

The interesting thing to me is programming languages rarely die out because of obsolescence (after all, look how long C and Fortran have lived). What makes Java unique also makes it still relevant today, but, it's dying because Oracle makes for a bad parent.

Mate, I don't know how you can claim such biased statements with no corroborating evidence whatsoever. Google is your friend, there are a gazilion proper, scientific-process-derived market research reports out there! Read them with an open mind!

Seriously though, if you're in the IT industry, in SDLC, or DevOps, or whatever, and been in there for years, you would have to be living at the seabed of the Mariana trench to never notice how ubiquitous Java is and believe that Java is going the way of the Dodo! It is absolutely everywhere! In fact there are major companies out there that have transitioned from small-medium enterprise to large enterprise, that start to realise that they get more bang for the buck on infrastructure whether in the cloud, in-premise or hybrid with Java and have made the decision to migrate to Java! Dropbox and Spotify to name a couple of examples of thousands in many countries!

King of the hill for the next 50 to 100 years in my opinion! You can tell me "I told you so" if I'm wrong, but I doubt I will be unless Python3.x or the future Python4+ starts fixing its poor performance and how you end up absolutely lost in the code once you go over 50,000 lines of code or hundreds of files due to dynamic syntax and not knowing what a bloody type is easily and having to reverse engineer through 20 files to find what type a bloody object is meant to be because some other developer/s didn't document properly. That scenario just doesn't exist in Java.

You want something rapidly developed/prototyped as a start-up in a garage or from a small donation from venture capitalists, sure, do it in Python or Ruby or similar, but when you've made it after a few years, and you start playing with the big boys with public shareholders demanding bang for the buck and efficiency of ops everywhere, you start to ditch those and settle for the middle ground which can be coded once and run anywhere just like Python, Ruby and all the others,... more often than not, if not 90% of the time, you end up with Java!

[schmidtbag]

Mate, I don't know how you can claim such biased statements with no corroborating evidence whatsoever. Google is your friend, there are a gazilion proper, scientific-process-derived market research reports out there! Read them with an open mind!

Ironically, Google is one of the very reasons to back up my claims. Oracle basically forced Google to transition away from the usage of Java with Android. You think that's going to help the longevity of Java? Meanwhile, Chrome dropped Java support somewhat recently, and I don't think Edge has ever supported it (or ever will). Not sure about Safari or Opera.

Seriously though, if you're in the IT industry, in SDLC, or DevOps, or whatever, and been in there for years, you would have to be living at the seabed of the Mariana trench to never notice how ubiquitous Java is and believe that Java is going the way of the Dodo! It is absolutely everywhere! In fact there are major companies out there that have transitioned from small-medium enterprise to large enterprise, that start to realise that they get more bang for the buck on infrastructure whether in the cloud, in-premise or hybrid with Java and have made the decision to migrate to Java! Dropbox and Spotify to name a couple of examples of thousands in many countries!

I work in IT and everywhere I look, I see people whining about how Java updates break things and programs dropping support for it in favor of other languages. But seeing as you hadn't read my last post, you'd realize that I pointed out myself how popular Java still is. I mentioned myself Java has potential. In other words, I'm aware it has advantages that other languages don't have. I'm aware Java isn't going anywhere for a long time. But somehow you missed the point that I said Java is like a person in his/her 30s - anyone would agree that such a person isn't going to die any time soon.

Also, I don't know about Spotify, but Dropbox is mostly dependent upon Python.

King of the hill for the next 50 to 100 years in my opinion! You can tell me "I told you so" if I'm wrong, but I doubt I will be.

Java is not and never will be "king of the hill", even if you ignore security issues and breakages due to updates. However, I do think it will likely remain within the top 10 most used languages in the next several decades.

You want something rapidly developed/prototyped as a start-up in a garage or from a small donation from venture capitalists, sure, do it in Python or Ruby or similar, but when you've made it after a few years, and you start playing with the big boys with public shareholders demanding bang for the buck and efficiency of ops everywhere, you start to ditch those and settle for the middle ground which can be coded once and run anywhere just like Python, Ruby and all the others,... more often than not, if not 90% of the time, you end up with Java!

Not really. Interpretive languages like Python and Ruby are better than just prototyping. They're commonly used in the server market because they allow you quickly modify code while the program is still running, and they're light on resources. Graphical software sold to consumers (whether that's office suites, production software, games, etc) are written in languages like C/C++/C#, Rust, and other compiled languages because they don't depend on a runtime machine or an interpreter and they're more efficient.

Java is really only used in some servers (where the runtime machine is advantageous), in-house software for businesses, client-side web-based applications (which is being obsoleted by HTML5 and WebSocket), schools (Java is a relatively easy language to learn OOP), and hobbyists who care about performance (where languages like Python and Ruby won't keep up). That's not as diverse as it sounds.

[duby229]

Also worth a mention here is that Python is highly useful to automate stuff. It can actually be a half decent scripting language. I've never used it to write a "program", I don't have the skill for that, but I use it almost constantly to write scripts.