Hacker News new | past | comments | ask | show | jobs | submit login
Re-affirming Long-Term Support for Java in Amazon Linux (amazon.com)
137 points by pritambarhate on Nov 6, 2018 | hide | past | favorite | 58 comments

No real surprise - a lot of Java underpins AWS and a lot of enterprise customers that Amazon covet are Java shops.

Keeping everyone happy for the sake of profit and market share.

Only loser is Oracle who were a loser before so no change for them.

Except that everyone else in the Java space, including Sun, already charged for support and had their own commercial JDKs.

Also OpenJDK lovers seem to keep handwaving the fact that the majority of its code is written by Oracle employees.

Elite GC and JIT compiler algorithms are not implemented for free during weekends and all nighters.

Precisely. OpenJDK is not only one of the most popular software environments around, it is also one of the biggest (in terms of features, size of standard libray etc., and certainly in terms of total continuous R&D investment) Companies like Netflix, Amazon, Apple, Google, Linked in and Twitter are either entirely or significantly built on top of it (some of them even have internal forks), and even Facebook and GitHub increasingly use it, yet contribute little back (SAP and RedHat do make some significant contributions), while Oracle spends tens of millions of dollars every year on the several hundreds of full-time OpenJDK engineers it employs. We must support that somehow by getting more companies to contribute either directly in code or financially through support subscriptions.

As of JDK 11, Oracle open sourced (or discontinued) all of the remaining commercial features in the Oracle JDK and contributed them to OpenJDK.

(I work on OpenJDK at Oracle, but speak only for myself)

"while Oracle spends tens of millions of dollars every year on the several hundreds of full-time OpenJDK engineers it employs. We must support that somehow by getting more companies to contribute either directly in code or financially through support subscriptions."

An unsustainable situation will continue until it can't --some quote I'm butchering.

I'm inclined to agree with you that we need more contributions from outside Oracle.

Right now, Oracle is doing most of the work, but isn't happy with the money they're making. That allows those companies to contribute less back, but if Oracle stops doing that work, they'll have to step up, and I think they will. They'll shift some of that effort from internal forks to maintaining the base (those internal forks work well for them precisely because they're building on such a stable base).

As an OpenJDK user, thank you for your work.

While I acknowledge your frustrations, that's one of the points of free software. If someone takes OpenJDK, uses it, and has zero problems, they don't have any changes to contribute, and the license doesn't obligate them to contribute money.

> Except that everyone else in the Java space

I think these days, people are comparing to everyone else in the programming space. Lots of comparisons on past work (and existing ecosystem), present work, and future work can be done but we shouldn't rest on history when evaluating. The days of making real money on the language/runtime alone are probably gone.

Even stuff like GCC and LLVM doesn't get made without lots of dollars/euros/yen/whatever.

Check how many of their key developers are on pay slips from multinationals, including Oracle.

Bills need to be paid and advanced compiler optimization research isn't doable as hobby.

Of course, it's just about where it comes from. There's not a GCC premium edition, or a hidden LLVM test suite, etc. I would never question the value of money in development. I think that the business model for languages and runtimes has (for a while now) clearly shifted towards no-strings-attached usage because the sponsoring/stewarding corporate interests align with the open ones. It seems the only barriers towards language and runtime use are put up these days by those clinging to business models of the past.

Thanks to Oracle, there is no premium OpenJDK or a hidden test suite as in the Sun days (Oracle contributed the remaining commercial "premium" features to OpenJDK a few weeks ago, including those that came in through the BEA acquisition). There is premium support offered by several companies (Oracle, RedHat, Azul, IBM), and a commercial JCK, which is not the OpenJDK test suite but a Java specification test suite for implementations that are not OpenJDk. The JCK can be freely used by OpenJDK.

Premium is something like discontinuing JS engine people rely on, move to Graal where certain performance features are not free for everyone. Hidden test suite is like the JCK I can't download to test my non-OpenJDK-based open source Java impl against (or Harmony or whatever). Regardless of details, the point is that it isn't more open like a MIT licensed Rust, Go, (recent) .Net, etc. There are obviously differing levels of openness by language and runtime regardless of specifics and differing levels by stewards of these languages/runtimes to make money on them.

We're not discontinuing Nashorn, simply not investing more in its maintenance any with the resources we have, which we can put to better use; if someone else wants to step up -- they're more than welcome to. Rust, Go and .Net Core (I think .Net isn't open source) are much, much smaller than OpenJDK. They require far smaller investment from the companies/organizations behind them.

Exactly. All they are saying here is that nothing is changing, we continue to provide openjdk builds, we continue to provide fixes for it, just like we do for other packages in our lts releases, just like we've always done. Business as usual. Ubuntu, Red Hat, Azul, and others are doing the exact same. The only thing that changes is that Oracle's jdk builds are no longer a safe option if you want long term support.

So, where people may have wanted the oracle jdk for mostly irrational / non functional reasons in the past, now they'll think twice.

With so much hate for Oracle, how long before they crash?

It's not like the hate for Apple, Microsoft or Google (which have enclaves of distrust) - in my experience Oracle seems universally hated.

We've bought some of their high end ZFS boxes. While I cannot comment for the hardware and experience (I temporarily left the team after we bought the HW), during the buying process I gained some valuable insight about them.

It looks like, unlike Apple, Microsoft or Google they are motivated solely by the money. Not the monopoly, not the domination, not the technology, just the money.

If they can earn the same amount of money, without doing anything, they will happily do that. Yes, they have very good technologies, but they are just built for the money.

Every corporation is in for the money, but money is generally a byproduct of good services and technologies, however it's backwards for Oracle. Services, products, and technologies are a byproduct of money. Money is not needed for products. Products are needed for money.

This is why Oracle is used also as an acronym for Larry Ellison's character and material wealth.

See also this VP's blog post https://seclists.org/isn/2015/Aug/4 which reinforces Oracle's worst stereotypes

Wow - that's shockingly tone-deaf.

He's virtually goading people into releasing discoveries to the public first (but of course they have already discovered these bugs and the fix was ready anyway - so THERE!).

Aren't we past the days when decompiling code would offer any real advantage?

In his defense, there is a real problem with people running code analysis tools and assuming the results to be correct. It can be difficult to deal with a constant barrage of incorrect "security findings".

It still doesn't excuse some of the things said in that post, though, of course.

I think he's doing his job very, very well [0]. The vulnerability in linked post is disclosed directly to the public.

[0]: https://news.ycombinator.com/item?id=18395855

"One Rich American Called Larry Ellison"

I heard that acronym expansion with a different A-word which ends with 'hole'.

Tangentially topical, fuck censorship, "asshole" is what people were calling him.

While that is true, I am under the impresssion OpenJDK project is mostly Oracle developers

That is correct. Oracle funds ~90% of OpenJDK development, and will continue to contribute most of the security patches (just not backport them).

There'll be a diaspora soon enough, as other companies beef up on their OpenJDK rosters to bolster credibility in the face of Oracle FUD.

> unlike Apple, Microsoft or Google they are motivated solely by the money

Unlike Google? What other motivations Google have? MS and Apple have solid products (hardware and Office/Windows). What Google have?

I'd risk to say, any business is motivated solely by money :) The differentiator is what and how are they selling. Oracle sucks in the 'how' department. Google sucks in 'what' department - phishing for people's data mostly, and selling it to businesses.

I think what he's implying is that they don't ever start with a product idea without a monetisation strategy or customers, like Google have done a few different times now.

That might be true but I don't exactly see what's wrong with that? After all by identifying what is going to be financially successful they are very much trying to serve the demands that people have on the market.

If anything not enough companies actually ask themselves if there is a market for their product, and instead spent a lot of manhours on building products around 'ideas', when at the end of the day someone actually needs to buy your stuff.

That last part isn't the side-effect of a business, it's what it's about, and it generally makes both consumers and businesses better off.

By experience, I divide the businesses / businessmen into two main groups. First group is motivated by serving the demands and needs of the market, and they earn money as a result. Second group wants to earn money, and they earn money by trying to serve the market's demand or by bullying people into buying their products.

First group earns way more money, produces way more higher quality goods, and people are happy to work with them, because don't feel ripped off or short changed. The second group makes you feel ripped off, produces inferior products and never supports you if something goes wrong.

Earning money is a side-effect for the first group. Yes, they earn money, but since they provide quality goods or services, they earn money without effort. For the second group since earning money is the primary target, they don't care about the products they provide unless the money stops flowing.

I didn't intend to imply that business shall be free. I tried to say that good business inevitably brings money in, and Oracle doesn't provide good business to earn money. They just want your money, and give you something in return to act as a memorial item and as an instrument to get more money from you.

@faceplanted, exactly that. Thanks for the clarification.

If hate could crash Oracle, the site of its corporate headquarters would’ve been a giant “crack of doom” since the early 90s. The enterprise world runs on a different strata of popularity.

They won't crash while they are trusted by the small cohort of decision makers in the "Enterprise" software market.

This is exactly correct. The guy I work for has friend who does sales at Oracle; as soon as he mentioned to his friend that he was starting a software company, the friend started telling him that Oracle was needed to make sure our data didn't get "corrupted" and started blaming random problems on not having Oracle; so the guy I work for started asking me if we needed to move to Oracle immediately and front the money for it. I was like, "hell no", we are a very small company, no need to pay a huge licensing fee just for a database that won't make a bit of difference to our customers... but he still sometimes floats the idea.

> telling him that Oracle was needed to make sure our data didn't get "corrupted" and started blaming random problems on not having Oracle;

It is and it has always been Oracle and Larry Ellison to dump on others tech to promote his own.[0]

Now you know why enterprise love Oracle so much.

[0] https://youtu.be/xrzMYL901AQ

Wow, that guys is really good at hustling. Truth is very few people are technically well-grounded enough to audit the differences between two database solutions. And the solution(s) being closed source make it even harder. For the same reasons there exist literally no independent comparisons to different database solutions. So the choice will be highly fashionable, eg. based on how good you will look, rather then the qualities in the materials and handcraft.

Not everyone has someone as convincing as you, it seems: https://thedailywtf.com/articles/A-Software-Problem,-A-Marke...

"It's not like the hate for Apple, Microsoft or Google"

This is not about love or hate, it's about how much value they provide and at what cost.

Yes, they spend a lot of money convincing execs that they need Oracle, but they actually do have a good product. And don't discount all the other things like support etc.. In many applications, the cost of a DB is simply not hugely material.

So I don't like them, don't care for them, Ellison is one of the bigger douche bags around. But I'd still buy their product if it made sense to.

Genuine question: is Oracle actually still superior to alternative solutions these days?

I used to work with Oracle in my telco days, and RAC in particular was a nightmare to deal with. But then as now, the standard defense used to be "oh, the product is actually great, you just need to get set up right by one of the 3 Oracle DBAs on the planet who know what they're doing".

I actually think this is what Oracle intented: Broad support for Java. They will probably be able to extract some service fees from a huge professional market. Much better than have developers fleeing in all directions and being left with a dying platform.

Maybe, but a simpler hypothesis is that what Oracle intends to do is to scare up a few billion dollars with FUD.

It’s good that Amazon (As well as IBM and Redhat which are supporting OpenJDK, and Microsoft which made a deal with Azul) are supporting Java...

However, the writing is on the wall. Oracle is desperately trying to monetize Java, and they will be successful for a little bit, this also dilutes their control of the ecosystem. That can lead to fragmentation. It also changes the equation about what kind of adoption corporations will be willing to do with java, and not have a lawyer worry about a per-vCPU licensing term.

We’ve started discussions about a long slow gradual transition from Java. I’m very very sure we are not the only one.

This is good news. Red Hat and IBM had already made similar announcements, but now that’s one company rather than two. Since we are talking about long term maintenance here, this is definitely a more the merrier situation.

This is great. Seeming that Oracle's policy changes have caused other big player to step in and support Java in a convincing manner. I am not sure if that was Oracle's intent given the company's reputation but I think it actually might be ...

>> We are collaborating and contributing in the OpenJDK community to provide our customers with a free long-term supported Java runtime.

So hopefully other distros will also be able to support OpenJDK8 for a few years to come.

All of these comments are not necessarily fair about Oracle.

They are moving faster in the last 2-3 years.

They are also creating uncertainty. That AWS felt this announcement was necessary is evidence of that. Whatever goodwill Oracle may have created recently is paying off a vast deficit. They don't get the benefit of the doubt.

Microsoft .Net team take note.

Classic .net deployments are regarded as system components, they will receive security updates as long as the supporting OS is supported(that's at least the sales pitch we received) So 4.6 on Windows 10 will be supported at least till 14th October 2025 and 4.7 on Windows 10 at least until 13th October 2026 (if you are willing to pay Microsoft for that)

> (if you are willing to pay Microsoft for that)

So ... same as with Oracle. The "if you are willing to pay" part is what people are crying about.

This amazon release specifically says that amazon will only support up to 2023, what makes you think they won't do the same?

The Amazon release promises support out to that date.

There's fairly good business reason to think that, if OpenJDK continues to be popular and relevant (as I expect it will be), that Amazon will continue to extend that date, keeping it roughly four to five years in the future each time. They're giving confidence in the long-term support of Java on their platform, not making an indefinite promise about the future.

Also to note James Gosling works at Amazon since May '17. They clearly have a vested interest in Java.

.NET Framework was released in 2014 and will be supported until 2025. I'm not sure the timeframes are comparable regarding Microsoft vs Oracle.


On the contrary, Java is a very nice language in its core. It was dubbed C+- or C++ for lazies, and the language fits to this definition very nicely.

For most part, it can do a lot of things easily, but since it was a heavy language when the computers was as powerful as a Raspberry Pi (MK-I), it became infamous because of its resource usage.

This problem was compounded by Java's "conceal programmer's errors" mentality, especially in Java Swing. Swing has a very specific way of initializing things, and if you mix them up, Swing fixes them internally, but with a performance penalty. This results in performance penalties and increased resource usage.

As a result, Java was used as programmers' MATLAB for a decade or more. This is because it's very widespread, and very hard to monetize. It became a commodity runtime like Python in early 2000s (and Sun was actually happy about it).

Huh? I feel like this kind of emotional opinion can only come from someone that has little or no experience with Java beyond reading a few random blogs about it.

Yes, obviously anyone who doesn’t like Java is a moron, and couldn’t possibly have used Java, because it’s so perfect.

Java and its ecosystem is great. It helps our teams easily solve business problems. And at the end of the day, what matters is business needs, I'm really tired of all these purists trashing a great framework.

I've always said that Java is the best for solving organizational problems moreso than programming problems. In fact, I think the feature that catapulted it to the forefront was probably javadoc. The reality is that there is very little difference in the effectiveness of different program languages unless you're deep into niche problem domains. Java is a perfectly good general-purpose language with a powerful toolchain.

A) Java isn’t a framework.

B) Not liking Java doesn’t make someone a purist.

C) You can solve business problems in other languages besides Java. It’s a common misconception amongst Java programmers, I know.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact