Considering the amount of tooling and or services on AWS that relies on Java this is a great fit. I wonder what kind of services he could help AWS create (a cloud service that would support J2EE and also be lower cost / integrated)?

Considering his recent advisory work related to Scala, there's a good chance he'll lead an effort to build new services in Scala. A guy like James is clearly passionate about language design, so I be he'll help shape the future of Scala as well, using input from his AWS experience.

FWIW (and we're hiring) ... several AWS services are already written in Scala.

If I were to join Amazon, what orgs/teams should I lean towards if I prefer to do FP? I have more experience in Clojure but Scala would be great, too.

Amazon is super Java oriented so the overwhelming majority of backend work will be in Java. I would just search Amazon.jobs for Clojure or Scala, since even within specific teams you might see a combination of OOP and FP.

that is a bold hypotenuse. Also, I guess for AWS is more a marketing tool to get more developers. The same for heroku(matz), dropbox(Guido)

Hypothesis.

:)

Any chance that Amazon hired him with the mandate to develop a "new Java", in the same vein as Ken Thompson developing Go at Google?

That's exactly what I think. Nobody trusts Oracle and want to move away from Java no matter what it takes. See Kotlin.

I hope he creates a Kotlin/Swift evolution, I am already in love with both of them.

Those of us on Java community are pleased with the work done on JavaFX, Graal, SubstractVM, JRuby, Nashorn, research on eventually add value types and improvements to generics on the JVM and support for AOT compilation.

So far, Oracle has managed to push for more Java improvements than Sun was doing at the end of its history.

And those devs on OpenJDK, guess who pays the majority of the salaries?

Kotlin wouldn't ever been born without Java in first place, nor JetBrains would have a business.

JetBrains does a lot in the .NET world too. They'd still have a business.

Developers trusted Sun much more than Oracle with Java. Java wouldn't have been born without developers and an open ecosystem. Now, Oracle might not have a business if they destroy trust with Java.

First of all, JetBrains IDEs run all on top of Java, except for the Visual Studio plugins they also develop.

Second, while it is great that .NET is now open source, the libraries that would allow for something like InteliJ to be build across multiple OSes aren't quite there yet.

Finally I was actually referring to the fact, which I hate to see done by some in new languages that target JVM communities, is spiting on the plate that gives them food. Without the JVM those languages wouldn't even exist in first place.

Yes Sun had a more engineering friendly culture than Oracle has, but they also did tons of missteps. Specially in what concerns making Java appealing for the desktop, something that for incredibly as it may sound, has been better handled by Oracle.

IBM is a pretty good check on any theoretical Oracle crazyness in the java arena via the JCP and is very invested in java.. I wouldn't be too worried here.

They seem pretty keen in turning J9 into a multi-language VM and pushing Swift on GNU/Linux and their other OSes.

> "new Java", in the same vein as Ken Thompson developing Go at Google?

Go is certainly not the "new Java" at Google. In fact, Go pretty much failed at any wide adoption at Google. Go is mainly successful outside Google.

> Go pretty much failed at any wide adoption at Google

Interesting, first time I am hearing this claim. Do you have any good sources for this? Most of the Google engineers I know of speak highly of Go.

I didn't mean to imply Go was meant to be a new Java.

I meant that Thompson had the chance to develop a new language under Google's auspices, and that Gosling might do the same at Amazon. That's all.

I don't think that's how Go was conceived or Ken was hired.

Ken Thompson was brought on because he was Ken Thompson, and if you had the means, then not hiring him would be a strange thing to do. The amusing thing about that is that he never jumped through the new hire hoops (a language exam, or something) and therefore could not commit his own code.

https://www.theregister.co.uk/2010/04/21/ken_thompson_take_o... (quote here but it's worth checking out the book)

>> he never jumped through the new hire hoops

To paraphrase an old Steve Yegge joke [1]

"What? How could they? He doesn't even know JavaScript!"

[1] https://sites.google.com/site/steveyegge2/math-every-day

I think Rob Pike half-heartedly mentions in a couple of the twilight videos introducing Go that Ken and he got together to conceptualize go after he walked into the office one day and got tired of a google server binary taking ~40 minutes to compile. Not sure if it this is the real reason, but it's a great story.

Me neither besides I think there is not a room for "yet another Java/C#/Replacement". You need to offer much more now days to convince people to switch.

Maybe James Gosling would have a few ideas on what that "much more" should be?

He has worked on Java for nearly a quarter of a century. He's not young, but might still want to do something else.

He is (was?) a strategic advisor for Eucalyptus, a sort of AWS compatible on-premise cloud, also is/was a director at Jelastic. So I imagine he's already got some ideas for AWS.

Great. He is also a JavaEE guardian.

https://javaee-guardians.io/who-we-are/

This bodes well for AWS. I think this is good news.

Great news for AWS and Java since AWS is a major Java shop. What will he work on at AWS?

Probably some deep instrumentation of Java in the cloud

Dude lasted less than a year at Google, and the only thing he produced was a huge google doc containing a diatribe on how wrong Google's approach to software development was. Googlers had a good laugh and then carried on as before.

Any chance you could post that (with proprietary details anonymized?)

I'd love to read this.

I think he is also pissed off with Google due to Android thing.

is there a link to that document?

... but where is the creator of Kotlin going to join?

I'll speculate that he will design a new cloud-first programming language. Almost all letters in GAFAME has their own programming languages:

  Google has Go (and Dart)
  Apple has Swift (and ObjC)
  Facebook - Hack and ReasonML
  AWS - ?
  Microsoft has DotNet-based PLs, Typescript
  Ethereum - Solidity, Serpent, Viper, etc.

Including Ethereum in this group is laughable. Apples (literally) and Oranges

> Facebook - ?

Facebook created Hack[1]

[1] https://en.wikipedia.org/wiki/Hack_(programming_language)

also ReasonML

Google also has/had Borgmon, Sawzall, GCL and a few Python derivatives.

Why is Ethereum included in this group?

I extended GAFAM to GAFAME, but that's not a good thing for Ethereum:

https://twitter.com/nivertech/status/860970968143060992

I don't think Ethereum belongs on the list, considering that almost noone has heard of it.

As much as I disagree with Ethereum's design choices and all the speculation on it's ERC20 tokens, I still decided to list it as a 6th major tech platform TODAY - I hope that in the future we'll have something much much better.

Go to any blockchain or fintech hackathon and you'll see that Ethereum is a number one choice for devs (hobbyists, startups and even financial institutions and F500 companies)