How I Learned to Stop Worrying and Love the New Axis of Evil (Oracle) | Hacker News Reader

How I Learned to Stop Worrying and Love the New Axis of Evil (Oracle)

lessonsoffailure.com

30 pointsgacba10 comments

> Java not good enough for Android? Fine, let’s make a new language that finally innovates on the mobile device, unlocking us from the collective disasters of Objective C, mobile Windows, and bloated Java ME.

While I appreciate the sentiment of this post, it displays some ignorance of the actual legal issues at play. Oracle's patent attack has nothing to do with the Java language. They have to do with virtual machines, and any VM (even clean-room like Dalvik) running any language is going to run afoul of them since they're absurdly broad.

Never mind that VMs are as old as UCSD p-code.

http://en.wikipedia.org/wiki/UCSD_Pascal

edit: Interesting voting here, would someone please explain what's so bad about UCSD p-code being a primitive byte code interpreter + associated VM ?

Modern VMs use some techniques for optimizing runtime dispatch, or for security, or for dynamic loading of code, etc... techniques which are newer.

Sun did research on Self, and Strongtalk which they've bought was heavily influenced by Self, which later became Hotspot.

So Oracle might own really relevant IP to modern implementation techniques for a VM ... for which there might NOT be prior-art. And some of those techniques are really not trivial.

This is just speculation btw, without a careful analysis of their patents. But if I'm right, Oracle can nuke all modern VM implementations.

The parent comment wrote explicitly: "and any VM (even clean-room like Dalvik) running any language is going to run afoul of them since they're absurdly broad."

But its actually not that bad.

This applies to Apache Harmony, too. Spot the difference here? Apache Harmony is open source, not big business. Android actually incorporates at least parts of Apache Harmony.

If Oracle sues the Apache Foundation, you know what's going on.

Who votes for Scala developers free scala from java vm and .net with compiler for own vm or runtime?

And a self hosting clojure.

Let the engineers build the vm modular and plug everything that deserves it.

But i dont't know if this could be really fast accomplished, because the java vm has such a maturity.

Goodies could be modern concepts deeply incorporated and specific configurations for current and future hardware usage.

You mean start their own VM implementation? The cost of that would be huge (see, e.g. Parrot). The Scala developers are better off concentrating on making Scala better, IMO.

Not a fan of these strangelove references. First time was good, but it has become cliché

Reply on news.ycombinator.com