Hacker News Reader

Comment by adamc | Hacker News Reader

Makes me wonder whether they will bring out a competitor to Java, maybe for Android.

They already have at least one world class language designer in Guido. They also have the Go language.

Point being that if they wanted a competing language to Java, they probably already had the talent to get it done. (Or just, you know, use Mono/C#).

Arguably, they already did create a competing language (ie: Android)

This is a very solid point. Having an internally developed language like Go (which is, according to Rob Pike, being used "for real stuff" inside of Google), the creator of Python (one of the most widely used languages in recent years) and the creator of Java (one of the most widely used languages for at least a decade now) under one roof has to be an embarrassment of riches for any group intending to innovate in the programming language space.

Nah, that's too many cooks. They all want different things.

Guido is interested primarily in usability, and Python is notoriously poor in performance.

Rob Pike writes systems languages.

With Java, Gosling has written a very enterprise-friendly language, that drags C++ halfway to Lisp.

Nit: Halfway to somewhere, but that somewhere ain't Lisp.

I'm under-rested and under-caffeinated, but the only thing I can think of that I'd call an intentional similarity between the two languages is the fact they're both garbage collected.

Cobol? :-P

Smalltalk would be a good non-snarky answer and one with some basis in history, although Smalltalk seems to have been a much stronger influence over the JVM than the Java language.

FWIW, none other than Guy Steele has cited Java as dragging C++ programmers halfway to Lisp: http://people.csail.mit.edu/gregs/ll1-discuss-archive-html/m...

I'd agree with him, given that this statement was about programmers and Java really did open the floodgates for a diversity of programming languages to become accepted, or at least tolerated.

The Java language itself, however, is not very close to Lisp at all.

I think that Guido probably has his plate full with the open revolt over the transition to version 3.

Javascript is much closer to a halfway point to Lisp than Java. I don't really see Java as being at all similar to Lisp... witness the 3-4 year (possibly longer) debacle trying to get closures into the language. Talk about banging a square rhinoceros into a round hole.

Java's point of greatest similarity to C++ is undoubtedly Generics, which has all the syntactic disadvantages of C++ templates, but none of the meta-programming advantages.

But one of the ways of seeing that C++ and Java are quite dissimilar is to first grok Java Interfaces, and then read through the smalltalk/C++ version of the GoF book Design Patterns, with a view to how many of the Patterns don't even make sense in Java or are obsolete in Java because of Interfaces.

Thought I should point out MS also has quite a concentration of languages folks that were famous before joining MS:

Butler Lampson, Erik Meijer, Herb Sutter, Burton Smith, Anders Hejlsberg

+1 for Anders Hejlsberg. He created Turbo Pascal, Delphi and C# - that's 3 very successful languages/development platforms.

He created Turbo Pascal and Delphi when he was young and hungry. C# when he was well-fed and older. Not hungry in any way. Thus the over-engineering and bloat-ware.

... and Simon Peyton-Jones (Haskell)

Add Josh Bloch, Google's Chief Java Architect to that list.

I threw up a little in my mouth at the thought of James Gosling going to Google to work on Mono/C#. That is quite possibly the most retarded piece of speculation I've seen on the internet all year. How on earth you got 18 upvotes I do not know. Maybe Bruno has a spider crawling the net for references and then sicked his elite team of Argnetinian cyber-ninja-commandoes on it.

Yup, that and saying that Python and Go are well designed languages.

Once you're done throwing up please re-read what I wrote. I merely suggested that if Google wanted an alternative to Java they could use c# which is a similar (and probably better) language instead of inventing their own.

Be that as it may I still think you're delusional for suggesting that Google should jump on the Mono bandwagon, since Mono directly benefits their biggest competitor.

Moreover, in what way does hiring Gosling suggest that Google are seeking an alternative to Java? Is Gosling a closet C# fancier? I strongly suspect the opposite.

Mono directly benefits their biggest competitor.

Mono directly benefits Oracle?

I think you need to re-read the whole thread more carefully. I never suggested that Google needs to adopt a java competitor (and I don't think they do), much less Mono. I was only responding to the hypothetical that someone else raised and said that if Google wanted a java competitor, they could just use C#. I personally can't see why they would want one.

If anything, they'll try to continue with Java instead, I'd expect. They're a big user internally.

Even if Java is thrown out there is still NDK, not to mention a host of other language inventors at Google ;-)

Gosling would provide the credibility necessary if Google wanted to usurp Oracle as the provider of an "official" version of Java and/or the JVM. Big Corps are terrified of Oracle. I worked for a Fortune 100 which was all about using Weblogic vs. open source offerings like JBoss. Then, after the Oracle acquisition, the directive was to "explore competing options". Oracle's history of shakedowns is so well established that Google or someone similar might have a shot. I was surprise that IBM didn't make a better attempt to steal Sun's place.

One thing to note is that I don't entirely understand the IP issues surrounding Java.

IBM was a competitor to Sun. An IBM acquisition of Sun definitely would have triggered the government to get involved and explore the anti-trust/monopoly angle.

What surprised me was how IBM decided to get close to Oracle instead of fight for improved position in the F500 using Java as lever.

I doubt that. Gosling hasn't worked on anything related to Java since JSR-1 (that was before 2000).

He checked out from the Java world a while ago.

I'd counter that it's important to specify the time-frame you're talking about. For instance, would you say that they'll try to continue as a big Java user for the next 15 years?

When you stretch out the time-frame, your assertion sounds substantially less credible. At least to me.

Google has an immense amount of Java code, in addition to all the Java programmers it took to write them, a good number of whom are programmers that have written Java for their entire professional careers.

If the bar for getting away from it is rewriting millions of lines of code and retraining 1/3 - 1/2 of their engineers, it seems credible to me.

I'd be very surprised if they replaced all their Java code within 15 years. But even if they do change over, it's still a very good investment for them to continue maintaining and updating Java for the short term.

Or something as cool as NeWS (http://en.wikipedia.org/wiki/NeWS)

They're certainly stacking the decks with lots of ex-MSFT-ies with experience working on Microsoft's version of the JVM - the CLR. What will the name be? GLR? GVM?

Reply on news.ycombinator.com