Gosling on Ruby

Does that mean we might see Visual Ruby :)

The thing is, Ruby *isn't* a bystander, it is a serious threat to Java, at least judging by the number of Java programmers learning RoR. Chances are that it may be a bigger threat to Java than C# is, as C# at least is bound to one platform (mono being incomplete and all).

Ross,

My comment about innocence was really about the lack of a large (and easily villified) vendor behind Ruby.

BTW, I agree that Ruby stands to eat at least as much mindshare from Java as we do.

I also am hopeful that we're SLIGHTLY less vulnerable to Ruby than Java is, but I hope none of my colleages in the big house get any comfort from that hope ;-)

One thing I do know is that Ruby has pretty much eclipsed Java on my competitive radar.

Once the Ruby community has its own Eclipse competitor, I think the fat lady will have sung for good and Java will take its rightful place as the successor to COBOL. BTW, that's an enviable market position, but it's definitely NOT where the action is.

DB

google hired Guido (python BDFL)...Microsoft hired Jim Hugunin (jython, IronPython). C,mon Anders....Turbo Ruby.NET.

I don't know about Visual Ruby, but this might interest you:

http://plas.fit.qut.edu.au/rubynet/

The quote is somewhat out of context. See http://www.artima.com/forums/flat.jsp?forum=276&thread=153221 for a fuller transcript.

Regardless of the quote though, is there a contradiction in your comment? If Ruby is a genuine threat rather than an innocent bystander, then doesn't it follow that Gosling should play it down and be (unfairly) dismissive?

Don WinFX is (I hope) going to kill Java. C# is already everything Java should've been but now that you've (finally) got the whole WS-* stuff sorted out it's over for Java. Don't you find it ironic that Java proponents who praise open standards are now the ones who are not embracing them and instead are going down the JMS dead end?

Adrian,

I think it's the job of marketing folks to dismiss/fud the competition.

I hold technical leaders to a higher standard.

DB

http://beust.com/weblog/archives/000382.html
"The problem is that by now, there should be other obvious success Rails stories, and not just ones developed by the Rails Society.... lot of companies evaluate Ruby on Rails but will only take the jump if they can find evidence that other companies have done that before them."

This very sharp analysis. Ruby is *thought* leader right now, yet in thought only. Far less are willing to bet the farm on it and develop significant ruby-only codebase. It would be great if Anders is indeed loosing sleep over it. The road of adopting Ruby via ruby-ization of C#, PHP and other legacy languages is far easier to accept (via introduction of non-breaking syntax constructions)

Couple of years ago I got some C++ work thru Rentacoder, while this is a highly unscientific observation, it's still interesting to compare the number of projects (looking for coders) in different languages.

Surf to http://www.rentacoder.com/RentACoder/SoftwareCoders/BrowseWork.asp and scroll down to the Language Specific offers. The number of C++/C projects are approx. double that of Java projects, and that quota has been the same say since 2003 or 2004. And C# runs about the same number as Java (same as in 2004 I think). But now PHP has 50% more projects than C++/C. And I don't remember Ruby even being in there in 2004...

Ruby isn't a *thought* leader, while vastly better than Java, it's still just a poor implementation of Smalltalk with agol syntax. Ruby has one, and only one, contribution to language design, it's mixins. Beyond that, it does nothing that hasn't been done for years and years by Smalltalk and Lisp, that makes it a follower, not a leader.

Ramon,

Smalltalk and Lisp WERE though leaders and in fact both languages altered the face of the PL industry deeply.

In terms of impacting the way people think today, the innovation is happening downstream in other languages that hopefully are informed by the lessons of ST and Lisp.

In my own work, I remain deeply inspired by ST and especially Lisp, but in terms of influence on the generation of programmers coming up the ranks now, Ruby is what's blowing people's skirts up and is the lens a lot of people will view features like lexical closures or call/cc through.

DB

'Its the productivity stupid'
We are still waiting for a decent Ruby and RoR Development environment for use by the masses, not just for the geeks. On my Desktop I have, Arachno Ruby, JEdit and Eclipse Ruby Plugins, and have tried varius other environments. But these are for Geeks, and I still don't feel the productivity a simpleton programmer like me can achieve with Visual Studio C# combo. Ruby will only be taken seriously by the development industry when its tools support real productivity gains.

[RoR also needs a decent web develpment environment, I can build simple Web Applications very rapidly, but for real applications I am still stuck with designers developing all those pesky CSS sheets.]

If I could pick just one thing that differentiated (sp?) Java and c#, it would be default final methods in c#.

In that light, I'd have serious doubts Anders is even the right guy for Microsoft to have on the job of imagining a Ruby-beater. Especially considering his reasoning behind default final methods.

Then again, I never developed in Turbo Pascal, so maybe I'm off base.

But it's the anti-freedom aspect of c# that makes Ruby such a danger, and Extension Methods aren't going to solve it alone. Duck Typing is just as, if not more important. Getting rid of some of the excessive amount of compiler cues in c# might help too. Half my code in c# seems to be nothing more than:

"This is a method, it belongs to this class, it's public and returns a string; it needs you to pass in a byte array and a boolean value. First we'll open a database connection in a using block so the IDisposable interface is called, we'll declare the Type of the connection twice in the same line even though it's never used outside of this scope." ...etc...

As many problems as I have with ActiveRecord's design, it's obvious why there's nothing like it for .NET, despite the Microsoft platform being much more mature than Rails. By crippling libraries like AspectSharp, that could make it possible, because default virtual methods would be too "dangerous", Microsoft stalls innovation. With only ASP.NET on the web-side, I for one seriously doubt c# or the platform in general will ever sway lots of developers who prefer an MVC model over a Component/Event Oriented model. Perhaps if Monorail had an official blessing of sorts...

> In that light, I'd have serious doubts Anders is even the right guy for Microsoft to have on the job of imagining a Ruby-beater. Especially considering his reasoning behind default final methods. <

Please tell me you didn't just say that.

Obviously you did say that, so let me just jump right on in and ask...

"Are you a ----ing idiot!?"

> Then again, I never developed in Turbo Pascal, so maybe I'm off base. <

Okay, so maybe your'e not an idiot, but you do need to spend some time learning more about what make Anders Hejlsberg the right man for ANY job.

I would start here > http://en.wikipedia.org/wiki/Anders_Hejlsberg

This isn't the first time Gosling has showcased his belief that if its not Java, its not anything important...

Please see: http://dev.extensibleforge.net/wiki/dotnet/dotnet%2Bvs%2Bjava%2Bvs%2Bagility for more on this topic (to many links to various .NET language projects to get past the spam guards, so I've posted it to the previous link instead)

Hi Don,

I 100% agree what you say about Ruby. It's the combined effect of Ruby the language plus Rails the framework that is capturing the enthusiasm of so many people. For those who point out the lack of commercial applications that currently use Ruby+Rails, the same was true of Java when it was in the early days. I think it's the *trajectory* of Ruby of Rails that indicates that it is worth taking seriously.

Regards,
Graham

p.s. I would *love* to see Ruby and Rails running on the CLR.

David, I can appreciate being a fan, but I think it's worth pointing out that default final methods have _castrated_ the .NET platform when it comes to O/R Mapping, AOP, DesignByContract, and any number of other genuinely useful technologies.

I'll take a few versioning issues over _that_ handicap any day.

The simple truth is Rails can't be done in .NET because of that decision. Sure Hammet created a very nice MVC Web Framework in Monorail, and developed the technology to get NHibernate as close to POCO as it is, but it's not Rails, and it's not really anything like ActiveRecord.

Anders may be a genius, but we all do stupid stuff, no matter how smart we are, and IMO, default final methods was a bone-headed move.

Don,

Why don't you take your own advice and make Matz a deal he can't refuse?

Via Don Box
I ran across James Gosling statements about scripting languages @ Sun’s Worldwide...

Patrick Logan does make an interesting point though:
http://patricklogan.blogspot.com/2006/05/losing-war-no-not-that-war.html

Microsoft has an answer to ruby: IronPython!
Why so little talk about IronPython?

In terms of capabilities the languages are similar and both easy to use.

In his blog, Don Box writes:

From where I sit, Ruby has the language thought leadership position and is the competitor I hope AndersH is losing the most sleep over nowadays.
Then later in the comments to the same article, he says:

I agree that Ruby

Ruby's star&nbsp;is&nbsp;rising fast at the moment.
It's star is rising fast it even has the thought...

Just in case someone reads this in the future and needs a Ruby IDE try http://www.sapphiresteel.com/ Yes - I know, Ruby on Windows - how bizarre ;)