Personally, I would love to see something like IronLisp, ie. an implementation of Common Lisp on top of .NET and the DLR, similar in concept to IronPython and IronRuby. Such a language could directly integrate with all the .NET framework libraries and tools (including Visual Studio).

For the Microsoft/.NET community, learning such a language would be relatively easy since they already know the tools and libraries, so all they’d have to learn is the language itself. Also, functional programming is all the rage in the .NET community now anyway, with F# soon becoming a “first-class” language integrated with Visual Studio etc.

On the other hand, those who already know Common Lisp would only have to learn the typical .NET tools and libraries.

IronLisp actually exists – see http://www.codeplex.com/IronLisp. However, the author apparently dropped it in favor of IronScheme (http://www.codeplex.com/IronScheme). But the code for IronLisp is still there, so a new project could be built upon it.

As IronPython and IronRuby have demonstrated, Microsoft is surprisingly open to input from the open source community these days – and they’ve got the resources and the market significance which it takes to make a programming language successful. Both IronPython and IronRuby, by the way, try to be absolutely true to the language, which is something I’d find of high value for an IronLisp project as well.

There have been attempts of implementing Lisp on top of .NET before, but AFAIK, they all deviate from the Common Lisp standard, or implement only a tiny part of it.

I’ve heard your counter arguments from the community before. The problem is simple: Lisp and it’s community has an image problem and it’s not friendly towards new developers. Ideally, the community would look in the mirror and come to terms with these problems. It’s a matter of empathy, you need to understand what the outsider is experiencing.

It isn’t my intention to be derogatory or demeaning to you or anyone, but to be blunt, do you realize how you sound?

If you’re one to simply state everything is fine the way it is, and it’s the newbies’ problem, then why did you bother to speak?

I’m a professional game developer, but a newbie to Lisp, and one of my biggest problems has been trying to find an implementation that is consistent across platforms, and one that can utilize other libraries and interfaces present on those platforms. Is it out there? Fuck if I know, because what documentation that is out there is too scatter brained for me to figure and sort out. There is no good starting point.

Why would I pursue this mess when there are more consistent avenues? I can conceptually understand there are merits to the language, and I can think of no better way to really put Lisp in the stage light than to make a bunch of noise that our XBox 360 game was written in Lisp and saved us time and money; but as it stands, I’d rather use Python.

I can’t fix these problems. I can’t do it alone. But if Lisp was somehow reborn, these problems were fixed, I’d jump on it in a second.

> 1. There’s no default/most prominent implementation – finish SBCL threading work on Windows and OSX

I don’t want a default implementation. Current ones are suitable for different purposes (embedding, running on JVM, number-crunching, etc)

> 4. The learning curve is steep – there are dozens of valuable resources out there that beg to be organized. Watch this space for an upcoming announcement related to this

I don’t think it is as bad as people complain about. The problem is that there is a paradigm change from usual languages. You *need* to feel confused in order to finally understand Lisp.

But a nice way to start is with Peter Seibel’s or David Touretzky’s books. Simple like that: “read this book”.

> 6. Lisp’s community is weird – no, it’s just headless. Viewed from the outside, Paul Graham is the biggest figure in the Lisp world, but he’s more of a philosopher than an evangelizer. He’s also not a teacher (at least not to the masses – I hear wonderful things about he treats his YCombinator companies!) The Lisp community is like the Mongol hordes: talented and powerful, but needing a Genghis Khan.

The Lisp community doesn’t need a leader for two reasons:

- Additions to the core language will NOT happen. The standard is there, and noone will change it.
- Lispers don’t even want a “standard set of libraries” or anything like that. They like Lisp the way it is. None of them wants a leader.

– w-g