Lambda the Ultimate

jhc is a haskell compiler which aims to produce the most efficient programs possible via whole program analysis and other optimizations.

This seems like an interesting project, for example: Region Inferencing, Compilation by transformation with 2 general intermediate languages, very modern design, using rank-n polymorphism, monad transformers, generic programing, and existential types.

Note, howver, that there are quite a few problems (scaling, memory leaks, etc.)

Maybe some of you might want to offer a helping hand...

Datatype Laws without Signatures

Using the well-known categorical notion of `functor' one may define the concept of datatype (algebra) without being forced to introduce a signature, that is, names and typings for the individual sorts (types) and operations involved. This has proved to be advantageous for those theory developments where one is not interested in the syntactic appearance of an algebra.

Does it sound like "a module without a signature"?

If you like programming with bananas, lenses, and other weird things you might like this paper as well.

PS: OTOH, if you are sceptic about bialgebraic programming, then dialgebraic is definitely not for you.

From the Haskell mailing list.

People often wonder about Y in Haskell, so I think it is worth to have this link in the archive. Nothing new here for Haskell mavens, though.

If you're a GHC user, the Glasgow Haskell Compiler HeadQuarters needs your feedback! See Simon Peyton-Jones original message, or go directly to the user survey. Here's a quote from the original message:

We'd like to hear from *absolutely everyone* who uses GHC: whether you
are a first-year undergraduate or a famous professor; whether you use
GHC for industrial applications or recreation; whether you use
higher-rank polymorphism or have only just learned what functional
programming is. Everyone!

We'll use the information to guide our future priorities, and we'll
publish some kind of summary of what we learn in due course. It's all
anonymous, of course, unless you choose to say who you are.

Gianluigi Ferrari, Eugenio Moggi and Rosario Pugliese

MetaKlaim - a Type Safe Multi-stage Language for Global Computing

This paper describes the design and the semantics of MetaKlaim, an higher order distributed process calculus equipped with staging mechanisms. MetaKlaim integrates MetaML (an extension of SML for multi-stage programming) and Klaim (a Kernel Language for Agents Interaction and Mobility), to permit interleaving of meta-programming activities (like assembly and linking of code fragments), dynamic checking of security policies at administrative boundaries and “traditional” computational activities on a wide area network (like remote communication and code mobility). MetaKlaim exploits a powerful type system (including polymorphic types ´a la system F) to deal with highly parameterized mobile components and to dynamically enforce security policies: types are metadata which are extracted from code at run-time and are used to express trustiness guarantees. The dynamic type checking ensures that the trustiness guarantees of wide are network applications are maintained whenever computations interoperate with potentially untrusted components.

Ωmega is a new programming language by Tim Sheard which is descended from Haskell and adds new facilities for defining static type constraints, such as allowing "users to write functions at the level of types, and then use those functions in the type of functions at value level". It also has "equality qualified types". See also Programming with Static Invariants in Omega and the manual for more information. Mentioned previously (in passing) on LtU.

(via Keith)

Higher-Order Perl is about functional programming techniques in Perl. It's about how to write functions that can modify and manufacture other functions.

Why would you want to do that? Because that way your code is more flexible and more reusable. Instead of writing ten similar functions, you write a general pattern or framework that can generate the functions you want; then you generate just the functions you need according to the pattern. The program doesn't need to know in advance which functions are necessary; it can generate them as needed. Instead of writing the complete program yourself, you get the computer to write it for you.

The book was published on the 8th of March, and the text will likely appear soon on the web site and will remain freely available.

What with all the foolish notions about lambda be passed around since the start of this month, a nice introductory article on Using functions for such higher order purposes as arguments, function-generating functions, and anonymous functions is just what the doctor ordered.

Functions are the wonderful and powerful building blocks of computer programs. Functions allow you to break code down into simpler, more manageable steps. They also allow you to break programs into reusable parts -- parts that are both reusable within the program and in other programs as well. In this article, learn how to create new functions at runtime based on templates, how to create functions that are configurable at runtime using function parameters, and how the Scheme language can be a valuable tool with functions.

A lot of language theory goes past here on Lambda the Ultimate, but we rarely see that theory directly impacting commercial programmers.

I'm a great fan of theoretical concepts like arrows, but at the same time I'm a self-employed programmer interested in solving my clients' problems.

Pugs is notable in that it profitably uses recent developments such as GADTs and Template Haskell for an implementation of Perl6.

I recently became a regular on the #perl6 irc channel and soon after joined the list of committers.

In just a few days I've seen a lot. I've seen enthusiastic members of the Perl community learning Haskell. I've seen myself learning Perl. I've also seen how daily Perl programmers work with abstractions like monad transformers. I've seen how some structures are easy to extend for programmers new to both the Pugs codebase and Haskell.

The Pugs project was started 64 days ago by Autrijus Tang, as an exercise while reading TaPL. Pugs already includes network and threading primitives. New tests and code are add at an amazing rate, as evidenced by the smoke tests.

I don't know if I'll end up using Perl after Pugs is written, but I am learning how to practice the theory of programming language design and implementation.

Along the lines of the quine discussion happening in another thread, Manfred von Thun wrote a new page for the Joy site, Reproducing Programs implement Lazy Lists. This connects nicely with Jeremy Gibbons spigot algorithms paper, one example of which is his IOHCC 2004 submission PiSpigot.