Lambda the Ultimate

There have been several threads recently about buying old computer science books, so I thought I'd make it into its own topic. What obscure books are out there that are worth searching for?

One that I've found interesting is "Lucid: The Dataflow Programming Language." There's also a followp, "Multidimensional Programming" (which haven't seen).

Occam books are also eye-opening.

And I have a copy of "Programming in Parlog" (a parallel Prolog variant) which I have yet to read.

For the time being I recommend section 2 of Wadler's Monads for functional programming.

In my neverending quest to grasp continuations (stupidly, apart from actually trying to use them in a supporting language), I found this "continuation sandwich" example interesting.

I wrote down my thoughts on what I still don't understand. If anyone would be willing to comment there or here and help me to correct my still-confused understanding of continuations I'd be very grateful.

SIGAPL is showing signs of life. After skipping a year while a new slate of officers reorganized, *Quote Quad* (the SIGAPL newsletter) Volume 34, Number 1 appeared in my mail box. The SIG has fallen on some hard times, but I think it's well positioned for a comeback. With cheap, fast hardware and service oriented architectures, the productivity advantages of *Array Programming Languages* are compelling for some applications.

Here's a flash cartoon outlining the future of Java after the settlement. Warning: Geek humor alert. :-)

(via slashdot and OSNews)

Bossa is a framework for scheduler development, including "a domain-specific language (DSL) that provides high-level scheduling abstractions that simplify the implementation and evolution of new scheduling policies". The DSL compiles via C to kernel code.

Interestingly, the DSL includes constraints such as "the absence of pointers, and the impossibility of defining infinite loops". A good example of a language that forsakes general-purpose features in order to provide verifiable safety guarantees.

The quickest way to get a flavour is probably to read the release notes.

There's a discussion about favoured programming language features on Slashdot.

Unification, parametric polymorphism etc. are mentioned.

From the OASIS group, responsible for OpenOffice formats among other things, comes a new schema for business processes. The gem here is the new standard for addresses, something long overdue in the XML world.

Ever since my early data modelling experience, with a pre-release version of IMS/DB, deciding how best to deal with addresses has been a major issue. So I looked at this area in particular and it does seem to work well and give the flexibility needed (this is partly due to the greater flexibility of XML over IMS) and I can easily see this becoming the standard.

I know I have made mention of this before, and I would like to thank many for making comments, espicially Dan Shappir. But I would like to throw an idea out to those of you, the many, that are smarter at this than I am. =)

The paper I have read concerning JavaScript 2.0 doesn't hold a lot of water with me; not necessarily because of the technical modifications being made, but because the future ECMA standard is leaning towards what is 'en-vogue' instead of sticking to what makes JS distinct. I do feel that JavaScript needs some changes, and I would like to ask the opinions of everyone as to what changes they see that need to happen, as well as listen to one of my own.

One change I would like to see JavaScript is the addition of being able to declare identifier's to be of a certain object type. I was always a fan of dynamic typing merely because I have always been bad at planning my code, so I always felt that static typing got in my way, and caused many edits in my code. I of course blamed this on the language, instead of what the real problem was... namely me, and my poor skills.

Although I see the benefits of typing, mainly for type checking. I do not think it is a panacea, and also think that JS's 'var' keyword provides a suitable syntax for declaring dynamic identifiers. Not to mention, 'var' can be implicit when an identifier is not declared, but instead immediately assigned.

This is just a thought I had run through my head, and over-all I think that it is a good idea. I would like to hear other's opinions concerning the problems with JS as well as my suggestion for a change in the semantics of JS.

Best Regards,

Mark

The Role of the Study of Programming Languages in the Education of a Programmer
Daniel P. Friedman

Surely we've mentioned this before, maybe not?