Customer Reviews

Elements of Ml Programming

5 star:		(4)
4 star:		(1)
3 star:		(1)
2 star:		(0)
1 star:		(0)

 

 

I appreciate the consistency of exposition, especially when explaining recursion of execution and recursive datatypes using basis and induction cases. It's a book for those who are interested in the crossover between CS theory and mathematics, but only introductory calculus is necessary (and that's only in one section). If you come from a procedural programming background, as I did, this book is an excellent introduction to the techniques of functional programming. I would recommend reading this in conjunction with ML for the Working Programmer by Paulson. If you are looking for a gentler introduction to functional languages, like LISP and its sister Scheme, try The Little Schemer by Friedman and Felleisen.

When tasked to quickly learn the basics of the ML programming language for class and be expected to program an Abstract Data Type, I was a little bit frantic. In the short time given to me to prep for said programing project, I raced around looking for sources that would improve my understanding of the language and one of the sources i found was this book that was also recommended by the professor.

This book, Elements of ML Programming, has its reader approach ML from the vantage point of either someone who has never programmed but has some understanding of Programming theory or someone who has programmed before but never touched a Functional Programming Language. In that respect it excels at explaining the unique features of the ML language and how best to use and write in it.

Although this is an introductory book, I also recommend it to advanced programmers. Concise and well written, it is a joy to read. Programmers who have little knowledge of modern functional languages will benefit immensely from this book. It is a great recreational read.

Examples and exposition are overall great. Jokes are clever and the tone is light through the book. It reads much like the Perl book by Lary Wall. It hits topics a bit slower than SICP. The biggest problem with this book is it left me felling "Ok, I know the syntax/semantics, now what?" This is a common problem with intro to programming books, and I don't fault it too much for that.

The biggest weakness of this book in my opinion is it didn't strongly teach a "style" that I can adopt into my own programs (but then, no FP programming book I've ever read does). It did great teaching functional programming, but as with most intro-to-fp books forgot that real programs don't fit in one file and are several thousand lines of code.

Sean

While I find functional programming, and ML/SML in particular to be most annoying, this particular book is helpful in illuminating some key concepts of the paradigm. However, there are certainly sections where complex concepts are either glossed over, or explained in a very confusing manner (take Section 5.3.1, for example, which attempts, in a somewhat confusing manner, the limitations of polymorphic functionality in ML and it's relationship to types. A dense topic to be sure, but also poorly explained.) Additionally, there are some technical errors to be found in the book (eg: things that 'dont work' which really do, incorrect error messages, etc), which I mostly attribute to SML having been updated since the author published the book (not really his fault).

Certain things, ARE however, the author's fault. In particular, he addresses the audience with needlessly confusing 'Basis'/'Induction' definitions for many many things. For example, a binary tree is defined in such a way. This can be incredibly confusing, espcially if one does not already understand the concept of a BT or a BST. While ML/SML is a mathematical language, the author assumes too much formalism in his definitions, enough that the reading feels like reading a thick algorithms book at times (not a pleasant experience). Suggestion to author: Explain things simply instead of trying to coerce definitions into formalisms in which they do not belong (in particular the induction formalism which is very often confusing, use induction only when TRULY needed).

On the whole, ok to learn from, but can be overly verbose in some areas (eg: the belabored explination of all the error messages) , yet unclear in others... (eg: the density of several sections).

Also, as a parting note, there is not too much assumed about the reader, however, it helps to have a broad general computer science knowledge (as well as basic math, eg: calculus) unless you pickup new side-concepts very quickly. For example, in one example, the author uses the trapezoidal approixmation as an illustration of higher order functions. This is well and good, assuming you know calculus. While he does 'explain' the ideas (eg: what a binary search tree IS) if you havent had it presented to you before, you're in too deep.

Important: This is a ACADEMIC book, this is not something to pick up and read, it is far too dry for that. If you're looking for a good learning book, look elsewhere.

This book covers the ML programming language. Unfortunately this language is rarely used these days as a functional programming language . This is a 97 edition which is again more than a decade old. If you are a person looking to learn a functional programming language i recommend learning the newer language such as Haskell. I used Haskell for my course recently.