This is a book for programmers who have experience in a non-Erlang language (C, C++, Python, etc.) but have not met Erlang and its "strange" properties (e.g., no variables (or at least variables that can only be written once which, to the uninitiated, therefore look a lot like constants)).

I had a particular professional need to get to grips with Erlang and bought this book as a paper copy (after failing to get it in eBook format on my reader---something that really annoyed me at the time). Like many of the O'Reilly books this one is thorough, authoritative and extremely clear. The examples are not "clever" as they are in many books: they are minimal and just cover the point being made. The authors have not allowed themselves to demonstrate their cleverness and ability with Erlang at the expense of the reader. I wish I could say the say for some other computing books.

If you are responsible for designing or implementing dependable (reliable and available) multi-threaded (or "multi-light-weight-processed" in Erlang) systems then get this book and switch to Erlang.

This book (also, incidentally available as a free pdf download but more convenient in book form) addresses the question that everyone working in the field of software in safety-critical systems must answer. We talk about probability of failure per hour of operation of our system being less than, for example, 1E-7 (IEC 61508 SIL3). Can we justify making such a claim?

The book does not present quantitative techniques for making the estimates and neither is it a statistics book (indeed, perhaps surprisingly, it contains no mathematics at all). What it does is make explicit the need for what it calls the "Three Es": Explicit Claims, Evidence, Expertise. I use the concepts from the book extensively in my work as a consultant in the area of software in safety-critical systems. I keep having to buy copies of the book because I keep giving my previous copies to customers.

The book has the weakness of a book written by a committee. It is somewhat repetitive but that is probably no bad thing. For specialists in the field it is a quick read (131 pages: about 3 hours) and contains a lot of gems. People familiar with Martyn Thomas' papers will be familiar with most of the content but it's convenient to have things gathered in one place in a form that can be left with customers.

Highly recommended.

This is a book on the direction in which a number of researchers think that the philosophy of mathematics is going.

Let me put my lack of credentials on the table. I studied mathematical philosophy many years ago (at the LSE under Lakatos, shortly before his untimely death) and have maintained a reading, rather than practising, interest since. I am therefore no more (and no less) of a mathematical philosopher than any other person walking along Sussex Drive in Ottawa. I was particularly taken a few years ago by Corfield's "Towards a Philosophy of Real Mathematics" as setting a new direction for mathematical philosophy.

So, this book contains essays (I would call them that rather than "papers", because none explores a topic completely) by 13 "young" mathematical philosophers setting their directions for the future. As would be expected, some of these seem to be potentially fruitful while others seem to be rather barren. In the first category, I would place Baker's "Mathematical Accidents and the End of Explanation", in the latter, perhaps unfairly, Colyvan's "Applying Inconsistent Mathematics". Reading Colyvan's essay, I was struck by how few examples of inconsistent mathematics were invoked: basically just naif set theory before the paradox and the infinitesimal calculus before Bolzano, Cauchy and Weierstrass. It's a fascinating question to ask: "does the (practical) use of inconsistent mathematics commit us to a belief in inconsistent objects?" Unfortunately for setting future directions of mathematics, Colyvan seems to give a clean, simple and acceptable answer to the question. Thereby closing it off.

If the essays in this volume are to be believed, the future of the philosophy of mathematics is in the pursuit of a resolution to the great Platonic/nominalist divide. It is understandable that this should be so (perhaps equivalent to saying that the future of moral philosophy will be dealing with "goodness") and Rayo's essay "Towards a trivialist account of mathematics" does make a stab at splitting the difference but I would have hoped for a more fundamental change in direction from these 13 researchers. It is understandable, but a little disappointing, to find in the index that the most quoted philosopher is Frege. The book seems to say that the future of the philosophy of mathematics is going to be much like its past (at least since Frege). I suspect that that is true but perhaps it shouldn't be.

I suppose a review should say whether a casual browser should buy the book or not. The answer is "yes". I have learned a lot and have been given reason to ponder and so would anyone else walking down Sussex Drive. The essays assume a little but not a lot of prior knowledge about mathematical philosophy from the reader; they are generally well-written (although too many are apologetic in their introductions: "I don't have room here to explore the full topic, so please bear with me as I address the whole topic superficially/just one aspect of the topic deeply") and are thought-provoking.

I have to declare an interest: I have used the TimeNet tool for drawing and analysing Petri Nets for a couple of years. This tool was created by Armin Zimmermann's group and is the focus of much of the book.

The book would be most useful to a reader arriving with some discrete event simulation background, preferably with Petri Nets. Without that background I suspect it would be heavy going: it embeds details of the various types of Petri Net within the text rather than dealing with them explicitly.

I was particularly pleased to see genuine (rather than toy) examples of the use of the techniques included.

A moderate level of mathematical, particularly statistical, ability is required in the reader (say that gained by a mathematics A level in the UK) to appreciate all of the text.

My reservations about the book are two-fold:

1. I wonder whether more could have been made of the practical issues of handling (and, in particular, maintaining) large, hierarchical Petri Nets.

2. The English is sometimes non-fluent and difficult to read. Several times on a page I have had to reread a sentence to get its meaning. I understand German sentence structure and that helps but, on occasion, I have actually been unable to understand a sentence.

I stumbled across this book and am amazed that I had not heard about it before. Since buying it, I have kept it by my bedside and have now read the whole book four or five times, picking up more of the subtleties at each reading.

The proofs are almost all magnificent (although I wonder how Buffon and his needles got in there) and even the well-known and time-honoured ones have a new twist or new extension.

The level of mathematics required to follow the proofs is reasonably low (high-school 'A' levels in the British system, no idea about other countries) although the book gives a deeper explanation in some areas (e.g. trans-finite arithmetic) than in others (e.g. number theory). I wonder if this unevenness reflects the interests of the authors.

But these are tiny nit-pickings. This is a wonderful and inspiring book and reading it should be made compulsory by the government in all high-school mathematics classes.