New DSL Book

Background

About five years ago, I co-wrote a book about Model-Driven Software Development. The original German edition was published in early 2005. The updated English translation became available in May 2006 and the further updated, second German edition was in stores in May 2007. While I don't have the exact sales numbers of the book available, I think in total 5.000 copies should have been sold. That's of course not earth-shattering, but also not bad for a technology textbook. I've received good feedback from readers, and I know the book is still used in universities today.

However, there is a big problem with the book: it is hopelessly outdated. The tools have changed a lot! Not just in details, but fundamentally. Also, my own thinking about the topic has evolved. This is reflected best in the fact that today I mostly talk about DSLs, and not MDSD anymore. So, an update is in order. However, I don't think that incrementally updating the old book makes sense, so I have decided to write a new one.

The New Book: DSL Engineering with Language Workbenches

The new book will reflect these changes. It will be called DSL Engineering with Language Workbenches and it will become available end of 2012 or very early in 2013. I had originally planned to write the book together with Eelco Visser, but we noticed that the two of us don't have a completely compatible approach to writing, and we don't have the same schedule/priority, so I am going to take the lead. Lennart Kats from Eelco's group will still contribute his experience with grammars and the Spoofax Language Workbench. And one section is heavily based on a paper I wrote with Eelco. See below for more details about the structure of the book.

Since the publication of the old book, a lot has also changed with regards to books and book publishing. So the new book will certainly be available as an ebook for the Kindle. I am not sure yet whether there will be a printed version; but I guess there is always on-demand printing, and I will take a close look at the possibilities. I am most likely not going to go with a "real" publisher, so in any case the book should be very cheap to get.

Structure

The book will consist of four parts. The first one will be an introduction to the topic and the book. It will be written in a very personal style and describe my "view of the world" regarding DSLs. The second part will cover DSL design: how do you structure adequate DSLs, and how can you design several DSLs to play together. This part is mostly tool-independent and will use a set of case studies to illustrate design alternatives and decisions. Part three will address DSL implementation. We're going to go through several DSL implementation aspects (such as structure, syntax, scopes, validation, IDE support, etc.) and then illustrate these with examples from Xtext, Spoofax and MPS. The book is not intended as a tutorial for any of these tools; we merely use them to illustrate the concepts. Part four addresses using DSLs in various aspects of software engineering. There will be chapters on product line engineering, requirements, architecture, and implementation. Some of these chapters will be case studies or based on existing papers and articles.

Current Status

As of January 2012, the introduction and the design chapter has gone out to reviewers and as I am getting the feedback, I am integrating it into the chapters. Lennart and I are also eagerly working on the implementation chapter. This should be ready for review by May. Part four will be addressed in the second part of 2012, so we can do copyediting in late autumn and then publish by the end of 2012 or early 2013.

Stay in Touch

I will update this page from time to time to reflect substantial new developments. However, if you're interested in the book, you should follow my blog, my twitter or my Google+ page.