rmathew.com

"Programming Pearls"

2010-03-03T00:00:00+05:30

Programming Pearls by Jon Bentley is a book based on a collection of articles written by the author for the eponymous column in Communications of the ACM. True to its name, the book presents several pearls of programming wisdom based on the hard-won experience of a brilliant computer programmer who was also lucky enough to be associated with some of the brightest minds in computer programming (the group at Bell Labs). The icing on the cake is the clarity and brevity of the book. I have read the book at least three times over the years and I continue to learn new things from it - no wonder this book is considered a classic in computer programming.

The book is divided into three parts with five chapters per part. The first part presents the fundamentals of programming, from programme design to algorithms, data structures, programme verification and testing. The second part focusses on performance and shows how to perform rough estimations of running costs, improving algorithms, tuning code and saving space. The third part applies the principles of the previous two parts to problems in sorting, searching and string-processing.

The chapters are short, clear and peppered with interesting anecdotes. These make them easy to read, though it would be a terrible mistake to zip through them. The author himself recommends that you take the time to understand the concepts presented in a chapter and attempt solving the given problems yourself before taking a look at the solutions. The chapters only give you a flavour for some of the concepts - you'll have to supplement it with reading a suitable book if you are not familiar with these concepts. Here the author has some recommendations and the usual suspects (The Art of Computer Programming, Code Complete, The Practice of Programming, etc.) all figure prominently.

The author ends the book with a couple of amusing epilogues presented as interviews with himself. He takes a dig at himself for the frequent references in the book to his colleagues from Bell Labs. One of the appendices contains rules for code tuning adapted from his now-out-of-print book "Writing Efficient Programs". I suspect that an absolute beginner would have some trouble understanding these recommendations since they are presented here without much elaboration or examples, though there are references to related material elsewhere in the book. Another appendix presents a model for estimating the costs of various data-types and operations. These can be used to make quick back-of-the-envelope calculations about the time and space that will likely be taken by a given function during its execution.

Some of the material, like that on performance improvement or programme verification, might not be very fashionable today, but no serious computer programmer can afford to ignore it. Much of the advice in this book is timeless and will definitely help you become a better programmer.

Note that the web-site for the book as given in it, "www.programmingpearls.com", is not valid any more as a domain squatter has snatched control of it - the correct web-site is http://www.cs.bell-labs.com/cm/cs/pearls/.

The sequel to this book was published as "More Programming Pearls: Confessions of a Coder". That book is a little hard to find in most book-shops as it did not become as famous as this one.

"Code Complete"

2010-02-24T00:00:00+05:30

Books filled with good practical advice about constructing software are rare. Code Complete by Steve McConnell is a well-written rarity in this field and has a well-deserved reputation as a classic. It is one of those books that every computer programmer ought to have read. I had read the first edition, published in 1993, as a budding programmer and the book left a lasting impression on me. With the benefit of several years of experience, I find myself agreeing almost entirely with the updated second edition, published in 2004. When someone asks me "How do I write good code?", I point them to this book without hesitation.

The book is about "software construction" - as the author notes, this includes coding and debugging, detailed design, construction planning, unit testing, integration, integration testing, etc. The bulk of the book focusses on coding since most of the effort in construction is spent in coding. The emphasis of the book is towards creating a toolbox of effective software construction techniques and on the importance of using the right tool for a given job. In particular, the author repeatedly advises you to programme into a language rather than in it.

An endearing quality of the book is that instead of pontificating, the author offers copious references to papers, articles and books to back up his claims. The tone of the book is entirely non-condescending and you can see that the advice is coming from a wary professional with years of experience working on real projects with real people rather than a wily snake-oil salesman. Rookie programmers will do well to heed the advice in the book and experienced programmers will find themselves nodding every now and then as they read the book.

At around 960 pages the book might seem intimidatingly big, but it is worth spending the time needed to read it from cover to cover. The book is divided into several parts, each part comprising several cohesive chapters. Each chapter begins with an overview of the material covered in the chapter and ends with pointers to additional resources, a summary of the key points of the chapter and a check-list to help you verify if your code adheres to the guidelines prescribed in the chapter. The side-margins of a page are used to provide cross-references to material covered elsewhere in the book, insightful or funny quotes, a short URL on the book's web-site that contains additional material and icons to bring something to your attention. These icons are "Hard Data" (typically a reference to a study backing some claim), "Key Point" and "Coding Horror" (an example of badly-written code). This structure makes it easy to find the material you're looking for as you refer to the book from time to time.

The focus of the book is on imperative programming and it mainly uses code in Visual Basic, C, C++ and Java as examples. Functional programming enthusiasts might particularly take offence to the material on recursion where the author seems quite disinclined to use it and advises minimising its usage. The author gives an example of a recursive implementation of a factorial function and says that it is slow and has an unpredictable impact on memory. In fact, the code uses a tail-call and is therefore a prime candidate for tail-call optimisation using a decent compiler. A good programmer can use recursion to create neat functions without adverse effects in many a case.

The author rightly points out the perils of premature optimisation of programmes, especially that done without proper measurements to identify the real bottlenecks. The primary emphasis should be on creating correct and maintainable code with performance improvements taken up at the end, if it is needed. I generally agree with this, but would point out that many a time you would barely have time to tackle performance in the end if you're working on a tight schedule (as most of us are forced to do). Many programmes also have a flat profile in that there are no obvious bottlenecks and yet the programme is unacceptably slow. If you have automated regression-testing set up with comprehensive code-coverage (as the author recommends elsewhere in the book), performance becomes important even if it doesn't matter during a single run of the programme or during a typical user-interaction. In such cases, it might be better to keep an eye on the performance of a programme during most of its construction.

There are at least three important aspects of software construction that the author unfortunately leaves out of this otherwise excellent and comprehensive book: security, supportability and concurrency. Writing secure code is difficult, but very important and precious few programmers are aware of the necessary techniques. The supportability of software determines how effectively you are able to diagnose problems encountered during its deployment - logging, state-dumps, etc. are techniques that help here. Concurrency is important but is made difficult by constructs available in mainstream programming languages. To be fair to the author, it is only relatively-recently that security and concurrency have come into increased focus with the Internet and many-core CPUs being the prime drivers respectively.

With an additional 10 years of experience gained before the publication of the second edition of this book, the author has slightly revised some of his earlier recommendations and freely admits it. For example, the first edition recommended a pretty-formatting of code where the "=" operators of consecutive assignment statements line up like this:

int foo       = bar;
boolean snafu = wombat;

It turns out that while this looks pretty in the beginning, it leads to additional effort as the code changes (e.g. variables change names) during subsequent maintenance. It is not much worse to have a simpler formatting that does not place unnecessary burden on maintenance. The second edition also covers object-oriented programming and agile development methods, but presents them in an even-handed pragmatic manner. Much of the advice from the first edition remains intact in the second edition.

The quality of the Indian reprint, published by WP Publishers, is unfortunately quite disappointing. It has very thin pages because of which some of the text and graphics on the opposite side of a page show through. The ink doesn't have a good consistency and fades a bit in some places. These make it a bit hard to read this reprint. At a price of Rs 600, they could have done a better job.

If you are a computer programmer and wish to improve yourself, get this book and read it.

"The Practice of Programming"

2010-01-23T00:00:00+05:30

In their book The Practice of Programming, Brian Kernighan and Rob Pike aim to advise computer programmers on things like testing, debugging, style, performance, design, portability, etc. that they are not usually taught in computer science classes or programming courses. This is what they call the "practice" of programming. Many pick these up over the course of their careers with some trial and error; many simply don't. This is the kind of book that has lessons for both rookie and seasoned programmers and that deserves multiple readings over the course of one's career.

I first read this book about a decade back when I was a relatively-new professional programmer with a couple of years' worth experience. Reading this book again after all these years makes me appreciate the wisdom contained in these pages and agree with almost all that the authors have to say. In fact, I think this is a great approach to becoming a good programmer - learn the basics of programming, spend a couple of years writing lots of programmes for fun or profit, read books like this to improve yourself and come back to such books after several years, having applied their principles in the interim. If you read such a book immediately after learning the basics, you might not really understand the reasons behind the advice in the book and will have to just take the words of the authors at face value.

This book is a typical "Kernighan book" - it is short and to the point, the language is clear and simple, the material is logically ordered with a natural progression and some of the examples show that it is simple to implement things (e.g. a regular expression parser) that might appear hard otherwise. It is also an old-school Unix book in that it uses the command pipe-line:

grap | pic | tbl | eqn | troff -mpm

for type-setting instead of using something like a system based on TeX.

The programming languages used in this book are C, C++ and Java with an occasional sprinkling of AWK, Perl and Tcl. Most of the examples use C however and some of the advice can be considered specific to this language rather than being agnostic. That is not as bad as it seems since every serious programmer must have C in their arsenal and not be wedded to a single language or a narrow set of languages.

The book is full of pearls of wisdom that you might miss on a quick skim through it. For example, in a section on debugging a problem, the authors say:

Resist the urge to start typing; thinking is a worthwhile alternative.

When faced with a bug, most of us either fire up a debugger or start modifying our programme in the hope of making the bug go away. With experience we learn that carefully reading the code in question and thinking about it lets us resolve the issue faster and some times exposes other latent bugs. As another example, in a section on performance, the authors say:

Thus the first principle of optimization is don't.

Once again many of us spend a lot of time trying to optimise rarely-used programmes or portions of a programme that we determine to be bottlenecks based on reasoning and intuition alone. We have to ask ourselves if the programme in question is worth speeding up and then use measurements to determine the real bottlenecks. Even then, using better algorithms and data structures coupled with clean and simple code can provide greater benefits than complex code that tries to be clever.

The last chapter in the book goes into topics like creating domain-specific languages or compiling code on the fly with an emphasis on using the right language for the job. It is better to create little specialised tools that work with text files on their standard I/O streams and that can then be combined in various ways (using Unix pipes for example) to solve different problems. This is, of course, the Unix philosophy. As an example, the authors create scripts for getting the document for a given URL, removing HTML tags from a document and formatting a text document into 60-character lines (written in Tcl, Perl and AWK respectively). They then combine these scripts to create a very simple, command-line-based, "web browser" that shows a web-page as formatted plain-text.

In short, read this book if you want to be a better programmer. If you have read it in the distant past, read it once again - you might have missed some things the first time round.

"Economics: Private and Public Choice"

2010-01-10T00:00:00+05:30

If you are looking for a comprehensive and accessible introduction to economics, "Economics: Private and Public Choice" by James D. Gwartney, Richard L. Stroup, Russell S. Sobel and David Macpherson is the book for you. It covers both microeconomics and macroeconomics in addition to the core principles of economics. Though it is a textbook meant for an undergraduate course in economics, it is also suitable as a gentle introduction to the dismal science for the lay person. I read the tenth edition of this book that was published in 2003.

This is a "modern" textbook in every sense. It is quite colourful and is richly illustrated with several pictures, graphs and icons. The chapters are not too big and are written in a simple language. They begin with an overview of the chapter and end with a summary of the key points learned in the chapter, a sneak preview of what is in store in the next chapter and several critical analysis questions for testing your comprehension (only some of which are answered in an appendix). Several boxes sprinkled throughout the text either talk about applications of economics or introduce a noted economist. Teaching aids for the book like presentation slides, instructor's manual, etc. can be separately obtained from the publisher. The book comes with a ridiculous CD-ROM containing a single HTML file that in turn redirects you to the "Gwartney Xtra!" web-site containing supplementary material for the book. I cannot help but wonder why the publishers could not just provide the URL for the web-site in the book.

The book is divided into six parts, each containing closely-related chapters that progressively expand upon the respective subject of the part. These parts are:

The Economic Way of Thinking
Markets and Governments
Core Macroeconomics
International Economics
Core Microeconomics
Beyond Basics

I was surprised to see macroeconomics presented before microeconomics - I would have expected it to be the other way round. The last part is a collection of 14 special topics on subjects like the burden of social-security, rising healthcare costs, increasing national debt, etc. that still seem quite relevant. The focus of the book is on the United States, but international readers will not feel lost while reading the book as the core issues that are discussed remain universally relevant. The thrust of the book might be termed right-wing or republican as it argues for a reduction in taxes, smaller role of the government, etc.

Unlike some other textbooks on economics, this book is not math-heavy. Like some other textbooks however, there is a lot of hand-waving and the reader is expected to just believe the authors. For example, at least in the beginning the authors could have shown how supply curves are derived from empirical data and how they shift under changed conditions. This book also seems to be revised every couple of years like other college-level textbooks - this is a very expensive textbook and these superfluous revisions can be seen as an underhanded attempt by the publisher to kill the used-books market.

The book presents a strong case for capitalism, but does not explore why communism or socialism became more popular in other countries. It does provide excellent insights into why the democratic process does not quite work the way people expect it to work. It shows how people are forced to choose among candidates who might not represent their stand on all the relevant issues, only some. It explores the rent-seeking behaviour of special-interest groups who use campaign financing and intensive lobbying to get politicians to pass laws that strongly benefit them but weakly harm the general public, sometimes in not-so-obvious ways. It explains how politicians have a strong incentive to take decisions that have short-term benefits but long-term hazards.

I felt that the book did not cover the very basics of economics and finance in much detail. For example, the origin of money or the evolution of the banking system could have been expanded upon. While the book presents almost all the fundamental concepts, it does not put it all together to illustrate how the student can perform economic analysis on their own in a given situation. For example, how can you arrive at the value of a house or a stock and decide whether an offered price is cheap or dear? Concepts like opportunity cost, interest rate, present value of money, inflation, etc. could have been used in the analysis for such a case.

It is my firm belief that everyone should be familiar with the basics of economics since it is so important in our times. If you have never been introduced to the subject before, this is a good textbook to get started provided you can afford the price.

"The Story of Art"

2009-11-08T00:00:00+05:30

"The Story of Art" by E. H. Gombrich is a classic and popular book that charts the evolution of art through the ages. Using a plain language and written in a non-condescending style (sadly quite rare for the art books that I have seen in my life), it is a book that I wish I had read much earlier in my life. It would make a great gift for a teenager to gently initiate them into the wonderful world of art.

The book mostly covers paintings, sculptures and architecture in its coverage of art. The period covered begins from pre-historic times and ends in modern times. As the author himself acknowledges, the book almost exclusively deals with western art in order to keep its size manageable. Within these limits, it does a splendid job of noting the contexts in which art developed, which provides valuable insights into the development of particular styles. You cannot fully appreciate a work of art without understanding the context in which it was created.

The book has more than 400 illustrations, works of art reproduced in full colour, that aid the text. In the main editions, these are put right next to the relevant text for easy reference; in the pocket edition, these are collected in a set of plates towards the end of the book in order to reduce its bulk. The pocket edition has two ribbon bookmarks so that you can keep track of your place in the text as well as in the set of plates. Flipping back and forth between the two as you read the text becomes painful after a while, but that's the price you pay for a less bulkier (and cheaper) edition of the book. Unless you want to read this book while travelling, I would suggest that you get one of the main editions (which are much heavier and, of course, more expensive than the pocket edition).

I fully agree with the author that the illustrations, all of them scaled down reproductions, do not do full justice to the respective works of art. I have had the privilege of viewing many of these works in person and no photograph or video properly captures the colours or scale or texture or the brush-strokes (in case of a painting) of a work of art. For example, something like a painting on the ceiling of a chapel is best appreciated in situ and in the normal lighting for the place. Of course, this involves a fair bit of travel, which is not possible for all of us. As the author notes, this is one of the biggest challenges for a lover of art.

While the author keeps referring to various elements of the composition of different works of art, he does not fully elucidate either the elements or the principles of art. You get a general idea of these things, but you'll have to look elsewhere for a fuller exposition. This is understandable in a book on the history of art. It is also understandable when you consider that these are not considered sacrosanct in modern art.

If you read this book looking to "understand" abstract art, you will be disappointed; I certainly was. The author uses a somewhat vague language to describe such art and very carefully tries to not pass a judgement on such art. His defence is somewhat credible in that enough time has not passed for us to judge "modern art" and to sort its place out in the grand history of art.

There is an essay in the final chapter of the book where the author ponders over the current state of art. He highlights the shift away from representationalism possibly as a reaction to the fidelity with which a camera captures a picture in modern times. He notes the widespread non-conformist attitudes of modern artists and a "rebellion against the rebellion" of some of the modern artists in going back to representationalism.

This is the sort of book that every person ought to have on their bookshelves. I would certainly recommend that you read such a book before you visit art museums in order to appreciate the context in which art was created. After you read this book, make it a point to visit the art museums in a city you travel to. It is quite an enriching experience.

"Paris"

2009-11-05T00:00:00+05:30

"Paris" by Colin Jones is a history of the city of Paris, covering the period of about 2,000 years from its days as the Roman camp of Lutetia to the present. The author chronicles its rise to prominence as one of the greatest cities in the world and a major centre for art and fashion. He does not shy away from talking about the horrible attrocities of its past either. This is a book written by someone who clearly loves the city for what it is and what it has been.

The coverage of the 2,000-year period is not uniform, with the author understandably concentrating on some intervals longer than others depending on their influence on the city. There are numerous "feature boxes" spread throughout the chapters of the book that go into more details of a particular topic, ranging from the Louvre to the public toilet. Unlike such boxes in other books, your flow is not interrupted if you start reading them after the text immediately preceding them and the topic in question is usually something alluded to in the last paragraph you were reading in the main text.

Though the overall narrative is fairly linear, there are occasional references to topics that are yet to be covered but that are still relevant to explaining the context of the current topic. This turned out to be not as bad as I had initially feared. The book manages to remain quite entertaining as well as informative throughout the text. This is a well-researched book written in an easy-to-read style. I am sure that even die-hard Paris fans will discover something new about the city if they were to read this book.

You must have stayed in Paris for at least a little while to make sense of some of the things (e.g. street names) that the author talks about in this book. French is thankfully not used extensively in this book - the meanings of the French phrases that are used can easily be guessed or looked up in a dictionary. You should also have a fair knowledge of world history in general and French history in particular to place some of the events and characters in mentioned this book in their proper historical context.

The book has a fair collection of paintings and photographs spread throughout, though all of them were reproduced in low-quality black-and-white prints in the hardcover edition that I read, which was a big disappointment to say the least.

"Buyology"

2009-11-02T00:00:00+05:30

"Buyology" by Martin Lindstrom is a book that purports to show that our subconscious drives our buying decisions in ways that we rarely suspect. Marketers can successfully sell products to their target consumers by understanding these factors; otherwise their campaigns are a waste of time, effort and money. The author tries to back these claims by citing the results of some studies.

The studies in question involved over 2,000 volunteers and used brain-mapping with Functional Magnetic Resonance Imaging (fMRI) to study their reactions to products and marketing messages. By noting the parts of a volunteer's brain that "light up" when presented with a test, the researchers draw various conclusions about how the product or marketing message affects their subconscious.

This is where the studies seem to be flawed, as far as I can tell. It seems to be a case of the classic logical fallacy usually known as affirming the consequent - "A implies B" does not necessarily mean that "B implies A". In other words, suppose a part of my brain lights up under fMRI when, say, I'm hungry. If you now observe that the same part of my brain lights up when I see a product, it does not necessarily follow that the product makes me feel hungry. Maybe it does; maybe it doesn't.

I feel that there is not enough material in this book that justifies its existence. It should perhaps have been published as an article in a magazine instead. There is a fair amount of fluff in this book seemingly to fill out the pages. In addition the author comes across as quite egotistical dedicating several passages to tell you (again and again) that he surely has influenced your life in one way or the other and that he is constantly travelling all over the world meeting his big and important clients. It seems as if this book did not have an editor.

"101 Essential Tips: Wine"

2009-09-12T00:00:00+05:30

"101 Essential Tips: Wine" by Tom Stevenson is an introductory little book for those interested in wine. It is a copiously-illustrated book with easy-to-read text and is so small that it can be finished in just a single sitting. It serves its purpose fairly well, though the title is a little misleading.

This book is supposed to be a collection of 101 tips, though most of the text can hardly be described as "tips" - only part of the text, for example the bits dealing with storing wine or what to look for while purchasing wine, can really be thought of as containing tips. The division into 101 tips looks entirely arbitrary and shoe-horned to fit the title of this series of books from the publisher.

The book is not entirely suitable for absolute beginners to wine in that it does not tell you just what is meant by terms like "body", "tannin", etc. This is important for people who have grown up in cultures (e.g. India) where wine consumption is not prevalant. There are also some terms like "terroir" and "sommelier" that are surprisingly missing from a text of this nature.

The book would have benefited from a wine vintage chart (something like this one by the Wine Society of India) to help beginners select a good wine while shopping. It would have also benefited from recommendations for selecting wine that can pair well with Indian or Chinese food. Those of us for whom the correct pronunciation of French or Italian words doesn't come naturally would have appreciated a pronunciation guide for some of the terms.

These quibbles aside, I liked the book for its welcoming, unpretentious and non-condescending text. Numerous colourful photographs and figures alongside the text ease the reading further (though I felt that the colour reproduction was not entirely faithful to the respective subjects) and mostly help in visualisation. Some of the tips are really useful if you are not a connoisseur of wine.

A book like this can only help you get started with appreciating wine. You need to sample a variety of different wines to really appreciate what such books talk about. This book however is cheap and accessible enough to merit a dekko.

"An Introduction to Database Systems"

2009-06-05T00:00:00+05:30

"An Introduction to Database Systems" by Christopher J. Date is a classic text-book on database concepts by one of the foremost writers and thinkers in the field. The book has been kept fairly up-to-date over the years by the author and includes topics like logic databases, temporal databases, decision support systems, object databases, object-relational databases, etc. Most of the book however concentrates on the relational data model and relational database management systems. This is not surprising since most of the popular database management systems today are based on this model and the author too clearly prefers this data model over all the others.

Almost all of the text in this book is a wistful reflection on the pure relational data model and its imperfect implementation in real database management systems. This is actually quite good since it is quite important for a serious student to know how things ought to be in addition to how things actually are. The book teaches you how to design good databases in general; it does not teach you the specifics of a particular product like Oracle or DB2. Beyond a preliminary overview of the general concepts, it will also not teach you writing complicated SQL queries, query optimisation for a particular product or writing stored procedures in PL/SQL. This book is what you need to get a solid understanding of the concepts underlying modern database managements systems.

The book is divided into a few logical parts, each part comprising well-defined chapters. Each chapter has an introductory overview, detailed exposition, summary of concepts, self-check exercises, well-anotated bibliography and answers to selected exercises. There are chapters on all the important concepts like the relational data model, relational algebra and calculus, database integrity, SQL, database normalisation, query optimisation, transactions, concurrency, recovery, security, etc.

The book provides most of the examples in a weird language called "Tutorial D" instead of SQL. It does not tell you much about the data models and database systems prevalent before the relational data model, depriving the student of a useful perspective on the historical significance of this model. The book does not provide any information on how relational databases could be implemented dismissing it as a detail that varies from product to product. This is quite unfortunate since it is very instructive to see how an abstract data model is concretely implemented in an efficient manner.

"Computer Networks"

2009-05-17T00:00:00+05:30

"Computer Networks" by Andrew S. Tanenbaum is a comprehensive text-book on the topic written by an author with many popular text-books to his credit. There are more than 800 pages in this book, though the prose is fairly accessible and peppered with the entertaining wit and sarcasm that is the hallmark of the author's text-books. Even though its 2002 vintage shows through every now and then, the book on the whole is fairly up-to-date and still quite relevant. It is therefore a great text-book from which to teach yourself about computer networks.

While the book talks about the seven-layered OSI model for computer networks as well as the much simpler TCP/IP model, it uses a hybrid model to describe the relevant concepts and protocols. This model comprises the Physical Layer, the Data-Link Layer, the Network Layer, the Transport Layer and the Application Layer. I personally feel that the model used in this book is a more natural fit to computer networks as they exist today.

There are enough protocols described here, across all layers, to sate the appetite of almost every person. At times the author describes the politics behind the promotion of some of the protocols and the reasons for some protocols becoming more dominant than others. You also get to appreciate just how much effort has gone into these protocols to cope with all sorts of network errors and the evolving needs of users. The book covers a lot of wireless and mobile networking issues as well in this edition to reflect their increasing importance in the modern world.

The sections on multimedia, cryptography and copyright issues, while no doubt very interesting, seem a little out of place for this book. Some of the figures do not appear anywhere close to their sites of reference, which makes it a bit cumbersome to flip back and forth between the relevant text and the respective figure. The book uses a proportional font for code listings instead of the usual monospace fonts, which makes it difficult to read through the code.

While this book is a great reference for the concepts related to computer networks, it is not so great as an actual network programming reference. You would be much better served for that purpose by one of the books by W. Richard Stevens.

"Operating System Concepts"

2009-04-19T00:00:00+05:30

"Operating System Concepts" by Abraham Silberschatz, Peter Galvin and Greg Gagne is a text-book that teaches the basics of operating systems design. As is implied by the title, the book focuses more on the concepts than their implementation. Almost all the concepts that you need to know to appreciate the working of an operating system are explained here and illustrated with numerous case studies.

The book covers all the operating systems concepts that you find in a standard text-book like process management, memory management, concurrency control, device management, etc. Unlike some of the other text-books, it also covers security, networking, distributed systems, etc. This makes the book slightly bulky, though it also makes it a comprehensive reference on the topic for beginners. A couple of chapters towards the end bring it all together by taking a detailed look at Linux and Windows XP. The book's web-site provides additional appendices on BSD Unix, Mach and Windows 2000.

The biggest criticism I have for this book as a computer programmer is that it only covers the concepts, not their implementation. While it is true that most of us will never ever implement an operating system, it is my experience that only by looking at and tinkering with working code do we programmers gain a solid understanding of a system. When you look at the code you realise that it does not take black magic to write an operating system. The tinkering of fearless programmers with such code creates a breeding ground for fresh breakthroughs in this field - after all, Linux is the result of such tinkering by a very talented hacker. This focus on implementation is what makes "Operating Systems Design and Implementation" by Andrew Tanenbaum such a valuable book when compared to this book.

This book has gone through several editions over the years, gaining a couple of co-authors and losing one. It refrains from passing judgement over design alternatives. The language is simple and non-condescending. Some of the topics should perhaps have been dropped and some moved to an "advanced" text-book. For example, the chapters on networking, transaction processing and distributed systems really belong to other text-books devoted to the respective subjects.

If you are looking for a refresher text-book on operating systems or if you are just casually interested in operating systems design, this is a good book to have. If you are more serious, you will need to supplement this book with reading code from a real operating system like Linux or FreeBSD.

"Computer Architecture"

2009-03-30T00:00:00+05:30

"Computer Architecture" by John Hennessy and David Patterson is an advanced text on the subject that is a sequel to their other book "Computer Organization and Design". The authors assume in this book that the reader is already familiar with the topics introduced in the other book and build upon these topics. Even then there is a fair amount of overlap among these books that is a bit irritating if you have only recently read the other book.

The book covers the exploitation of instruction-level parallelism and thread-level parallelism in modern microprocessors in great detail. It goes on to cover memory cache design and organisation, virtualisation and secondary storage systems in detail. It shows how instruction set design can affect the implementation of pipelining and virtualisation. The authors illustrate the principles introduced in the book very well by using recent microprocessors as examples. This makes the book quite current and relevant, though it also makes the book look a bit dated after a few years.

As with the other book, a fair portion of the book is in its appendices most of which are on an accompanying CD. This makes the book less bulky and more cost-effective to produce but it also makes it cumbersome to refer to the off-book appendices. I can appreciate this compromise as the alternative would have been to make the book more bulky and omit a lot of the material from the appendices.

A lot of the material in this book, especially the case-studies and some of the appendices, has been contributed by other authors. Impressive as the primary authors' credentials are (one is the father of MIPS and the other of SPARC), they cannot be expected to be experts on every topic in this domain and this extensive collaboration brings others' expertise to bear on specialised topics like floating-point arithmetic, embedded systems, etc.

Throughout the two books the authors have stuck with their "quantitative approach" to this field. Instead of making hand-waving statements like many others, they provide concrete formulae to measure various aspects of a computer's performance and extensively use various benchmarks to test existing systems. These benchmarks include the SPEC and TPC benchmark suites to cover a range of performance aspects.

This book would make for a great text-book for an advanced course on computer architecture. It would be particularly useful to computer architects, operating system designers and compiler writers. For the rest of us, this book provides a great insight into just what goes in the making of modern high-performance computer systems that most of us just take for granted.

"Computer Organization and Design"

2009-03-20T00:00:00+05:30

"Computer Organization and Design" by David Patterson and John Hennessy is an introductory text on this subject for budding computer scientists and computer engineers. The field has been in a state of constant flux in recent times and this book serves as an excellent reference even for professionals, since its authors have kept it up-to-date with all the latest developments. The downside is that such a book becomes dated pretty quickly by its very nature, assuming the field continues its breakneck pace of development even into the near future.

The authors are the designers of two early and successful RISC microprocessor architectures (MIPS and SPARC) and this expertise is demonstrated very well in the book. They do not shy from providing details that other authors might have omitted in an introductory text. The flip side is that the book spends a lot of time on the architecture of a microprocessor and very little on the other aspects of computer architecture and design.

The book primarily uses the MIPS architecture to illustrate the concepts. It also uses the x86 architecture in places to illustrate more complex implementations of the basic ideas. The MIPS instruction set and architecture turns out to be fairly regular and minimal, making it a good example from the real world for teaching, unlike the x86 instruction set and architecture with its overwhelming complexity and idiosyncracies.

The book describes the basics of the MIPS instruction set and then shows how programmes in a high-level language are translated into this instruction set. It describes how instructions are decoded and executed, how arithmetic is implemented, how pipelining works, how caches work, etc. With each chapter, you explore the relevant parts of the MIPS instruction set and architecture. Throughout the book, the authors maintain their "quantitative approach" to this field by showing how to measure various things rather than just describing the concepts.

Quite a bit of the book's material has been moved to a companion CD making it a bit cumbersome to refer to that material, especially since the book makes references to this material throughout. This also means that the book is not as bulky as it would have been without the CD and that the authors have been able to provide more information on the CD than would have otherwise been possible in a regular book. I would have personally preferred the "Computers in the Real World" sections after each chapter on the CD rather than in the book since they are so incongruent to the material described in the book.

The book stands quite tall compared to the other books on this subject and its current popularity is entirely justified. I wish this book was available when I was studying computer science. If you are a practising computer programmer who graduated quite a while ago, you must get this book and go through it to keep up with the advances in modern microprocessors.

"Data Structures and Algorithms"

2009-03-15T00:00:00+05:30

"Data Structures and Algorithms" by Alfred Aho, Jeffrey Ullman and John Hopcroft is a relatively-short introductory text for data structures and algorithms useful for computer programming. It covers a surprisingly-broad set of topics considering its size. It is fairly dated now but it still has all the essential data structures, algorithms and algorithm analysis techniques.

Early on in the book the authors describe algorithm analysis techniques and the effect of badly-designed algorithms on performance as the size of the input data increases. They also introduce the concept of an Abstract Data Type (ADT) and use it to describe various data structures. The algorithms are described using a pseudo-language that is a bit too similar to Pascal - programmers these days will find it a bit weird and distracting, though it is still close enough to the currently popular programming languages like Java or C++ for easy translation.

If you are an absolute novice, this might not be a good book with which to start the study of data structures and algorithms. It does not clearly explain why you need to study so many data structures and under which constraints you would prefer one over the other. It also leaves many of the details of the algorithms to be worked out by the reader. There are no solutions provided for the exercises given in the book. Many a time the book uses unnecessarily formal descriptions when a much simpler language would have sufficed.

Some of the "advanced" data structures like red-black trees, AVL trees, splay trees, etc. are either not described here or are just alluded to without any details. An important and basic algorithm like binary search through sorted data is only mentioned in passing. There is not much information on either parallel or cache-friendly algorithms, both of which are important on modern systems.

If you are already familiar with the basic data structures and algorithms and do not want to carry around some of the other unwieldy tomes on this subject, this is a good book to have around as a refresher. If you need a comprehensive up-to-date treatise on the subject, this is not quite the book.

"Irrational Exuberance"

2009-02-19T00:00:00+05:30

Robert Shiller (of the Case-Shiller housing index fame) is one of the few level-headed economists who have been able to recognise and point out market bubbles in the making and who have had the courage to stand by their analyses even in the face of ridicule. His book "Irrational Exuberance" became famous for calling out the stock-market bubble in the US when it was published in early 2000, just some time before the bubble burst. The second edition of the book has again been remarkable for pointing out the housing-market bubble in the US when it was published in 2005, though this time it took a little over a year since then for the bubble to burst.

The book starts off by providing a historical perspective of the stock market and the housing market in various cities. It then examines the structural, cultural and psychological factors that result in market bubbles, followed by some of rationalisations that people usually resort to when they find themselves in the midst of such bubbles. It ends with examining the ways in which influential people can prevent such bubbles from forming and hurting a lot of people when they burst.

The author has been able to strike a fine balance between academic rigour and prose interesting to the lay person. Almost all the assertions are backed by the relevant data, method of collection, references, etc. There are quite a few endnotes in the book for each chapter so that the serious reader can get more details. At the same time, an intelligent non-economist should also be able to follow and appreciate most of the arguments in this book. This is a rare accomplishment.

If you believe in the efficient market hypothesis or that stocks are always good investments for the long run or that real-estate is a sure way of significant capital appreciation, you must read this book and consider the arguments and data provided here. If you were already sceptical of one or more of these theories, you will obtain substantial ammunition here for your debates with people who believe otherwise. The author provides cogent arguments against each of these beliefs that have now become a part of conventional wisdom.

The book is somewhat biased towards examining the irrational expectations that people have from the stock market. The book talks about the housing market in detail only in the beginning and in the end. I would have preferred a more uniform treatment of the two.

"The New Cambridge History of India: Vijayanagara"

2009-01-06T00:00:00+05:30

After our recent trip to Hampi, Anusha and I became quite curious to know more about the history of the Vijayanagara empire. Our high-school text-books on history barely touched upon the rise and the fall of this great empire that ruled over almost all of southern India for about three centuries beginning in the 14^th century. We picked up Burton Stein's "The New Cambridge History of India: Vijayanagara" mainly because at about 150 pages it looked like a more manageable read than the other such books. It was also far more recent than the other books and therefore had a much better chance of incorporating the findings from recent research into this aspect of Indian history.

The book turned out to be a serious study of the history of the Vijayanagara empire as well as a commentary on the published literature on this topic. It seems therefore to be better suited for students of history than the general public and it does get dull at times. That does not mean that it does not have fascinating information for the patient lay reader. This is especially true for those in India since the history generally taught to us in our schools is quite lopsided.

The main focus of the author is on the economic, political and social factors affecting the rise and the fall of this empire. Reliable collection of taxes over such a vast and disparate region coupled with the pressure on the treasury by the need to maintain a full-time professional army made for tricky economics. Since this empire was ruled over by three separate dynasties coming into power via means such as usurpation, there is plenty of politics to go over. An empire like this that ruled over a diverse population speaking Kannada, Telugu, Tamil and Tulu spread over the entire Indian peninsula meant that local chieftains were both necessary as well as threatening for running the empire. Finally, the empire was constantly threatened by bordering kingdoms. It was therefore quite amazing that the empire survived and thrived for as long as it did. Quite remarkably, the book shows that the very factors that enabled the empire to form and sustain itself ultimately led to its downfall.

To fully appreciate the book, you should have some familiarity with the other main dynasties that ruled over parts of southern India like the Chalukyas, the Cholas and the Pandyas. There are extensive bibliographic references in this book for the serious student of history. Having visited Hampi is not a pre-requisite for reading this book, though it definitely helps in appreciating the few plates that are there. There is not much information about the flourishing of art in this empire as is evident from the ruins of Hampi - you will have to look elsewhere if you are interested in that aspect of the empire.