C++ Standard Template Library, The [ペーパーバック]

It's hard to pin down exactly why, but this book was not quite as pleasurable a read as its predecessors. One of the reasons might be the typography: The use of underlining for emphasis of the actual makes the standards sections of the book unpleasant to read. There might be an issue of the subject: For all its power, there is not all that much interesting algorithmic stuff going on in the STL. Lastly, it seems that C++ template code as such, no matter how brilliantly written and how useful to the library client, is rather unpleasant to read-a somewhat sobering insight to a C++ aficionado as myself. As a result, the code that *was* algorithmically interesting was quite hard to understand-I would not recommend this book to somebody trying to learn about red-black trees, for instance.

If you buy just one book about the STL, buy Josuttis' _The Standard C++ Library_. If you want additional in-depth insight into the workings of the STL, and are willing to invest the time it takes to study the code, buy this book. I certainly never regretted reading it, and I hope that Plauger will update his implementation of the rest of the C++ library to publish a standard compliant version of the iostream and string libraries sometime in the future.

But, if you're serious about extending STL, especially if you want to write new container and iterator classes, this is *the* book you need.

Personally, I make use of the information provided in this book to write a 3D container class and a couple of highly complex 3D iterators for an academic study demanding high speed and reliability. I have other STL books like Austern's Genetic Programming and the STL or, Musser's STL Tutorial and Reference Guide which are both extremely good references for *using* STL. But, those books did not help me even a little bit when I was trying to write a 3D iterator. I believe, The C++ Standard Template Library is the only book around that's really meant for serious STL developers.

So, if you ever want to develop a new container with a fair amount of new features or a new iterator with fancy tricks, buy this book. But, If you're only after using STL and/or developing new algorithms, stick with Austern's Genetic Programming or some other similar book...

The book is a reference to the STL, why things were done the way they were, and what the features of the various components are. The extensive sample code is a reference implementation of the STL itself, comprising a good deal of the text. I found few examples of actually using or extending the STL, these were left as exercises for the reader.

The book is a strong reference for language purists (You know, those people on the newsgroups who quote from standards documents) or for someone implementing the STL (why?). In my opinion, someone learning to effectively use the STL would be better served with another reference.