I agree with some of the other comments on this thread. The STL can be hard to learn, the error messages generally stink (but that's really the compilers rather than the library), and it's easy to get yourself in trouble because of the complete lack of error checking. I also agree that people who only learn how to write C++ programs with the STL and don't know how to implement any of their own data structures or algorithms are going to be pretty useless when they get out into the real world.
All that said, however, I believe it has a very important role in allowing highly efficient programs without locking you into one particular data structure or algorithm. I just think it should be kept inside user-friendly classes rather than put in the main program, if you have a choice in the matter (and you almost always do).
Before I would teach someone the STL, though, I would make sure that he understood the basics of C++, like variables, classes, inheritance, and polymorphism. Then he would be able to use the STL and a productive way without being overly dependent on it. After all, you can't expect every algorithm for data structure you ever need to be provided as part of a standard library.
By the way, if anyone is looking for a good C++ programmer, you may want to look here.
I feel it is important for me to correct the misstatements of fact made by some disgruntled former students.
First, I was not "canned", fired, or otherwise dismissed from UT Dallas. To the contrary, I was asked to return the next semester to teach C++ again. I did not do so for two reasons: the meager pay, and the few lazy, whining students who did not feel like working for a grade.
Second, the majority of students in my beginning C++ class got an A. As I recall, all of the Vietnamese students got As. So did most of the other foreign students. The only group of students who did not do well on the whole were the American students. I'm at a loss to explain this on the basis of anything other than their attitude. Surely they did not have the language barrier that many of the foreign students did, and yet their average performance was much worse. There were exceptions, to be sure, but that was the overall outcome.
Obviously, the student or students posting negative comments here were among the few who endeavored to make my life as difficult as possible as a teacher. As for the rest of the posters, I can only suggest that they take a look at the free Web version of the book, which is available here, and make up their own minds about its worth.
First of all, let me thank everyone who has contributed to this thread. It's nice to see so much interest in my book.
Now, to clear up some misconceptions:
I'm sorry that my former student doesn't care for the book. However, many of my students from the same class found it extremely useful in learning the basics of the language. I guess that just proves that you can't please everyone.
I do not use toy examples to illustrate fundamental concepts, where realistic examples are possible. For example, I use an inventory control application to illustrate the basics of writing a class. I abhor the general approach of using zoo animals or equally irrelevant examples, when realistic ones are possible.
The reason that I do not cover inheritance and polymorphism in this book, to put it is simply as possible, is that they would not fit the deadline that I was under when writing the book. I remedied this problem as soon as I could by writing a sequel, which oddly enough is called "Who's Afraid of More C++?". That book continues exactly where this one leaves off, and covers some rather advanced concepts that I believe are necessary for C++ programmers to know if they are going to write good programs.
Unfortunately, both of these books are going out of print shortly, if they aren't already. Therefore, if you want to buy them, you should do so as soon as possible.
However, the good news is that I'm under contract to produce a book combining the two above-mentioned books into one volume. It should be out in the fall.
Slashdot Mirror
I agree with some of the other comments on this thread. The STL can be hard to learn, the error messages generally stink (but that's really the compilers rather than the library), and it's easy to get yourself in trouble because of the complete lack of error checking. I also agree that people who only learn how to write C++ programs with the STL and don't know how to implement any of their own data structures or algorithms are going to be pretty useless when they get out into the real world.
All that said, however, I believe it has a very important role in allowing highly efficient programs without locking you into one particular data structure or algorithm. I just think it should be kept inside user-friendly classes rather than put in the main program, if you have a choice in the matter (and you almost always do).
Before I would teach someone the STL, though, I would make sure that he understood the basics of C++, like variables, classes, inheritance, and polymorphism. Then he would be able to use the STL and a productive way without being overly dependent on it. After all, you can't expect every algorithm for data structure you ever need to be provided as part of a standard library.
By the way, if anyone is looking for a good C++ programmer, you may want to look here.
First, I was not "canned", fired, or otherwise dismissed from UT Dallas. To the contrary, I was asked to return the next semester to teach C++ again. I did not do so for two reasons: the meager pay, and the few lazy, whining students who did not feel like working for a grade.
Second, the majority of students in my beginning C++ class got an A. As I recall, all of the Vietnamese students got As. So did most of the other foreign students. The only group of students who did not do well on the whole were the American students. I'm at a loss to explain this on the basis of anything other than their attitude. Surely they did not have the language barrier that many of the foreign students did, and yet their average performance was much worse. There were exceptions, to be sure, but that was the overall outcome.
Obviously, the student or students posting negative comments here were among the few who endeavored to make my life as difficult as possible as a teacher. As for the rest of the posters, I can only suggest that they take a look at the free Web version of the book, which is available here, and make up their own minds about its worth.
Now, to clear up some misconceptions:
Unfortunately, both of these books are going out of print shortly, if they aren't already. Therefore, if you want to buy them, you should do so as soon as possible.
However, the good news is that I'm under contract to produce a book combining the two above-mentioned books into one volume. It should be out in the fall.