The Art of Assembly Language. [Englisch] [Taschenbuch]

Well, after four years of reading these reviews, I thought I'd put in my two cents.

One recurring theme you see in all of these reviews is the following: if someone already knows assembly language, they tend to dislike the use of HLA as the teaching vehicle for learning assembly language. On the other hand, if they're a newcomer to assembly language, they tend to like the approach that Art of Assembly uses. Quite frankly, I wrote "Art of Assembly Language" (AoA) for this latter category, not for those who already know assembly language, so I am rather gratified by the response from those who are actually using AoA to learn assembly language.

When someone sets down to write a book on x86 assembly language, one of the first decisions they have to make is "which assembly language syntax do I use?" The x86 is blessed/cursed with literally *dozens* of different assembly language syntaxes. No matter *what* assembly language syntax I chose, there would have been someone complaining about it. If I'd gone with GNU's as (gas), there would have been complaints about the syntax. Had I gone with FASM, the NASM crowd would have been put off.

Probably the "safe" choice would have been to go with MASM (which the earlier, 16-bit version of the book, used). No doubt, many of the complaints about how I used HLA instead of a different assembly language syntax would have gone away had I done this. The funny part is that MASM is *also* a high-level assembler, having almost all the same high-level control constructs found in HLA. The same is true, by the way, for Borland's Turbo Assembler (TASM). From a language feature point of view, there really isn't much difference between the high-level facilities of MASM, TASM, and HLA. Maybe it's just the name that freaks people out.

Some reviewers have commented that this is the wrong way to teach assembly language. Well, having taught assembly language at the University level for over 10 years, I must respectfully disagree. I've used HLA (before AoA was available) and the students did *far* better in the course. They got much farther along because they were able to apply their HLL programming knowledge to problems early in the course. By the time the course covered the low-level machine instructions, they were doing quite well. The courses I taught with HLA worked *much* better than the comparable courses I taught with MASM. The bottom line is that this teachnique technique has been classroom and laboratory tested. Interested individuals might want to check out my white paper on this subject:

I will make the following observation about AoA: if you already know assembly language, you're probably not going to like the presentation because it's completely different from the way *you* learned assembly and most people seem to think that the only way to learn something is the same way they learned it. On the other hand, if you don't know assembly language and you want to learn it, pay particular attention to those reviews from the people who used AoA to learn assembly language.
Cheers,
Randy Hyde

Some of the above reviews have claimed that this book does not teach "real" assembly language, and that it uses 'c'-like wrappers instead of pure assembly instructions. This is a misconception most likely caused by these reviewers lack of knowledge, and/or failure to read the book of which they have submitted a review.

First off, what is Assembly Language? It is an attempt to make the actual machine instructions more readable to us humans, back when I first learned assembly language on the 6502, I programmed using hexademical instructions, so for example, changing the background color on the good old C64 would be:

$a9,$00,$8d,$21,$d0

Now, this isn't exactly readable as far as code goes, so later I got hold of an assembler, and the above code was written as:

lda #$00
sta $d021

This was suddenly alot more readable, and generated exactly the same code. Onwards assemblers have evolved, including things like macros, local labels, etc. HLA is one such evolution, it contains for example alot of control structures to avoid the need of labels, but that does not mean that you have to use them. For readability, it's lot easier for you to make a function call as:

Foo(1,2,3);

But if you really want to, you can write the code yourself,
push 3;
push 2;
push 1;
call Foo;

Still, this is exactly the code that will be generated by the above Foo(1,2,3), so it's really just a matter of taste.

Likewise, the high-level constructs such as IF... THEN works just the same way:

if(eax == 1) then
endif

could be written by yourself as:
cmp eax, 1
jne Label

But again, this is the same code that the high-level construct will generate. There are most likely situations where high-level constructs may generate code that could be written slightly more efficiently by hand, but it's entirely up to the programmer to use them or not. For beginners in assembly they are likely a godsend, and for experienced programmers they are simply an option.

Now, valid criticism towards this book is that the focus on HLA, although helpful, may also confuse the beginners, since it detracts somewhat from the low-level fundamentals that is the basis of assembly programming. For instance, although excellently explained, the way the stack operates could easily drown in the information sea of HLA's STATIC, VAR, READONLY, STORAGE sections described in the chapter beforehand, and make it hard for a beginner to grasp.

That said, the book still covers all basics of assembly language, from system bus to the individual cpu instructions.
And if you actually read the book, rather than firing up the examples directly, you'll have a good grasp of what these high level constructs do, and how to write your own code without using these constructs if you so please. And do not believe the above reviews stating that this is C-programming rather than assembly, if your programs consist of nothing but function calls then yes it will look like a C-program, but if your program actually does something rather than calls, you will use mov, and, or, add, sub, inc, dec, mul, div, shl, etc. like in any other assembly program, and these instructions are explained perfectly within this book.

The reason I don't give this book 5 stars is simply that I feel the focus on HLA should be mentioned in the books title, like "the Art of Assembly Language using HLA", since people using other assemblers will have to wade through alot of HLA specific content of which they have very little, if any interest.

This is truly amazing piece of work on High Level Assembly (HLA). It's important to know what you are getting is a book on HLA because the back cover says, "The most comprehensive guide to assembly language". Which is both hyperbole and factually slightly inaccurate. It's a very, very good book on an assembly language (HLA), but not all assembly languages. Nor should I think there would be a good book that covered all assembly languages, but that's beside the point.

There is some general value in the book that applies to almost any processor. The basics of registers, operations, pointers, the stack and other basics. But as you get deeper into the book it's clear that this is a work on HLA and HLA alone.