• The contents of the Windows and UNIX versions of this archive are essentially identical.
• One difference is the use of a .tar.gz file for UNIX and a .ZIP for Windows.
• Another is that I have tried to use the "native" end-of-line character for each platform.
• There are code fragments here that are OS-specific, but they are all reproduced together.
• In case of errors in this archive, please contact the author, Hal Fulton, directly.

File naming conventions

• Chapters 2 through 9 are represented by the directories ch2 through ch9.
• The files here are (for the most part) named according to this format pNNN where NNN is the page number in the text. Example: p81.rb
• Note:In a directory listing, the files will not necessarily appear in numerical order.
• In some cases there may be a letter appended (when there is more than one code fragment on a page). Example: p124a.rb
• When the code fragment is an actual numbered listing, the listing number will be appended after a lowercase "l" (with the dot converted to a hyphen). Examples: p92l2-6.rb and p124bl2-9
• Many of the filenames are NOT "8 dot 3" as Windows users might expect. This should not cause problems even for Windows users.
• In some cases, it made sense to preserve the "original" name of the file. This is done, for example, when the filename is referenced from another related fragment. In this case, the "canonical" file (e.g., p442.rb will usually just be a placeholder that names the real file (e.g., params.rb).
• Some of the files are not Ruby code and thus may have other extensions such as .xml or .html (embedded Ruby files are given the extension .erb).

• Every numbered code listing is here (2.1 through 9.20).
• Many smaller fragments are also included (more than two hundred of them).

• There is no code from Chapter 1 or the appendices. These fragments are in general less significant than most of the others.
• The shorter, less important fragments were omitted.
• Listing 9.13 was omitted, as it is not actually code.
• Code that couldn't possibly run (e.g., with embedded pseudocode or with nonexistent resources) was usually omitted.
• In some cases, adjacent fragments may have been lumped together into one single fragment (if they were short enough and closely related).

• All the code has been tested, but there may still be bugs or errors that crept in during publishing.
• The code here is (supposedly) consistent with the initial version of the errata for the book, and thus may differ slightly from the first printing.
• Certain pieces of code will run only on the Windows family of operating systems or only on the UNIX family.
• Some code fragments have dependencies on prior code fragments. The content of the file usually does not reflect this, but the book's text will. (Some of the binary tree examples in Chapter 3 are especially problematic.)
• Some of the examples depend on "imaginary" resources such as files, directories, URLs, IP names, user names, and so on. These will typically not work without minor changes.
• There may be minor variations in the behavior of different versions of Ruby (and even different builds, e.g., Cygwin versus MinGW).