This book addresses the exchange of data between applications that are run on micro and mainframe computers. It gives a good, quick, management level overview of the issues involved and a narrow solution to one part of the problem: the data format conversion problem. It does not contain an in-depth study of the problem and is mainly of interest to those dealing with IBM PC compatible systems and IBM mainframes. For example, the author discusses the Bisynch and HDLC/SDLC protocols but ignores common PC file transfer standards such as the Xmodem and Kermit protocols.

The first half of the book, pages 1–159, contains the actual text. Pages 160–269 contain the user manual and source listings for the author’s data format conversion system--the Text Descriptor File/Descriptor File Processor system. The TDF/DFP system is a package of tools to allow the conversion of files from one format to a TDF standard file, and then to another format. This includes a utility for building a TDF description file for any ASCII-based data file. The author proposes a standard for text descriptor files but gives no indication that there is any movement toward making the TDF format an official standard.

This book is not acceptable as a textbook or a reference book. Even though it makes references to other books, IBM technical notes, and ISO standards, there is no bibliography or information on how to obtain the referenced documents. Although there is an index, it is incomplete. For example, on pages 98 and 99, there are tables describing the internal structure of DIF files and dBASE III files. The dBASE III table is in the index; the DIF table is not.

The numerous programs and tables that are included in the book could have been an asset, if not for some obvious mistakes in format. First, there are no comments in the program listings except copyright statements. Standard BASIC is not that readable to begin with, and, without comments, discovering what some of the code is doing is more trouble than it is worth. For example, the very first line of code in the first program (TDF.BAS) is “80 DIM A$(13),B$(13),F$(20),H$(30).” Don’t you enjoy the nice descriptive names? “H$” turns out to be an array used to contain messages. I did not bother to track down the others. The second problem with the file listings is the format. The 80-column listings are printed 64 characters wide, leading to a lot of wrapped lines, which further reduces the readability of the programs. Finally, there is the typical BASIC problem of following the flow of uncommented programs. Again, the first program includes “1240 LOCATE 23,29:LBL=23:LEB=28:LY=3:GOSUB 4200:F2$=RIGHT$(“000”+YZ$- ,3):LOCATE 5,11:PRINT F2$”

If you want a brief overview of the dataflow problem, this would be a good starting place. However, unless you want to take the time to type in the 69 pages of source listings, you may as well skip the rest of the book.