The CORBA Reference Guide: Understanding the Common Object Request Broker Architecture - Softcover

Pope, Alan

Softcover

ISBN 10: 0201633868 ISBN 13: 9780201633863

Publisher: Addison-Wesley, 1998

Synopsis

The Common Object Request Broker Architecture (CORBA) is a pivotal technology that facilitates and promotes the development of distributed applications. CORBA has taken a stronghold in the computer industry, primarily because it is a standard architecture that provides common interfaces and descriptions for objects. The emergence of CORBA has been a blessing to the programming community-now there is a framework in which distributed object-oriented applications for many different systems can be more easily constructed. This book offers you the clear explanation of CORBA that you need and provides a complete reference to the standard. The CORBA Reference Guide provides a general background in distributed systems, a technology that is vital for building scalable distributed object systems. The book explains the base architecture as well as the services and facilities that extend this architecture. This explanation also provides historical context discussing why certain features were selected.

Of particular note, this book details the most sophisticated security framework developed for any architecture to date. The book also covers interoperability with other ORBs, especially between OLE and CORBA. Additionally, key details about the Object Data Management Group (ODMG) Standard are included.

Distributed systems managers will gain a better understanding of CORBA by quickly reading the first section of each chapter and browsing the informative margin notes. Software architects and project leaders will appreciate the in-depth explanations of the various interfaces and the rationale behind CORBA.

About this title

From the Inside Flap

The noblest function of an object is to be contemplated.
-- Miguel de Unamuno,
from Mist (1914)

The motivation to write this book came from a desire to help people understand CORBA and its basis. A great deal of misunderstanding arose in the first few years after CORBA appeared. In the time it took to write this book, however, many people became aware of and well informed about CORBA. The misstatements heard today are not nearly as prevalent as they once were. Still, I believe the need for a book such as this has not gone away.

Today people have a number of requests for features they want to see become part of CORBA. Occasionally these features have been given a great deal of consideration before being discarded or shelved. This book tries to record some of the thoughts behind the discarding of several ideas. It does not attempt to examine all of them because that would be an immense task. Rather, the most fundamental appear here.

This book is for system architects, designers, and programmers, specifically those working in information systems. It is not a programming guide. Instead, it provides a general background and reference to the architecture and various services. I suspect programmers will find this background material useful but not very meaty. The intent is to present CORBA with enough detail and background to help people fully understand and properly utilize its features.

This book started out as an explication for system managers and architects, but then several such books came onto the market. A more in-depth reference instead of a cursory explanation became more interesting. As more and more people began committing substantial resources to constructing systems using this technology, it became obvious that someone was appropriately informing management.

The Object Management Group (OMG) submission adoptions include details for both application builders and those implementing Object Request Brokers (ORBs). Frequently the bias is more toward ORB implementers. I attempt to remove most of this unless it is important for a good understanding. In some cases, I may have included too little or too much of this auxiliary information.

A separate book that gives good application examples is desirable. Documenting all aspects of the architecture and providing coding examples for each would be an extremely large and complex task. Either the examples or the descriptive content would suffer. In the future, it is more likely we will see books that either focus primarily on examples, or books that select some subset of the entire architecture, each concentrating on a different area of emphasis and focus.

A number of factors made this book hard to write. For one thing, the rate of adoption of proposals and their complexity is rapidly increasing. It is much harder to stay completely on top of every topic with so many different groups of people working in parallel. It also is impossible to capture in a single book every up-to-the-minute adoption because the production time leaves a window open for new technology that cannot be included. For the short term, anyway, this is a losing battle. The most current source of information is always at the OMG, within the submissions themselves.

A second factor, for me, was unexpectedly becoming a single parent to my two-year-old daughter. With a full-time job at Quantitative Data Systems, Inc., the book was already a second full-time job. Finding myself suddenly with a third full-time job, priorities being what they are, the book became the lowest priority of the three. Had I known what would occur over the last two years, I most likely would have had the sense not to sign up to write a book, which is immensely time-consuming.

I am responsible for all errors that appear. Although some very astute people in this field gave me very sound advice, I may have totally misinterpreted or entirely missed their points. The people I name are not accountable for any of the book's content. They made their best attempt at making sure I understood things clearly. I made my best attempt to listen and incorporate their advice.

In some sense, I had the pleasure of building two ORBs while I was in the employ of Sun Microsystems. The first ORB was built in eight weeks by about sixteen engineers. This is the original HP/Sun ORB that became CORBA with the incorporation of the DII (from the DEC and Hyperdesk submission). This incorporation was the first submission merger and its revision was to appear in 90 days. Construction of this second ORB took much longer and involved quite a few more people.

Specifically, from the original eight-week stretch, I want to thank: Sally Ahnger, Dave Brownell, Steve Byrne, Alain Demour, Jacob Levy, Joe Pallas, Michael Powell, Sami Shaio, and Bart Smaalders. Unfortunately, there are several dozen others from my ORB-building period as well as from the time I was at Sun but both space and memory conspire against me, so this list is artificially short.

A number of people managed to read through my rough manuscript, something I still find quite astonishing. I can only thank and account for those that I know: Dr. Daniel R. Edelson, Bret Hartman, Jishnu Mukerji, Dr. Richard Mark Soley, Mark G. Wales, and Andreas Vogel.

Some folks also have no idea that they were contributing to anything. They require thanks, though, because they did. The following list of people either said something to cause an epiphany or clarified some aspect that had not previously been so clear (although the new clarity sometimes was 180 degrees from the original intent). They are Rick Catell, Ralph Johnson, Douglas Lea, Jeff Michinsky, Alan Snyder, Drew Wade, and Jim Waldo.

Certainly not least are the staff of Addison Wesley Longman who patiently waited for me to exhibit signs of life and made extremely valuable suggestions. Among them, certainly John Wait, Mike Hendrickson, Marina Lang, Katie Duffy, and Sarah Weaver deserve several rounds of applause. Both Marilyn Rash and Ann Hall deserve special thanks for helping to get this book in useful form.

My current employer, Quantitative Data Systems, Inc., probably deserves some award, specifically Jerry Conrad and Robert Morse, because writing this book sometimes took priority over my real job.

In addition, for helping me survive I'd like to give a special thanks to Lisa Walker, Louise Allen, and Charlotte Einar. Without the help they gave me I would never have been able to finish this book.

Last but never least, although she had little or no choice in the matter, I would like to thank Ashley, who got short shrift and not nearly as much attention as she deserves but loves me anyway. 0201633868P04062001

From the Back Cover

The CORBA Reference Guide provides a general background in distributed systems, a technology that is vital for building scalable distributed object systems. The book explains the base architecture as well as the services and facilities that extend this architecture. This explanation also provides historical context discussing why certain features were selected. Of particular note, this book details the most sophisticated security framework developed for any architecture to date. The book also covers interoperability with other ORBs, especially between OLE and CORBA. Additionally, key details about the Object Data Management Group (ODMG) Standard are included.

Distributed systems managers will gain a better understanding of CORBA by quickly reading the first section of each chapter and browsing the informative margin notes. Software architects and project leaders will appreciate the in-depth explanations of the various interfaces and the rationale behind CORBA. 0201633868B04062001

Edition details

• Publisher: Addison-Wesley
• Publication date: 1998
• ISBN 10: 0201633868
• ISBN 13: 9780201633863
• Binding: Paperback
• Language: English
• Edition number: 1
• Number of pages: 407