Data processing: software development – installation – and managem – Software program development tool – Translation of code
Reexamination Certificate
1998-09-18
2001-08-28
Powell, Mark R. (Department: 2122)
Data processing: software development, installation, and managem
Software program development tool
Translation of code
C709S201000
Reexamination Certificate
active
06282697
ABSTRACT:
TECHNICAL FIELD
This invention generally relates to computer processing and programming methods, and more particularly, to methods that employ a new type of autonomous data handlers.
BACKGROUND OF INVENTION
Computer processing has been developed from its earliest theoretical foundations to the present as a machine-oriented paradigm in which machine instructions are executed by a central processing unit (CPU) to convert raw data and input in sequential steps to a useful output. To accomplish more and more complex tasks involving greater and greater amounts of data, the development of computer hardware has proceeded with exponential increases in speed, power, and capacity, now approaching the physical limits of the atomic scale and the speed and wavelength of light. The programming of complex tasks by linear programming methods increasingly requires intensive code writing which can take up millions of lines of code and is susceptible to errors and cross-code conflicts which can require intensive debugging efforts. It also inherently requires that the programming be accomplished in one chosen programming language for a selected operating system, leading to obsolescence and legacy systems.
Object oriented programming (OOP) methods were developed to break up large programming tasks in modules that are self-contained and self-executing. Modules that have been optimized for specific functions can thus be catalogued and re-used in modular form in a variety of larger programming tasks. However, OOP methods require the enforcement of class definition hierarchies over a programming environment as a whole, resulting in closed systems and libraries of OOP modules that must be recompiled to support other environments.
Parallel processing methods were developed over the past decade in the attempt to obtain increased capability by breaking up a processing task into parts executed by an array of CPUs in tandem. However, conventional approaches to parallel processing have required centralized management and control of programmed tasks which imposes stringent requirements in the coordination and synchronization of tasks. Parallel processing methods to date have also continued to employ linear programming methods which are intensive and susceptible to fault, especially when operating in parallel, and are captive to the chosen programming language and operating system.
In recent years, with the rise of large-scale distributed networks, there have been proposals to perform very large or complex tasks by breaking them up into small pieces and distributing them among cooperating computers on a network, often referred to as “distributed computing” or “network computing”. The proposals for distributed computing represent a shift in the computing paradigm from the processing of tasks of high computational intensity towards one in which data are widely distributed among resources on a network and processed locally with lower-intensity CPUs through distributed efforts. However, since the same linear programming methods and centralized management and control have been used, the distributed computing paradigm to date has also required stringent programming and coordination of efforts to achieve worthwhile results. The failure of one or more individual processing efforts or time delays associated with communicating or coordinating with centralized processing management can produce bottlenecks which bog down the effectiveness of this approach.
One method which has utilized multiple asynchronous computer processes is called “Flow Based Programming” (“FBP”), as explained in greater detail in
Flow-Based Programming: A New Approach To Application Development,
by J. Paul Morrison, published by Von Nostrand Reinhold, New York, 1994. The reference describes Flow Based Programming as being based on three principles, “asynchronous processes, data packets with a lifetime of their own, and external definition of connections”. In FBP applications are defined as networks of “black box” processes, which exchange data across predefined conditions. FBP is a process of stepping tagged data packets through networks of “black box” processes and thus is a data-centric process. The FBP approach needs extreme predictive capability to work well.
Therefore, it is a primary objective of the present invention to provide a computer processing and related programming method which decentralizes the management of computer processing tasks and allows distributed or parallel processing of data to proceed autonomously with adaptive options such as taking the most efficient path available.
It is a further object of the invention to provide a new programming method which is optimally designed to facilitate parallel or distributed data processing by avoiding the need for captive programming environments such that simultaneous use of multiple programming languages can be supported and centralized processing management is eliminated. It is also intended to reduce the intensity and case specific nature of programming required for complex processing tasks and the susceptibility of such programming to fault.
SUMMARY OF INVENTION
In accordance with the present invention, a computer processing and related programming method calls for creating a plurality of software entities (referred to herein as “molecules”) which can be dynamically configured to process data autonomously. The molecules can send and receive signals indicating the state of a molecule, whether or not a processing task is fulfilled, the results of a processing task, and whether or not the operation of a molecule is to be terminated, interrupted, reconfigured, or continued by creation of one or more “next” molecules. The molecules are created from a common set of software micro-components, which may be programmed in any programming languages to run in any operating system environments.
The molecules created for an overall processing task may reside with a single computing resource (CPU and database), however, they are ideally suited to be deployed with distributed computing resources of different types on a network or in parallel processing environments. Each molecule may be programmed in a programming language or languages suited to the computing resource where it is intended to operate. The molecules can function with a mixture of compiled or run-time compiled code, allowing them to run on a computing resource with only a portion of the process adapted to run on that computing resource. The molecules can signal each other and send and receive data using common network protocols. They operate to complete their respective data processing tasks autonomously until the overall data processing task is completed. An overall data processing task can be performed by creating molecules and structuring their operations and interactions in a “logic web” which can dynamically adapt to fulfill the task.
Adapted to a parallel processing environment, the molecules can be dynamically configured to run autonomously on parallel computing resources. A logic web is defined to allow efficient access of each molecule to a data object it is intended to process, and to any available processor of the parallel processor array as an interim CPU resource. The scheduling of processor access can be handled by a main controller program, or by a separate scheduler molecule or mini logic web. The molecules are programmed to signal their state, result, and termination, reconfiguration or generation of next molecules. The dispersed, autonomous operation of the molecules avoids the rigid scheduling structures and time delay conflicts in conventional parallel processing methods, in order to accomplish data processing tasks more efficiently and quickly.
As a software-configured entity, a molecule is comprised of the following basic software micro-components: a signal handler, an input handler, an interface handler, a method handler, an associated method, and an output handler. The signal handler is capable of sending and receiving signals for invoking other molecules and signaling other molecules and resources on the ne
Fables Wylci
Park Jore
Chong Leighton K.
Nguyen-Ba Hoang-Vu Antony
Powell Mark R.
LandOfFree
Computer processing and programming method using autonomous... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Computer processing and programming method using autonomous..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Computer processing and programming method using autonomous... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2516133