We have been studying obsolescence and security
problems over a 20 year period. We discovered that there must be some
fundamental principles to follow in information system architectures and
designs. The ten such principles are: (1) Inherent security,
(2) Obsolescence-resilient systems, (3) Application and owner centric focus,
(4) Bare machine computing (BMC),(5) Object-oriented (OO) methodology,
(6) Green computing, (7) Upward compatibility, (8) No middleware, (9) No
layering and (10) No heterogeneity. Some of these principles may be controversial to many
people, but they will considerably minimize obsolescence and maximize
security in information systems. In order to demonstrate a system
adopting some of the above principles, we have constructed numerous
complex applications based on bare machine computing (BMC) paradigm.
Our solution to the current computing problem is to simply eliminate all
middleware and make the hardware bare thus resulting in the BMC paradigm.
The BMC approach makes all computing devices bare by not including any valuable
resources (persistent storage and software) in the machine.
Thus, there is nothing to protect or monitor any valuable resources
other than vandalism. A computing box simply consists of CPU, Memory and
I/O as intended by von Neumann architecture. A computing device could be
a PC, Laptop, Large Computer, IPAD, Smartphone, or any other network
appliance. A bare machine will be activated using a detachable device
that boots and loads an application to run. When a given application or
a set of applications running in the machine, there are no other
applications running in the machine. The set of applications running at
a time are self-controlled, self-managed and self-executable. All
persistent data is stored on the external mass storage or on the network
which is secured and managed by the owners. When an application suite is
running, it will only provide functionality that is needed for that
suite, nothing more or nothing less. There are no open doors for the
intruders to entertain. This is not an open system, nor a universal
system to do universal things. It is application centric and user
controlled approach. When BMC becomes viable, then a given application
may run on any bare device which has no ownership. That is, one can use
any Laptop any where that is available. In such world, one can use any
one's cell phone by simply plugging in their USB flash drive. A user
carries his/her detachable mass storage device with their own
applications and data. This approach will change the computing as we
know today in every aspect of information technology. It will provide
enormous benefits in reducing obsolescence, reducing number of devices,
reducing computer dump, reducing rapid unnecessary releases, and
improving productivity in information technology. In additon, current
global Internet can be used to host "Bare Internet" applications, where bare
machine users communicate within their own domain of users as a an
"Intra-net." Current Internet entertains all users without
distinguishing between a bad and a good user. When information systems
architecture and design is based on bare machine computing, it will
result in systems based on software/hardware foundations, instead of products
based on short term profits and continued revenues. The solutions
proposed here span across many information system facets. It is a
daunting task to connvice industry, researchers and government agencies
to pursue such non-evolutionary approch. However, we pursue our
mission to demonstrate feasibility of information systems based on bare
machine computing paradigm.
Potential Applications
The solutions provided above can be
used in any computer application or information system. The bare machine
computing paradigm must be the backbone of information systems. The solution
provided here is not evolutionary, thus, for now, this approch can be used for some
"Intra-net" users using "Bare Internet". This will motivate industry,
research, government agencies to pursue this path to address current obsolescence and security problems.