The Flaws in the Binary Computer

General Purpose Computers use statically bound imperative commands on binary data.  The software is defined by mutable binary data that is a shared medium to attack. Static procedural programs cannot detect mistakes in this mutable binary medium. When John von Neumann proposed this design he ignored the scientific alternative of his colleague at Princeton University. Alonzo Church defined his Lambda Calculus to scale Alan Turing's digital computer as independent atomic functions using a Universal Model of Computation found in Nature.  Turing's imperative instructions are encapsulated as dynamically bound mathematical symbols by the other half of the Church-Turing Thesis. The Lambda Calculus compartmentalizes mathematics as independent digital objects dynamically bound into a DNA structure by a sequence of mathematical expressions as found on a university chalkboard. 

Shared binary data is endlessly stretched to virtualize industrial society without the strength to survive attacks, mistakes, and design flaws that take place as software interacts in cyberspace. Over decades of unbelievable change, from an isolated batch processing to advanced global society, binary data proves unreliable and unsafe. When batch-processing mainframes became the computer on a chip, networking exposed society to globally inspired attacks. Without experts to follow best practices, the blue screen of death frustrated personal computer users and hours, days, and even weeks of hard work were lost. Next malware tricks emerged with remote hackers to pry of communities and service organizations. Private secrets held in networked systems that concentrate personal data in a large monolithic configuration are attacked for commercial gain. 

The shared binary computer assumptions made by von Neumann in WWII prove inadequate for society digitally interconnected through cyberspace. The age of global cyber society requires fail-safe computer science for a citizen's democracy to function. The requirements of paying passengers, the citizens must come first. Simple binary computers based on monolithic compilations do not meet this need. The machine, not only shares data, this data is unguarded and easily corrupted. Worse yet the monolithic compilations exchange binary data that can fraudulently attack the binary conventions that are different in separate compilations. Thus the rings of privileges that separate any two compilations can be confused and further attacked without detection. This is how a simple email can freeze a corporation or steal emails from a presidential campaign.  

These new requirements cannot be implemented in software. Unlike the 
Titanic, the engine room of cyberspace must remain dry. It must be data 
tight to remain media tight. Any leaks in the engine room lead to errors 
and consequential disaster. A monolithic physical computer lacks 
functional digital boundaries. When the engine floods everything else 
fails. Shared physical memory means binary data cannot be trusted. 
Monolithic software only survived behind locked doors. Everything 
changed beyond expectations with personal computing and the internet. To 
sail and survive in a sea of malware and hackers, computers must evolve 
to support the requirements of global civilizations. The single hull of 
the pioneer is enhanced to support the functional boundaries of 
software. These functional boundaries are defined by the symbols on a 
teacher's chalkboard. 

Computers began with the Abacus and the slide rule. These functional 
machines are explained by the λ-calculus. The λ-calculus defines the 
rules of computational binding as found in nature. As Alonzo Church 
tutored his graduate student Alan Turing, a λ-calculus application as 
the atomic form of computation was expressed as a dynamically bound 
single-tape Turing machine. The binary architecture of General-Purpose 
Computer Science is not the same. Without the λ-calculus, von Neumann 
overstretched and distorted Turing's α-machine. These binary computers 
lack digital boundaries. Any mistake (a wrong instruction, some invalid 
data, or deliberate human sabotage) is immediately shared within the 
monolithic build. The engine room of the compilation is disabled for 
normal operations and the application takes a big hit. 

The binary computer is outdated. The pioneering assumptions for batch 
processing do not satisfy the needs of individuals, societies, and 
nations living as the paying passengers in cyberspace. The engine room 
of the cyberspace must detect infections to prevent attacks and operate 
safely. Neither fail-safe software nor obscure best practices mattered 
when computers were guarded in experts behind locked doors in top-secret 
computer rooms. Attacks were prevented by the 'air-gap' between 
individually locked rooms. Software problems were hidden by patches to 
home-grown programs. Infections could not spread beyond the locked doors 
of a physical room if the staff is trusted. 

Everything changed with the computer on a chip, quickly followed by the 
Internet. Cost collapsed and personal computing took off. Programming 
paradigms change when off-the-shelf and cloud-based software replaced 
bespoke, home-grown software on private machines. Despite significant 
hardware savings, software costs only grow because the software compiled 
for a binary computer is inflexible, wedded to the past by a statically 
bound memory space. Complexity grows as the network expands and software 
fails because the detains are monolithically exposed. Binary computers 
expose the compiled software to attack. 

The functional modularity of mathematical symbols is vital for 
mathematics to be understood to work reliably. As taught in school, 
children can perform mathematics without the mistakes that take place in 
a General-Purpose Computer. The Abacus and the slide rule are machines 
that follow the same rules learned at school. Fail-safe computers, 
including children, the Abacus, and a slide rule are all functional 
machines. While unknown at that earlier time, both the Abacus and the 
slide rule use the dynamic binding rules of the λ-calculus. They apply 
the rules of the Church-Turing Thesis in form as well as function. 

The reason binary computers fail, and monolithic software is so hard to 
debug is because the λ-calculus is missing from the General-Purpose 
Computer as a functional machine. It is the binary details of 
General-Purpose Computer Science that interfere with and expose 
monolithic software compilations to attack. 

The binary computer is shared without individual security. One mistake 
reverberates throughout the statically compiled, monolithic image and 
then infects other builds both local and remote. Worse still the system 
is beset by rapids that require data portage between each image because 
functional boundaries do not exist in von Neumann's binary computer.

Comments