This is a legally copyrighted work,
so none of it can,
legally or honorably,
be copied, used, or claimed by anyone other than the owner.

__________________________________________________
__________________________________________________
Computer Science Research
Section
Preface

Most of the following research took place between the turn of the century and 2015, when major AxleBase work was stopped.

All theory, research, system engineering, documentation, and web site maintenance presented here was done by one man.

The questions and postulates that gave rise to AxleBase seemed like foolish, unbelievable, and even embarrassing fantasies when first imagined; initialy so fantastic that they were not even shared with friends. Although now an operational system that seems obvious, when the project began at the turn of the century, such a system was undreamed-of.

The following questions and scientific postulates were developed over a period of years in the sequence shown as the AxleBase project progressed. The first three arose almost together near the turn of the century.

Scientific Rigor :
      Rigor is faulted. For example, the Data Entity Magnitude postulate uses the word "larger" instead of a specified quantity or bounds. However, an attempt to make the work appear more rigorous than it was in actuality might assign undue professionalism to the worker and his work.

Engineering :
      Science is a search and engineering is application. But they were necessarily interwoven here because sophisticated systems had to be built by the researcher to provide tools for the science. The researcher shares both of them in the belief that where enjoyment of one is found in a person, the other lurks nearby.

Please enjoy the following with me.

__________________________________________________
__________________________________________________
Computer Science Research
Section
Postulates And Findings

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Project Resources

Postulate :
      One man can build a major software system.

A simple proposition, but the fact that our civilization has poured hundreds of billions of dollars into developing the I.B.M., Oracle, Microsoft and open-source database managers made this simple proposition non-trivial. The task was strongly believed, even by the researcher, to be impossible.

Finding :       True.

Proof :
      A relational database manager is one of Man's most
          complex engineering tasks.
      AxleBase is a relational database manager.
      It was conceived, designed, and built by one man working
          alone and unfunded.
      To insure investigation integrity, books and internet were
          not consulted. All work was original.
      Proof is enhanced by the system's new technologies far
          exceeding the abilities of existing systems when it was
          finished.
      Ergo:   One man can build a major software system.

Qualification :
      Interestingly, the builder found that he had trouble entirely grasping the complexity of that which he had built after project completion.

Caveat :
      The task required years of seven-day work-weeks, so its accomplishment is not necessarily an indication of cost, but it certainly did not require billions of dollars.

Extensibility :
      Is the finding generalizable ?
      Yes. So long as new computer technologies support systems of complex logic structures, and so long as development tools are not denied to the individual, we can expect retention of validity.

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Programming Language

Postulate :
      A major system can be built with a common high-level language.

Finding :       True.

Proof :
      AxleBase was built with the world's most common language
            at the time; Visual Basic version 6.
      Ergo:   A major system can be built with a common
                  high-level language.

( Note :
      Visual Basic version 6 is totally unlike the subsequent language with the same name, which was marketing obfuscation. )

This postulate was addressed concurrently with the previous and following postulates.

Another way to state the finding is that the only meaningful measure of a computer language's power is, like any tool, the power that it delivers to the worker.

High-Level :
      When applied to a computer language, this term refers to the abstraction level. A high-level language generally encloses many operations within each command so that programming is speeded and simplified, thereby increasing the programmer's power.

Extensibility :
      Is the finding generalizable ?
      Probably not in the near future because the business world has found that a reliable high-level language allows the user community to compete with the manufacturer. Therefore, the sale of such powerful tools has been stopped, at least for a while.

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Platform Requirements

Postulate :
      Complex systems can run on cheap personal computers.

Finding :       True.

Proof :
      A high-end RDBMS is one of Man's most complex systems.
      AxleBase is a high-end RDBMS.
      AxleBase runs on cheap personal computers.
      Some of his test machines were built in the last century.
      Ergo:   Complex systems can run on cheap personal
                  computers.

This was addressed concurrently with the previous two postulates.

Extensibility :
      Is the finding generalizable ?
      Not only was AxleBase placed on personal computers, but it also contains more advanced technology than do the popular database manager brands. Furthermore, AxleBase is designed to run on computers of the last century as well as new ones. The investigator therefore believes that other kinds of complex systems can be built for very small computers.

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Data Entity Magnitude

Postulate :
      A relational database manager can build and manage tables larger than any conceivable size.

When postulated at the turn of the century, a table of a few gigabytes on a mainframe was considered large.

Finding :       True.

Proof :
      AxleBase can handle tables in the exabyte range and has
          been tested with petabyte-sized tables.
      Ergo:   A relational database manager can build and
                  manage tables larger than any conceivable size.

( Although simple internet search engines and data warehouses followed, they lack the power of relational database managers.)

Extensibility :
      Is the finding generalizable ?
      The result was achieved entirely through software engineering that allowed vast tables to be built on hardware that held only a few megabytes. The result is therefore felt to be generalizable to other software systems.

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Platform Capacity

Postulate :
      Large data entities can be managed on personal computers.

When postulated at the turn of the century, a personal computer disk drive with capacity of more than a gigabyte was unusual.

Finding :       True.

Proof :
      AxleBase began building tables of many gigabytes.
      A test table of 100 Billion rows was built and tested.
      AxleBase runs on personal computers.
      Ergo:   Large data entities can be managed on personal
                  computers.

Extensibility :
      Is the finding generalizable ?
      See the previous postulate.

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Software Expansion Of Hardware

Postulate :
      Software engineering can expand hardware.

Using AxleBase for proof added confounding requirements :
      1. The mechanics of the expansion had to be transparent to
          the user to retain industry-standard use of the software.
      2. Expansion could endanger neither data nor data returns.
      3. Expansion could not slow the database operations.

Finding :       True.

Proof :
      A permission list can be given to AxleBase.
      The list consists of computers and drives that he can use.
      He uses them as he needs them.
      He distributes large data tables across them.
      He manages the distributed data thereafter.
      The effect is the expansion of local storage to any size.
      System users are unaware of this behavior.
      A table of many gigabytes was built across many computers.
      Ergo:   Software engineering can expand hardware.

Observation :
      Expansion is virtual and from the software perspective, but the important facet is that it is functionally real to the administrators and users.

_________________________
Computer Science Research
Section
Postulates And Findings
Sub-Section
Software Enhancement Of Hardware

Postulate :
      Software engineering can enhance hardware.

Using AxleBase for proof added confounding requirements :
      1. The mechanics had to be transparent to the user to retain
          industry-standard use of the software.
      2. Enhancement could endanger neither data nor data returns.

Compared to the others, this postulate was especially daunting. There was no conceivable route to a proof at the time. It was easy to imagine solution by a huge organization, but not by one man.

Finding :       True.

Proof :
      A table of 100 Billion rows was queried in seconds on the old mechanical disk drives.
      The AxleBase system can be configured as a super-system.
      The super-system can use multiple computers for processing.
      The number of computers is unlimited.
      In his super-system form, AxleBase uses the power of
          many computers for database management tasks.
      AxleBase can drive desktop computers to surpass
          mainframe power in some operations.
      Ergo:   Software engineering can enhance hardware.

( The solution was found within a highly abstracted conceptual plane making the elegantly simple solution appear more sophisticated than it is. See the following Unexpected Findings section.)

__________________________________________________
__________________________________________________
Computer Science Research
Section
Unexpected Findings

Computing Power

A totally unexpected finding from all the work was possibly the greatest; i.e., that considered software engineering can overcome hardware shortcomings. A factor in this research that is obscured by the mass market and the mass marketers is the fact that the sophisticated AxleBase system will run just fine on a turn-of-the-century desktop computer. His development was, and in 2014 still is, done entirely on a Windows 98 computer with a four gig hard drive. His test machines included Windows 95 and an even older Windows NT machine until they finally died.

A result of that finding is the indication that software engineering in business and academia has been a farce. This is especially true considering the fine tools available and the billions available for funding.

Software Engineering Impact

The Google corporation was understandably proud of the speed of their Dremel database manager. But Google relied upon massive amounts of hardware for that speed, whereas the AxleBase code was laboriously designed and engineered for power. A comparison revealed that AxleBase was over four thousand times faster than the Google system due entirely to software engineering.

(See Performance Comparisons : Dremel on the performance page.)

Unexpected Benefits

It was found that care in engineering and coding frequently point to, and offer unexpected support for, additional design features. Features that are real and not superfluous "bloatware".
      For example, the distribution of the database manager across multiple computers unexpectedly offered system reliability that transcends that of the local computers. Only minimal engineering and programming were needed to capture it.
      The "unexpected benefit" effect became so great in the project that the researcher finally just smiled and said, "Enough.".

Solution Simplicity

The project required a great deal of time and thought because it was original research and because of its massive size, but it was found that many solutions within it are far simpler than expected. The alterations of hardware characteristics, in particular, appear far more complex than they are in actuality because the complexity lies within a highly abstracted conceptual plane where the solutions were found, investigated, simplified, and then brought back down to the implementation level.

( This simplicity section and the Problem Domain section do not conflict because different domain abstractions are addressed therein.)

Engineering Foundations

Software engineering was found to have extremely important basic principles that are not necessarily intuitive. Simplicity is an example. This researcher frequently interrupted work to return to supporting systems to simplify them. The effort delivered several tangible benefits, but most importantly for the researcher was that the beautiful simplicity enhanced the assessment of structural efficacy and also freed the mind to build more complex structures atop the simplified base; probably understandable by engineers in many disciplines.

__________________________________________________
__________________________________________________

Web site and technology
Copyright 1999 - 2022 John Ragan

Web site is maintained with Notepad.
By intent, so don't bother me about it.