an Underpants Gnome Model for Collecting Maths

Preparing my tools to study maths

Day 2 of the monthend. Good time to chill and reflect. Current attack plan is :

1 : the TPL generator

Outputs : TPL, via LDD and SGAS

Write a friendly UI, that helps the user to specific the parameters of an arbitrary target programming language ( TPL ), compiling to some programming language description document ( LDD ). From the LDD, generate exemplary source code in the described language. Upon the exemplary source code, show the user how a lexer and parser will operate. Inconsistencies in the LDD should become obvious, so it may be possible to lead the user to iteratively debug the LDD. The parser should output some sort of standardised generic abstract syntax ( SGAS ). Compile the SGAS to JavaScript. Use JavaScript because it is easy, and also a vastly deployable medium.

( In the future, hardware idioms should be presented, to demonstrate target implementation concerns, during the iterative LDD creation process. At least include : bare metal vs. virtual machine, memory layout, OS processes and threads, green threads and alternatives, memory management and locking, async, etc. Eventually it should be possible to model all intermediate storage representations and the operations conducted upon them, in varying degrees of detail, depending on the present needs of the user. ) 

Implement Program 2 using Program 1, as a matter of dogfooding.

2 : the TPL

Outputs : a study tool, via CND and GLDO

Design and implement a simple TPL, which has a lexical layer capable of easily representing the most common notation, used in the quantitative reasoning stack, of our current civilisation. In other words, the TPL should be good at representing the various notations used in STEM subjects, particularly the arbitrary 2D layout of text and numbers, without much concern for how that is subsequently processed. This capability is the capture of notational data ( CND ), terminating in a regularised object storage format. This should be the main frontend concern of the TPL.

The TPL's middleend concern should be, to have available internally, various idioms of operating upon arbitrary data, from first principles. This will require a review of information processing techniques, across rudimentary logic, grammar, set theory, arithmetic, and geometry ( the liberal arts ), and the subsequent creation of some sort of generic library for data operations ( GLDO ), during the iterative process of LDD development.

The TPL's backend should be mosly handled by the TPLG, so it should not be highly concerning to the user at the early stages of using the TPL.

Finally, the UX of the TPL must be facilitated with a practical guide to iteratively mapping, from the output of CND, to the GLDO. Thus the TPL will provide an environment for the user to study unfamiliar notational data used in STEM.

3 : Profit

Finally, should the above be accomplished to a reasonable degree, then I will finally regain hope that my study of maths can begin, upon reasonable tooling. 

link