User:Merbst

Hi everyone, my name is Matt Erbst. I am gifted by a surplus of ideas, and cursed by an inability to organize these for others to understand! I hope that by structuring my User subpages like this, I will be better able to collaborate with other Wikimedians & Open Content advocates in the future, starting at the Developer Summit 2018.

User:Merbst/Autobiography User:Merbst/DandelionLabs User:Merbst/OpenAmbitions User:Merbst/URIs_on_my_dev_servers User:Merbst/MediaWiki_Ideas User:Merbst/Wiki_Projects

I have been a software developer my entire life, a believer in the power of mass-collaboration & the WikiWay since 1997, an avid reader of Wikipedia content since its beta in 2000, and an author of OpenSource PHP web applications since 2000.

As an individual having exceptionally high levels of curiosity, idealism, passion, and hubris in the face of pragmatism, I have spent large amounts of my time over the last decade studying pure math & philosophy, an evolution of my previous interests: software engineering, web development, taking trivia seriously, and interpersonal infatuations. These personal idiosyncrasies, combined with a lifelong struggle to achieve my optimum environment to productively master my limitations of attentional focus balanced against the steeply varying levels of drive reflecting the relationship of my labors to my passions, have spawned many lifetimes of ambitious project ideas. My sole focus at this stage of my life is to bring into existence as many of my ambitions as I can achieve. Luckily my true ambitions are highly altruistic, and my own ideals are very congruent with the Wikimedia Movement, giving hope that within this movement I will find many altruists with whom I can work synergistically to achieve mutually beneficial alliances.

I was born in Southern California in 1981, where I continue to live, specifically in the city of Laguna Niguel, CA. I enjoy sharing the unique atmosphere of my tourism-friendly area with visitors, especially playing guide & offering my perspectives on what guests will find most worthwhile. Beginning this year, I will be experimenting with organizing workshops, meetups, hackathons for Open Source projects intending to introduce previously overlooked audiences to tools they could find valuable uses of, and encourage developer cross-pollination among complementary projects.

I registered the domain name www.merbst.com for me to gather the authoritative links of all my personal web profiles & projects going forward.

I owned many websites in the past reflecting several inside jokes, previous projects, and on behalf of my friends and family.

An abridged list of these (sometime soon providing Internet Archive links to their former contents) is:

erbst.org Previously mine... I would appreciate owning erbst.com in the future, but it has always seemed unreasonable to capitulate to the $3000 asking price of the cyber-squatting company owning it.

wikicms.org & .com This was my idea to provide a Content Management System implementing an extremely simple low-user-investment interface for quick modification of permitted items within a unified class-based namespace of CSS stylable term-keyed elements for use with customizable user-interface templates. The intended audience was highly custom community oriented websites built around custom written web-applications, with user authentication centralized, and permissions variable from fully anonymous submissions to very narrowly controlled user access.

autocoder.org My most ambitious PHP/MySQL project, inspired by reading about Table Oriented Programming, OnceAndOnlyOnce, and DontRepeatYourself on c2.com/wiki/ and recognizing the potential elegance afforded by a perfectionistic web application implementing a Declarative Language for specifying project requirements in terms of the datasets involved, the relations among these, the perspectives from which each user group will be presented their interfaces to these, and the secure bounds upon accesses permited of these. The foundational realization behind my AutoCoder project was that the SQL99 standard was entirely sufficient to represent all of these requirements with its CREATE TABLE and USER PERMISSIONS statements alone, and that many enhancements to the user interface could be inferred from the CREATE INDEX and CREATE VIEW statements corresponding to the database metadata suggested by best practices for performance-conscious Database Architecture. By design, the SQL99 standard implies a bijective mapping between its CREATE statements and the meta-data of the resulting databases. Homomorphic CREATE statements may be output by special forms of the DESCRIBE DATABASE commands of SQL99. I exploited this throughout my implementations of Autocoder to provide the functionality which inspired my choice to name it Autocoder, that analyzes existing MySQL databases to automatically provide a web application offering user interfaces derived by its structure, user permissions, and FOREIGN KEY to PRIMARY KEY relationships among its tables.

After I created a proof of concept for Autocoder in May 2001 implementing a C.R.U.D. (Create, Replace, Update, Delete) interface, I was eager to make use of it to rapidly develop data-intensive web applications intended for real production usage, as a demonstration of the principles I felt sure would take off as a new fad for Web Application developers worldwide if I made my source code freely available and published on SourceForge.net. I chose the LGPL license, due to my fondness for the work of the OpenSource Foundation, and pitched the potential for rapid development to my consulting client, GeneticsCenter.com which sought to migrate their legacy FoxPro and DB4 patient billing database application onto the WWW, modernize its interface security to be fully HIPAA compliant, and potentially offer this future-proof patient billing service to other genetics counciling businesses with similar requirements owing to their shared industry. After a year or 2 of intensive development & bugfixing, I had worked with a team of young developers to provide a working system meeting all of these requirements, while ensuring the continued modularity of AutoCoder's source code in a handful of PHP files functioning as a library. By 2004, I had a fully functioning FreeBSD+Apache+MySQL+PHP stack installed and running a clean Autocoder 0.7 instance on my own development server, which I had built myself out of commodity hardware, and colocated at the www.he.net facility in Freemont California for $90 to $160 per month. I continued development of the codebase towards the standard I had set for myself to achieve a version 1.0 release, which I reached in 2006. The desire to achieve compatibility with PostgreSQL & my contemporary WikiCMS.org project, as well as fully exploit the potentials of proper utilization of well formed HTML4, CSS 1.3, and JavaScript for AJAX features led me to fork out a new branch to be named AutoCoder2.0. My nascent interest in mathematics, and intense exploration of Computer Algebra Systems proved to be a distraction from any further AutoCoder development on either branch, until 2016. Financial limitations suffered in 2007 unfortunately caused the lapse of all my domain name registrations & prevented my continued monthly payment of the colocation service. It is interesting to note the commonality of the Python based Django project with the directions I was intending this project.

In 2016 I recognized how much development effort on a project idea I had named ErbstKeep could be saved if I used AutoCoder as a base. This presents an opportunity to modernize my AutoCoder dream with a version 3.0 emphasizing the timesaving potential of rapidly generated responsive HTML5 user interfaces to many of the most prevalant data-sources & datastores used today, expressing application requirements with cross-platform UX-aware UIs showing consistency of design & quality able to convincingly obsolete prototype mock-ups.

The essential featureset additions necessary to my requirements for AutoCoder 3.0 are:

Modularity of data-sources & data-stores to cleanly implement input/output of/to:

+Structured text files including XML, JSON, CSV formats specified by templates

+Arbitrary RDBMS interfaces, including SQLlite

+Arbitrary NoSQL and distributed databases

+(Symantic) Web-services

Modular extensibility of UI to cleanly present:

+Tree-structured representations for nested set structures

+Graph structures

+Information Theoretic metrics & modern Statistical measures

+MathML

+Unicode 10.0 symbols

+Code markup