Data Art Miles

17 Feb 2016

The Concept

I find myself often wishing for a digital outlet, a way to share simple programs, that manipulate data in some way or another. The data I want to manipulate is the everday common data that exists around us. The threshold for me a skilled programmer to do this is too high. We need a way to lower this threshold creating some digital “bot” art. The process should be more conceptual and less onorous than exists today. What we really need is geocities for “microservices”. Where microservices are written in javascript or the most widely available language of the day. If we’re able to utilize docker containers, and large pieces of infrastructure like kafka,storm and hadoop, we can free ourselves from writing unspecialized programs to execute within a custom context.

What it looks like Today

  1. Decide what you want to make
  2. Figure out what language context your going to write your code in* a. Web based code b. Mobile based code
  3. Figure out storage or interaction/engagement level
  4. Figure out how to get any data you might need
  5. Figure out how to host anything you might need,web,database, etc…
  6. Figure out how to share the thing you made and make it accessible

If you strip it down you get Almost 90% of the work building, scaffolding, and very little free form expression

What’s necessary

In order to reduce the threshold and have a better experience I suggest we build the following structures:

  • Build simple javascript web/native framework
    • The framework should accept a simple javascript object that accepts input and outputs data
    • The framework should only accept the javascript code and the end user shouldn’t know about wiring into the framework or anything like that
  • Build a simple abstraction that allows for UI input ‘
    • This abstraction could just be React/React-Native
  • Build infrastructure that allows users to upload data or stream data to their aformentioned javascript object
    • Without some data storage everything is lost data should be transient or permanent and the database that backs the process abstracted/ same for storage content
  • Host the content, manage all of these live processes in containers and allow them to be addressable

When we do this what do we really have

At the end of the day, its a spec for how to make a microservices platform. That is code that is highly specific, yet is able to leverage existing infrastructure to do alot of the ugly.

comments powered by Disqus