Skip to content

Latest commit

 

History

History
69 lines (56 loc) · 3.23 KB

UserInterface.md

File metadata and controls

69 lines (56 loc) · 3.23 KB
Raw

User Interface

The user interface unites visualizer, editor and debugger into one. It is an integrated development environment which is not only meant for developers.

Getting Started

It is best to checkout the umbrella repository, because it ensures the compatibility of all submodules / repositories. Then go into your local copy of the UserInterface repository and open index.html in a browser of your choice either directly (make sure to disable local file restrictions and cross origin policies) or host it using Node.JS for example. Alternatively, you can also try the bleeding edge online version. The user interface is optimized for Firefox but also actively maintained on Safari.

Reasoning

Triples

There are 4 mathematically equivalent different ways to visualize Triples:

  • Hypergraph: Every Symbol is a point in a arbitrary space (in you imagination it might be 2D or 3D), Each data field is like a leave hanging from a Symbol and Triples are a strange kind of edges but not between two Symbols but three of them, having a direction from ... via ... through to ...

  • Aggregated: Personally I think of this as the most appealing form. It is a modification of the normal graph. Entities and their Attributes are collected in a Block with edges leading to their Values.

  • Table: If you come from standard relational SQL databases, this might be familiar to you: All triples are in a big table with these three columns: Entity, Attribute, Value.

  • Voxel: And if you come from a mathematical school, you might find this form reasonable. Every triple is a point in a 3D cartesian coordinate system. The three axises are: Entity, Attribute, Value. Little extra: A slice from the Entity/Value plane can be seen as a logical matrix of an Attribute.

Visualizations of Triples

Projectional Editing

If you are not familiar with the idea: Here is a good introduction.

Advantages:

  • Things that become unnecessary:
    • No scanning / lexing / parsing needed
    • No syntax errors and different styles anymore
    • No need for names (usually called symbols) to connect the sequence / hierarchy into a graph
  • Easier to release new versions / standards
    • Ground truth: Enforces clear distinction between semantics and representation / decoration
    • More homogenous structures / interfaces
    • Far less primitives, most can be moved to the standard library
    • Can be far more customized / domain specific (like DSLs)
  • Transformability (Program is just data like any other)
    • Better meta programming and reflection / reification
    • Better for machines to operate on
    • More and potentially better code complexity measure than LOC (lines of code)

Disadvantages:

  • Geometry improves orientation inside the source code for some people and it is hard to project this additional non-semantics forward and backward
  • Currently tools like GIT are optimized for sequences / text
  • Can not make errors / produce an invalid outcome on purpose