Prepare for actual first public release (pre-release):

- cleanup old, irrelavant docs
- write a README that offers a brief overview of Touch's ideas
- add the MIT license
- allow for elements to be disabled via config which enables "removing"
  elements from the default config
- add extjson tool to convert extended JSON to plain JSON so anyone can
  use the "better" configs (easier templates)

Many improvments from the previous tag, some of them are:

- simpler node—only one node type
- add core DOM-like manipulation methods to node
- printer now outputs proper canoncial form (previously it didn't even
  support many elements properly)
- simpler config
- document exported types, functions, and packages
- usable CLI
- config overrides (shallow merges)
- autolink URLs
- parser keeps newlines in text
- add position info
- don't replace illegal characters

Add formatter and support multiple export formats

Move packages out of internal dir

Change module path to be able to `go get` module

Public release (pre-release)

Fix parser skipping valid UTF-8 encoded characters. For example, a

Unicode character at the start of a paragraph was being skipped. Add a
few simple test cases.

Fix parser not replacing NULs at the beginning.

Define source code representation: encoding, whitespace, line endings,

and disallowed characters. Update parser to specification.

Fix mixed list identation: a tab equals 8 spaces when using mixed
identation for list items.

Escape sequences, disallow nested links.

Add escape sequences to represent special characters. Escape sequences
start with a backslash '\' and are followed by the special character we
want to escape. All delimiter characters and the backslash can be
escaped.

Fix links; do not permit nested links and allow '<' to be placed inside
the link (link text or destination).

Printer, renderer, CLI, and lines.

Add a pretty printer (decoupled from nodes), HTML renderer, and a simple
CLI which reads To and outputs HTML or pretty printed parse tree.

Add paragraph and list item lines. Lines are now represented the same as
in To source. For example, To paragraph lines can now be broken into
lines in HTML by line breaks.

Fix '>' causing infinite loop.