Substance website


Substance and related projects

by Jack Park
...

Demo here:
http://substance.io/explore
Slideshare here:
http://www.slideshare.net/_mql/substanceio-content-is-data

I am having difficulty keeping from going ballistic over this one.
Consider this lonely statement lifted out of a page on "getting
started" http://interior.substance.io/reader/#substance/getting-started
:

"Substance is different from traditional text editors as you arrange
content nodes that form your document, rather than writing text
sequentially. A content node denotes a fragment of a specific type
that can be Text, a Section, or an Image, etc.  Sections serve as
containers that can be nested and are used to structure your document.
However, as a user you simply need to click on a placeholder at the
position you want to insert new content."

What they are saying is that they are breaking a document into chunks,
individual nodes, some of which can be containers (sections).  From
their "document" description, using JSON for serialization, it seems
clear that each chunk, which I call an AIR (addressable information
resource) has identity, thus, it is an AIR:

http://interior.substance.io/modules/document.html

It is precisely the notion of an AIR which is exciting. This, I think,
blends with ideas brought out in Yuzuru Tanaka's many talks (and book)
about MemeMedia, the notion that there are libraries of "memes"
(information resources and tools) available for use, mixing, mashups,
and so forth.

We see similar ideas emerging in Ward Cunningham's Smallest Federated Wiki

http://wardcunningham.github.com/

A third project in this same space is TiddlySpace

http://tiddlyspace.com/

which is mostly pure javascript, but with varieties of servers,
including python in their version.  A "tiddler" is an AIR by my
definition, a discrete object with content and metadata.

In each of those three projects, Substance is the most complex, since
it borrows from Google's Wave some collaboration tools to ensure
"eventual consistency" among edits emerging during concurrent document
creation sessions.  Substance is also the least far along in
development.

A favorite desktop tool that animates this investigation is Tinderbox

http://www.eastgate.com/Tinderbox/

Comments welcome.
Jack

****

http://en.wikipedia.org/wiki/Scrivener_(software)
and
http://literatureandlatte.com/
and
http://www.lifehack.org/articles/communication/desktop-to-ipad-blogging-workflow-with-scrivener-elements-dropbox-and-marked.html

The comment here about "bidirectional" is interesting. It may be that
one "starts" with a perspective, then builds arguments to support.
What?

Jack

On Sat, Oct 27, 2012 at 7:53 AM, Mark Szpakowski
<markszpakowski@topicquests.org> wrote:
This is very relevant to the idea of a "Perspective Editor": an editor to create a narrative in the environment (closure?) of a framing, about an issue, with links back to evidence nodes supporting each discrete argument. That narrative could take the form of text, compound document, video (as in Popcorn Maker), etc, but the key is that it's structured as a DEM (Document Evidence Model), allowing exploding it into different mapping views. Writing tools like Scrivener are more and more supporting the idea of a document as a project, with distinct addressable nodes, which then get "compiled" into a result document in some sort of form (.doc, ePub, etc). In fact  using something like Scrivener as a front end to a supporting mapping editor might be a Good Idea™ (I think there may already be Scrivener/Tinderbox linkage).

I'm thinking more and more that this process needs to be bidirectional. Sometimes you want to generate a "current-summary" document from a structured conversation about an issue, going from nodes to text. But you could also start with a "thesis", which you then want to support by pointing to nodes in the evidence field: so you go from text to nodes. And of course going back and forth iteratively seems inevitable as well.

- Mark