The Literary Machine is more than a product. It is an ongoing project. Among wordprocessors, brainstorming tools and mindmappers, LM is steadily advancing in its own unique directions.

The Literary Machine is a workbench for writing:


The LM database with its specific suggestion for how human thought and memory can be efficiently mapped is the heart of the machine. The basic structure will not change. Another basic idea about this would require its own new product.

However, marginal changes usually come with new versions. The outlining service in LM may require changes. New added object types (like "LM Tables" and "LM Maps") add database objects. The endnote/reference service is another subsystem that requires own database arrangements.

LM uses a classic PC database tool: Paradox. This has proved to be a good choice. What remains to be done on the storage side is to construct any import and export tools that could be needed. The techniques that have been used so far are XML and ODBC.


2. Note-taking

The ideal note-taking application should be in the tiny PDA and it should export your text into the big and extremely flexible text database. We already have that kind of tools and I expect that they will develop further.

But of course you also want to retrieve and manage your huge database in the tiny PDA.

This is a demanding task. All applications that "replicate" data between separate systems and databases are complicated.

This is in a way the same task as sending data between different LM databases. If users want to share LM data and how it should be done is still an open question.


3. Outliner

LM is provocative in its unusual way to link data in a thought database. It however takes a few steps back to generally accepted structuring paradigms in supporting an outlining tool with traditional ups and downs and invariant connections in a traditional tree arrangement.

The outliner is still in development. The outline work is on the "text stream" application side in LM vocabulary (while LM words and LM concepts demonstrate the relational and timeless system side in LM).

Outlining in LM moves the database text up on the assembly line for text production. Outlines are sent to text files, HTML pages or to the OEBPS eBook format.


4. eBook delivery

Since more than a year LM features ebook authoring with the W3C-supported OEBPS standard. eBooks are developing at an agonizing low speed, but as with much other computer development it will become midstream and accepted one day.

The idea with (standardized) eBook authoring is that books often require minimal formatting. That is, the value is in the text. The LM eBook assembly line produces in the same way as original HTML or CSS stylesheet produces. (Which in fact comes close to an ordinary printed novel.)

If you make your HTML page (and book) like typesetting by hand, LM is not your tool.


Building an endnote tool with XML formats

Work with bibliographical references (like in research work papers) has many similarities with the LM text model and since long there are features in LM aiming at this important application area. The development has so far been slow, waiting for more participating users.

In LM version 2 most parts are in place, included a flexible XML database implementation.

The next step in development will be to develop the use of parallel LM program instances. LM Bibliographic databases could be independent LM applications. This would increase both database security and database flexibility. Data would be picked in those separate environments and sent by XML to your working environment.


Responding to the incessant demand for graphics features

Mind tools are very influenced by the dominating mindmapping paradigm and many users argue that LM should be mindmap-like.

However, LM starts in another direction. In fact, you can see LM as a kind of anti-mindmap project. But graphical tools are often good. Images have always been favored in LM and graphics with connecting lines could also have a place.