Docear - The Academic Literature Suite

Posted by Alexander Deliyannis
Apr 21, 2012 at 02:30 PM

Dr Andus wrote:
>I’m a bit wary of the Swiss knife type approach to writing software.
>While I’m impressed with efforts like Citavi or Docear that try to capture the entire
>writing process, I’m reluctant to give up my freedom to assemble my own system of
>specialist software for the various stages of the process. I guess I worry that it
>somehow constrains my creativity.

I’m a bit wary of the Swiss knife type approach to software in general. That said, assembling one’s own system implies the interoperability of the components which is anything but ensured. So far I can think of the following de facto ‘standards’ for information interchange among discrete programs:

- PDFs for reference documents
- Endnote format for citation databases
- IE Favourites and Firefox Bookmarks for collections of URLs
- OPML, tab indented text, MindManager and Freemind format for outlines
- RTF, .DOC(X) and .ODT for final drafts

These leave a lot to be desired, as they don’t cover staff like comments and versioning, as well as most formatting and metadata in the case of outlines. There is of course XML, but as long as there are no specific standards, one needs to be a programmer to create the necessary transformations among different implementations.

It would be great if programs could fit together like LEGO, even more like LEGO Mindstorms, but they don’t. Then again, even LEGO Mindstorms are made by one single company. In fact, one could argue that a lot of interoperability already exists among several applications under Mac OS thanks to Apple’s rules.

I have a dream for such seamless interoperability which can be described with the example below. I would note that there are workarounds to do such work already, but they can get substantially in the way of workflow.

- While doing research on a subject, I collect material from various sources, in HTML, PDF and multimedia form (audio, video and images). My collection tool (let’s call it Evernote) can handle them all, always maintaining a link to the original source for reference.
- As I scan the material, I highlight and annotate certain excerpts. The highlighted extras and notes are maintained as virtual notecards (e.g. in Supernotecard), linking back to the original material (local copy and source URL).
- I view my collection of notecards on a virtual corkboard, where I can arrange and group them together (e.g. in Stickysorter or Visual Concepts). Patterns begin to emerge. An alternative would be to organise them in a wiki like Connected text.
- I organise the notecards within each group, e.g. with CRPA, and write an abstract for each group, representing the main underlying idea.
- I view the grouped notecards in my outlining application, e.g. Scrivener. Each group represents a chapter in the high level outline, headed by the abstract.
- I start working on each chapter. I need much more granularity, so I launch Brainstorm or Sense for the actual editing.
- As work on a chapter progresses, I review the chapter abstract and check the updated high level outline for consistency.
- At any time, I can switch back and forth between the various views, which remain in sync.