Background

About three years ago, I bought a license for Scrivener, and have since written up almost all of my papers that don’t require math typesetting in it.

It is not an especially good word processor. In fact, it’s kind of terrible as a word processor and is geared mainly toward writing mostly-plain text. Its distinguishing feature is that, rather than working on a single document, it effectively manages a written work as a project, like in an IDE, with support for building the output document from multiple internal files, and accomodates storing various metadata, notes, arbitrary related files &c. in the project.

This can be helpful when working alone, but groupwork especially demands this sort of organizational aid to keep track of what’s being done on a project. Unfortunately, nothing similarly useful and accessible (to ordinary users) exists among collaborative software offerings.

Some issues with existing services:

  • only deals with a single document, without a good way to add metadata or attach related data (e.g. Google Docs). Many projects would benefit from being able to attach data files or even just a graph/photo/figure that should go in the document somewhere, but without deciding the precise placement until cleaning up the draft for publication.
  • rely on an explicit check-in/update workflow (hosted VCS, e.g. git). This only makes sense when we have either discrete changes or numbered drafts. It’s impossible to edit a file together, only make separate changes and hope that the results don’t conflict and require an arcane merge procedure.
  • only deals with downloading/uploading files, without providing an editor (e.g. Google Drive). This is here only because it sort of integrates with Google Docs to provide a way to organize related files, but the integration is awful and nigh-useless.

The collaborative editor element seems to be very poorly explored, while almost all project management tools marketed to individuals (rather than B2B) are only suitable for source code, perhaps because the only people who write these things are mostly interested in code.

The only service that seems to fill a similar niche is ShareLaTeX/Overleaf, and their product is focused on providing a TeX environment in the browser moreso, and so caters to a rather specific audience.

As a result, in my experience, groupwork that needs to produce a written artifact usually either

  • doesn’t get done without meatmeetings
  • involves emailing things around
  • involves emailing links to Google Docs around
  • the above, except with increasingly desperate Facebook messages instead of email, past midnight on the due date

Proposal

Build a web application for interactive collaboration on a written project. Basic requirements are file storage and organization tooling, a collaborative editor, export to a common format, and audit logging.

Beyond that, the application should provide functionality useful to writing up an article/paper/novel, like referring to reference sheets, or making it easy to look up earlier mentions of a term/name/reference. A lot of this is actually very similar to what an IDE does, but in a natural language context. It doesn’t necessarily have to be very smart to be useful, though.

Ulterior Motives

I spend a lot of time around amateur translation groups. Version control doesn’t get used because it’s too hard, and a lot of workflows involve copy-pasting scripts around collab tools with subtly different capabilities and/or defaults. A lot could be improved by software that better fits the use case than already exists.