Since starting my PhD the number of digital tools I use has increased. The current list includes Microsoft Office products, Zotero, Evernote and Trello, and I have talked to people about Mendeley, Scrivener and more.

As an archivist, and someone who has worked on research data management issues (including providing training and advice to postgrads, not to mention a few academics), my first question is always: what’s the exit strategy? So much software is marketed on how easy it is to get your stuff in. Often, finding out how to get your stuff out is harder.

For example, Microsoft Office isn’t too much of an immediate concern given its ubiquity (and the ready availability of alternatives like Open Office which will open Microsoft files, albeit sometimes at the expense of fonts or fancy formatting). Regardless, I always push key versions of things like Word documents to PDF to give me a fixed point-in-time record stored in what is (post-1 July 2008) an open standard.

Zotero has a whole pile of export formats, from bibliographic standards to .csv files, so makes life easy for its users. Trello has a json export which is unwieldy and not particularly useful, but I don’t keep a lot of content in the tool – though it’s a key part of my daily and weekly planning, it’s just a glorified ‘to do’ list – and if I get really worried I can upgrade and get a .csv export too.

And then we come to Evernote. I’ve only recently started using their note taking tool, after many recommendations, and from an ease of use perspective I can see why it has so many fans. Good syncing, clean interfaces, tagging, web clipping and more. But its exit strategies leave a lot to be desired. I can export my notes as HTML, but this leaves tags and other metadata behind,  which means it’s not an exit strategy at all – it’s a partial representation of a portion of my content, in a format that’s not particularly useful. The only other option is a .enex file, which is proprietary Evernote XML.

These .enex files are human readable, and someone with a little skill could no doubt write some code to parse them, or transform them into a more useful schema. But when a program is asking to become a significant repository of research notes, web references, images, videos and other content, I would like more.

I don’t (necessarily) blame the vendors. Evernote is a business, and proprietary formats help lock customers in to their products. I’m actually more frustrated with their customers. We should be asking for better exit strategies from our software, and if we don’t get them we should either be going to competitors who do, or (at the very least) putting in place workflows that avoid aspects of the software which cause lock-in. Make good exit strategies one of our conditions for using a paid product and software developers will soon sit up and take note.

Apart from  avoiding the paid version, I’m not putting anything mission critical in Evernote without putting it somewhere else as well; at the very least I print regular copies of key notes to PDFs which then get incorporated into my local storage and backup. Plus, I’m not going to use their tags or other features to encode important data, because I can’t get that data out easily. It doesn’t mean I won’t use their tags, just that I won’t use them as a key part of my notes or information structures.

More broadly, when choosing the right tool for the job I keep in mind the fact that the ‘job’ is often not just to solve my immediate problem, or even to help me for the next three years. It is to capture my research in a way which means I can access and use what I need now and in to the future. All being well, ‘into the future’ could be for 30 or 40 years which means I won’t be accessing it using the tool that captured it. That means I need an exit strategy now.

I want to finish with an idea. Finding information about exit strategies and export functionality is often harder than it should be. Sometimes it even involves downloading a program and just playing around to see what your options are. Others have good documentation, but it’s all in different places and different formats.

What I would like to see is a wiki (or something similar) which the whole community can contribute to, documenting exit strategies for software, such as:

  • how to get data out
  • available export formats
  • what data (if any) gets left behind by what format
  • known issues

When people write code to parse things like Evernote’s .enex files it would be great if people made this accessible too. That way we don’t have to collectively spend our time doing the same things. We should be able to build on other people’s knowledge and skills; stand on their shoulders, if you will. (If something like this already exists please let me know.)

And if no one else does, as soon as I clear some of my current commitments I’ll do my best to get it up and running myself, no doubt while I’m procrastinating over something else. The first priority for the project: finding a suitable tool with a good exit strategy.