grep jason_

Plain Text is Portable! Or is it?

This was originally written in March 2022 for the Hemispheric Newsletter as part of One Prime Plus for the Hemispheric Views podcast.

It seems to depend on what portable means to you. I have been a long-time backup person. Back up something, then backup that backup, and then go ahead and copy it again to somewhere else. No such thing as too many copies as they say... (Wait, who says that? I don't know, but let's just pretend they said it.) Having a plan B is fantastic, but a backup plan is a two-stage approach.

  1. Backup the data.
  2. Verify the integrity of the data so it can be repurposed.

A number of years back the term Data Portability started to be thrown around the internet. The idea is that your data should not be trapped in a silo of some company, but rather, you should be able to freely export that data so you can use it elsewhere. Well, that's kind of true...

I think we tend to confuse portable with convenient when talking about our digital assets. The idea of getting your data "out" of a silo is only part of the story. The format that data is stored in plays a huge role in what can be done with it after the fact. I have not been bitten by this too badly in recent memory, but I am having some mildly concerning thoughts as of late.

Example One: my blog.

I host my blog on Blot and love the service. All my posts are a series of Markdown (.md) files. Plain text. What could be more portable than that? Nothing, really. Unless I want to move those blog posts to another platform with any bit of efficiency. To be convenient I would really only be able to move to another service that works almost identically to Blot. I don’t have any reason to want to move away from them, but it does make me rethink this whole data portability thing.

Example Two: my notes.

I use Craft. I love Craft. I talk about Craft (a lot). They have a great export system that I went over in a previous newsletter about “How I Craft.” Similar issue here, however. I can export my data in a bunch of ways, which is fantastic, but then the pesky second half stands up in your way again when trying to import it into another service. Depending on the content of the notes this may be more or less effective.

So what are you supposed to do? Well, you can hope you pick the perfect service for you on the first try and then never leave. That would be nice, but probably not realistic. The next best thing is to look at what kind of data export options exist and see how “standard” they are across the industry. It’s obviously not strictly in a service’s best interest to make leaving easy, but I see that as a strength of a product or service to do exactly that. Make it so good and useful that leaving isn’t something I want. Should I outgrow the service or it simply doesn’t perform the actions I require, the platform should either adapt or acknowledge the fact and let me go in the easiest way possible.

It may seem counterintuitive, but the next time you are going to sign up for a new service, look at these two things first:

a. How easy is it to export your data and in what formats? (simply exporting to their proprietary format does not count)

b. How easy is it to truly delete your account and remove your data from their servers?