David Mytton

Keep a copy

Published in Privacy, Product Engineering.

Keep a copy

A few weeks ago I went to check the latest updates in my RSS feed reader, Feedbin, and it was down. I searched on Twitter and found a lot of people complaining, so it wasn’t just me. I couldn’t read any of my subscriptions.

Feedbin is a great service. It is usually reliable, fast, run by an independent company, can receive email newsletters, lets me save web pages, and respects privacy. It allows me to read all the blogs I subscribe to as well as following specific people on Twitter, because I don’t use any official Twitter client.

However, it is a SaaS product. I use their web browser client and everything is hosted on the Feedbin servers. If they’re down, I can’t access anything.

I don’t want to build my own version – replicating all the Feedbin features would be a waste of my time. I also don’t want to host an open source RSS feed server. The features wouldn’t match, but I also don’t want the hassle of running a server. Same as I use Office365 because don’t want to run my own email or file server. Microsoft does a better job at security, reliability, performance and functionality than I ever could.

But nothing is available 100% of the time, services don’t last forever, and there is always the risk of your account being locked out.

Where is your data?

Compared to the old days of running software natively on your computer and storing data on your local disk, we now access most things in the cloud.

If you use GMail, chances are you’re accessing your email from the web client. IMAP / POP3 is not enabled by default so most people are unlikely to use a native client like macOS Mail which stores all your messages locally.

If you use Google Docs, Sheets, Slides, there isn’t even an option to use a desktop client.

If you use Apple Photos, the default option is to store all your images in iCloud. It has an option to download everything locally but starts cloud-first.

What about your music? Most people have now switched to a streaming service, either Spotify or Apple Music. You pay a monthly subscription to get access to all the music in the world, but if you cancel the subscription then you lose access to everything. Files are cached locally, but they are protected by DRM and cannot be accessed from other clients, or if you cancel your subscription.

Kindle books are the same. They are stored on-device, but in a protected format that cannot be read except in the Kindle client. When your Kindle runs out of space, you must delete the older books. You can resync them on demand, but what happens if your Amazon account is closed, or Amazon revokes access like it did with an edition of 1984.

Video is even worse. YouTube is the defacto place for all video content, so it would be a shame if your content was removed, or your access blocked.

Losing access

The world has moved to a “rental” model. In most cases, this is a positive thing. Access is cheaper and available on demand. You no longer need to hoard CDs, DVDs or books.

But what happens when something goes wrong?

What if your Google account is suspended, or hacked? All your files, emails, calendar, todos, videos, photos would be unavailable.

What if Amazon closed your account? All your books would be inaccessible.

The big tech companies are almost unaccountable and lack transparency around how they make decisions about accounts and content. They legitimately take down a large volume of content, but there is inevitable collateral damage when accounts are wrongly marked as suspicious.

This applies to any service where your data is in a proprietary format, only located in the cloud, or the content is protected by DRM.

Ownership of content like video and books has always been a contentious issue. Technically, you are only “licensing” a copyrighted version of the film you watch, or book you read. But companies would not raid your house to take back physical books, DVDs or CDs. Now, they can remotely revoke the license, or suspend your account.

Keep a copy

This doesn’t mean you should stop using those services, or switch to only buying physical copies. It simply means its worth considering how you keep a copy of your data.

Almost all these services allow you to export and download your data. For example, Google Takeout will export all your files in standardised formats. Apple lets you request a copy of all your data, including your photos. Apps like Calendar and Contacts on Mac let you export all your data to an archive format. And if you are uploading videos and photos to YouTube or Instagram, you can keep copies of the originals or export them from those services.

Like an album? Maybe buy a copy of your favourites (from the artist directly to give them most revenue) and keep the audio files locally. Spotify can still find them.

Digital content needs the equivalent of legal deposit, but until then you can do it yourself. Don’t just rely on a single service. Direct everyone to YouTube, but upload a copy to Archive.org – they support all formats, from audio for podcasts to photos and software. Always keep the originals.

Work locally

If you store and work on a local copy of your files, you can continue even when that service is down, if you are offline, or if there are any problems with your account. It also means you can keep backups.

Have you ever looked inside the files that Google Drive creates if you run the desktop client? They contain a single URL which directs your browser to the web version. Non-Google formats are downloaded, but anything inside Docs, Sheets or Slides is entirely in the cloud. This means you can’t back them up like normal files.

It is better to work in open formats by default. Microsoft Office files can be opened by Libre Office, Google Docs and Apple Pages/Numbers. All the data is stored inside them, not on some cloud service. Sometimes you need the superior collaboration functionality of Google Docs. In that case, regularly downloading copies is sensible – Google downloads in Office formats by default.

Even better, work in plain text. This was one of the main requirements when I was looking for a new note-taking app. Plain text + Markdown means the files can be opened on any system by any app.

At least ensure the service can sync and cache locally. My new feed reading setup involves using Reeder, a local client which supports syncing with Feedbin. I can still benefit from all the Feedbin functionality but the app downloads all my updated subscriptions and caches them locally. Feedbin supports exporting to OPML so I can keep a backup of my subscriptions, but I have now introduced some redundancy by using a local client.

When selecting SaaS products, consider where the data is stored. Can you export it easily? How easy it is to work with the exported files? Or better, can you keep a copy locally whilst still benefiting from all the SaaS functionality?