Platform obituaries

I appear to have been writing obituaries for social networks for the better part of a decade, one at a time, without noticing it was a series. So here they’re together, with what each one actually cost me.

Google Plus, October 2018

I wrote about G+ when the shutdown was announced. What strikes me in hindsight is how little I minded. Google Plus was the only one of these where I never really moved in, so when it closed there was nothing to get out.

Lesson, such as it is: the platforms that hurt when they close are the ones you were productive on. Somewhere there’s a version of me who put real effort into G+ and lost it, and the only thing separating us is that I couldn’t be bothered at the time.

IFTTT, February 2019

The smallest entry and by far the most instructive. I published a post titled, in its entirety, an observation that the IFTTT link had broken.

What had actually happened was that an automation quietly stopped working and a chunk of this site stopped receiving things it was supposed to receive. No announcement, no shutdown, nothing that warranted an obituary. A service simply changed and my site was collateral. There was no drama at all, which is exactly why it took years to appreciate that this was the important one.

Instagram, April 2021

The post was called “Ugh Instagram” and I stand by the analysis.

This is the one that eventually did real damage, but not in the way I expected. It wasn’t the algorithm or the ads. It was that the plugin I had used to bring photographs onto this site had never actually saved the photographs. Years later every one of those posts was an empty box. That has its own post and I won’t repeat it here, except to note that the platform didn’t have to shut down to take my content with it. It only had to change a URL.

Twitter, March 2023

I downloaded the archive. That’s all the post said, and it turns out that was the right response too.

Twitter is the only one where I acted before rather than after, and the difference is enormous. I have the file. I haven’t looked at it once, which is fine. The point of an archive isn’t that you read it, it’s that the decision about whether you can is yours.

Bluesky, November 2024

Which brings us to the current one, written while it’s still alive, which feels like tempting fate.

I don’t think Bluesky is about to collapse. I thought that about the others too, at the equivalent point. What I have changed is the assumption underneath: I now treat every account on a platform I don’t control as a broadcast channel with an unknown expiry date, and I keep the original somewhere I pay for.

The pattern

Five entries, and only two of them involved anything actually shutting down. The rest were services that carried on existing while quietly breaking the thing I was relying on.

That’s the part I got wrong for years. I was watching for shutdown announcements, which are loud and give you notice. The real risk was never the funeral. It was the maintenance window nobody told me about.

So the rule I now use, which cost me about fifteen years to arrive at: if the file isn’t on a disk I pay the bill for, I don’t have it. Everything else is just a link to a decision somebody else gets to make.