All Good Things Must Come To An End

This post originally appeared on the online community Hi.co, which shut down as planned in September 2016. All the old posts from Hi.co will be archived by micro-etching them into nickel and storing them in a number of archives around the world. The original URL for the post was: https://hi.co/moments/0x9s5sl0

On the eve of Hi.co's imminent shutdown, I felt I should log in to leave some parting thoughts. There's a poignant finality to such a definite, scheduled end... and something darkly mortal about the idea of entombing our information in an archive, etched atop a piece of zinc (or zirconium, or zinclipscombite, or whatever mysteriously robust mineral or crystalline metal it is.)

I was never a huge contributor to Hi.co — but I arrived early on to deliver a rough-hewn satirical vignette on the life of Hi.co developer Gueorgui Tcherednitchenko, or whatever I could perceive of it from thousands of kilometres across the ocean. Gueorgui had encouraged me to drop in and try it out back when it was still hosted on "sayhi.co" — he'd been involved in the development of the project and even right after launch it was an impressively ambitious concept. I wasn't sure it would work, but it looked cool.

I came to Hi.co by way of App.net, where I followed Gueorgui.

App.net had an interesting past. Envisaged as a data backbone for all kinds of modern social networks and tools, it had an important difference from existing social networks: it would host no advertising, and was entirely funded as a commercial service users would pay for directly, rather than through the unscrupulous sale of their collective data, or their attention. A raised middle finger to the idea that nobody would pay for a service online, the very idea of App.net was central to many a fiery debate between those who believed ad-supported was the only way to fly — and those who wanted a break from it all.

The launch of the service was bright and optimistic; dipping its figurative toe in the water with a crowd-funding campaign targeting a bombastic (for the day) US $500,000 — a goal it ultimately succeeded by a good margin.

Tech glitterati from all over came to try out the new service, hedging their bets between this utopian vision of the future and still carefully maintaining their stage presence on Twitter and Facebook. New client apps appeared, crude at first but rapidly refined. Jamin Guy and Jared Sinclair's app Riposte, in particular, remains one of my most loved apps on iOS despite having been pulled from sale. The early warts in ADN were mythologised by its primarily technical users ("_xyx_pau: Never forget!") and a distinct culture rapidly developed.

Every day on the newly birthed network was exciting. New ideas emerged, conversations came thick and fast, debates were robust and (especially compared with Twitter) particularly respectful. A frequent theme of discussion was the question — "Just why was conversation on App.net so good?" Was it the 256 character limit that afforded the use of pleasantries, reducing the likelihood a given conversation would become a bitter fight? Was it because conversations were threaded like email, and had a hierarchy to them? Was it that only certain people were attracted to the service? Maybe we were all elitists and we'd self-selected a bunch of agreeable people who could afford to pay fifty dollars for the privilege of casual banter? (Who knows. The conversations were great though.)

App.net (or as App.netizens called it, ADN) followed me through my life, growing as I grew, and the people I met there became real friends. I maintain close offline relationships with many of them. In a very real sense, the course of my existence was altered through participation in the network, and to myself and the rest of the community it became much more than just another silicon valley flight-of-fancy.

Perhaps around 2013, the community reached a tipping point — maybe twenty thousand regular users, maybe a hundred thousand (it's hard to know) where it stopped growing, and started shrinking. Many of the original participants declined to renew their accounts, and disappeared from the community — sometimes immolating in a virtual gasoline fire as they did so, but mostly they just faded away quietly, their mothballed profile pages a sad reminder of online friends that once were. In valley parlance, the platform churn could not support the ambition of the App.net company, and the founders, as well as everybody involved, strolled off into the sunset to leave the network as it lay.

Followed by a screaming exodus of series of vocal personalities, and lots of arguments about the future, the community mostly ran as normal for another two years — every now and again another person would drop out, and would be missed. I made it my business to stick around, even though I knew the writing was on the wall. I had — have — nowhere else to go.

Twitter wasn't the town square any more. It was a screaming urban junction of a size and busyness for which there is no suitable real world analogy. The constant vitriol and rage was so palpable you could probably bottle the froth and sell it as a stimulant. While I'd become accustomed to the friendly and familiar faces of App.net, Twitter felt both too constrained and too open. It wasn't eloquent, was developer hostile, and lacked the privacy App.net somehow afforded. Facebook was creepy and impersonal. None of the boutique networks like Ello seemed to offer what App.net did.

Even as App.net dies, instead of jetting off to another network to continue the revelry, I prefer to sit with my friends overlooking the dying platform and reminiscing, like the last representatives of a star system might cheer and clink glasses as they watch it go nova in the Restaurant At The End Of The Universe. It was beautiful while it lasted. All good things must come to an end.

For now, App.net is still lumbering along, but there's the undeniable sound of a death rattle. The news feeds I subscribed to and enjoyed have stopped. Image uploading has become unbearably slow. The SSL certificate for the whole service came within nine hours of expiration today.

Maybe it'll be another year, or maybe there are mere hours left. But while there are little splinters of the developer community working to create App.net's successor in the image of their own philosophy (Simpler! Decentralised! Open! Closed! More visual! Different interchange formats!) — and while I'll certainly stick around till the final moment, the network is on the way out.

It's sad, and especially upsetting to think that I might lose touch with people with whom I'd spent so much time, and from whom I learnt so much.

Watching the Hi.co service turn from a community scribbling board into a time capsule, to descend into an archive where it'll stand as a record of a loose band of humans who once laughed, and cried, and walked in the sun upon the surface of the Earth — reminded me of the mortality of our online services. I can only hope that we don't forget they were once outpourings of human emotion rather than simple tubes, mechanistically routing our communication without thought. They're what we built on top, and even in death, they're incredible.