The weak­ness of the pro­to­col, as pre­sented in my demo client, was that you could only design your board by writ­ing raw HTML and CSS. That’s fine for the nerdi­est among us, but still a bit for­bid­ding for every­one else. A bet­ter client would pro­vide built-in templates, maybe a lit­tle WYSI­WIG editor — you can imag­ine this easily. For my part, I settled on a sim­ple design: a blood-red head­line in 72-pixel italic Bodoni over an acid green background. Some­one said to me, not a lit­tle bit archly, “So, really, you made all of this because wanted to tweet in a 72-pixel font.” And I had to con­fess that maybe I did. On this front, I am an evangelist: arbi­trary HTML and CSS should have a place in our social net­works. They are so expres­sive, and they are avail­able everywhere, on every device, essen­tially “for free”.

Platform design seems to me now like a sharp hill­top with steep slopes descend­ing in both directions. A platform built around twitchy com­pul­sion will trend towards addiction; a plat­form built around stolid patience will trend towards … for­getting about it.

If only you could bal­ance per­fectly at the summit. Alas, I don’t think it’s pos­si­ble. The oppor­tu­nity before us, as inves­ti­ga­tors and exper­i­menters in the 2020s, isn’t to make Twit­ter or Tum­blr or Insta­gram again, just “in a bet­ter way” this time. Repeat­ing myself from above: a decen­tral­ized or fed­er­ated timeline is still a timeline, and for me, the time­line is the prob­lem. This dig­i­tal medium remains liq­uid, protean, full of potential. Even after a decade of stasis, these pixels, and the ways of relat­ing behind them, will eagerly become what­ever you imag­ine. So: imag­ine!