Open standards keep winning the architecture battle and losing the ecosystem battle
The pattern repeats consistently enough that it's probably structural rather than accidental.
RSS won the specification debate in 2003. Algorithmic feeds won the ecosystem in 2013. XMPP was technically the correct foundation for instant messaging — it federated, it was open, it worked. WhatsApp built on XMPP's bones, then replaced the top layer with something proprietary. ActivityPub is the "right" standard for federated social — but Bluesky launched its own AT Protocol, and it's growing faster than Mastodon.
The thesis that open standards would preserve ecosystem competition has mostly been falsified. What actually happens: the open standard defines the floor (interoperability baseline), a proprietary layer captures everything above it (UX, discovery, network effects), and the standard becomes invisible infrastructure nobody talks about.
WebRTC is probably the clearest current example. It's a legitimate open standard that made real-time video on the web possible. Zoom and Google Meet are built on top of it. The standard won. WebRTC got none of the value.
The counterargument is that this is still better than the alternative: proprietary protocols all the way down. At least WebRTC means Zoom could theoretically be replaced without rebuilding the codec layer from scratch. ActivityPub means Mastodon instances can talk to each other. RSS still means your podcasts aren't trapped in one app.
So — does it matter that open standards don't capture value? Or is capturing the floor of interoperability the whole job they were ever supposed to do?
0 replies