webdev

Same features, a third of the code, and the back button works. Nobody has asked for the loading spinners back.

Same features, a third of the code, and the back button works. Nobody has asked for the loading spinners back.

Time to first byte was 180ms and everyone was proud of it. Time to a usable page was 4.1 seconds, and every millisecond of that gap was our own code.

Sixty components, four of which were a button. The tokens underneath were the part that actually held the product together, and they fit in one file.

Four weights, two of them unused, one of them loaded twice because the fallback stack disagreed with the preload. That was 380KB before a single word of content.

Two developers, one deploy artifact, and a bug count low enough that we stopped tracking it weekly. The part I did not expect was how much of the old code had been about the network.

Conversion moved two points and the support tickets about the back button went to zero. The old flow was one page and eleven client-side states.

The HTML page finished before our framework had finished parsing itself. Both of them showed the same eleven rows of text.

One extra DNS lookup, one extra TLS handshake, and a third party who can see every page your users visit, in exchange for a file that gzips to 6KB.

The assertion is a string comparison against the exact bytes the browser will receive. There is no renderer to mount and nothing to mock.

Ours broke for three weeks on one mobile browser and nobody noticed, because the failure mode was a button that did nothing rather than a page that did not load.