Julia asked: why not OBS?

The plan was set: build an app to sync screen recordings with the camera footage, so the videos can finally show the code being discussed. Estimated cost, about three days. Then, walking the dog, Julia asked the question: why don’t you just use OBS? ▸ 0:50

The answer took one hour, not three days: OBS installed, Iriun Webcam turning the phone into a camera (after a detour through a fake clone app with a copied logo), and the new setup working on camera in this very video, sharing screen with no camerographer needed ▸ 2:31.

The accounting, done honestly at the end: “the setup didn’t cost much time to make; what cost time was thinking it, I was about to spend three days building a program that wasn’t even useful” ▸ 21:31. Worse than useless, actually: the homemade version would have added a half-hour transfer step to every future video, a small recurring tax the off-the-shelf tool simply doesn’t charge ▸ 21:47.

una pregunta en el paseo del perro > tres días de código →

This closes a trilogy the last four days wrote by accident. Entry 38-1: when no tool fits, build it, and that instinct just produced a real product. Entry 39-1: the same instinct, unsupervised, spawns projects faster than it finishes them. And now the checkpoint that keeps the instinct honest: before building, let someone outside the obsession ask whether it already exists. The layout editor survived that question (nothing affordable did what it does). The recorder didn’t survive it, and dying cheap in one question is exactly what the checkpoint is for.

The general rule: “I’ll build it myself” should be a conclusion, never a reflex. It costs one dog walk to check…