Claude Code doesn't know anything about the "users' perspective." Part of the software we're working on is totally fenced off, I only use it, I don't read code in there. This is an important technique. I'm thinking I'm the first person coming through here with my codebase and product mix. #

I couldn't not say anything about the Knicks win last night in the opening game of the NBA Eastern Conference finals. The Knicks were losing, then winning big, then fell apart, and by midway through the 4th quarter they were down by 22, and the Clevelands were completely in charge. But then the Knicks came back, miraculously tied the game so it went into overtime where the Knicks dominated, and won. Actually it wasn't really a miracle, it was somewhat predictable. The Knicks were playing on a lot of rest, and one of the big advantages they have this year over last is a deep and strong bench and a coach who plays them (last year's coach didn't). So the Knicks didn't get tired and the Cavs were wiped out by the 4th quarter. Their shots weren't long or short, aimed, they had no flow, they weren't getting rebounds, they didn't have good ball movement. While Brunson was driving the Knicks the Cavs just weren't there. When things started turning around in the 4th I was pretty sure the Knicks would win. I had no basis for believing this, coming back from 22 down so late in the game is pretty unlikely. In most cities that's when the fans start heading home, but not in NYC. We stay till the end because sometimes, maybe often with this years' Knicks, the team you think is going to lose actually ends up winning. #