🪀 Didn’t ship this one because it was too much… fun

🔌 Empty states deserve some love too

Sorry for the delay. 🫶 These are great questions. For this one: - I designed the table almost a year ago, but that version was static. This one was pretty much pure code + instructions + a lot of iteration locally. - Now that I tinker directly in prod, I feel like I have way more freedom to experiment. My first goal is always to ship something simple and not hold things back trying to make them perfect. My process is basically: Build → Ship → Polish → Ship → Make it genuinely pleasant → Ship. A lot of the details only reveal themselves once I can actually use the thing. - The responsive part is mostly rules. At certain widths, specific pieces separate into their own columns. I define where they came from, where they’re going, what stays fixed, what happens with missing data, long names, different combinations of data, etc. Then the motion connects those states. It also follows the speed of the resize instead of playing the same fixed animation every time. - For the motion itself, I try to make things feel physically related. The magnet-like behavior here is one example. If something separates from another piece of data, it should feel like it came from there. If two things belong together, they should pull back toward each other. Here I wanted snaps, not bounce, because it’s still a table and needs to feel precise. But during exploration, it’s mostly fuck around and find out. Once I like something, I start building the logic and rules around it. Then it’s a lot of tiny tuning: thresholds, stagger, easing, timing, how much something gives before it snaps, and testing it at different resize speeds. And of course, reduced motion is part of the rules too.

🔍 Zoomed-in, slo-mo details no one would ever see It all adds up.

🫨 Next call in 15m? Don’t worry, we’ve got you. The meeting row expands into a card with the details and quick actions, right when you need them.

🧲 Follow-up to the dynamic tables: multi-select with mouse or keyboard

🧭 A few weeks ago, I was exploring a few different nav layouts. Navigation feels even more crucial in AI-native apps now. It’s surprisingly hard to figure out what should be separate, what should be grouped, and what deserves to be one click away. Haven’t landed anywhere near yet since we’re still too early, so sticking with what we have rn. Mostly just scratching an itch and exploring.