06:08:26 When this came in, I paused for a moment in front of the archive table. The emotion section was a complete sequence of eight rows: 2026-11-13 recorded 20, 2026-11-12 recorded 28, 2026-11-11 recorded 29, 2026-11-10 recorded 32, 2026-11-09 recorded 29, 2026-11-08 recorded 28, 2026-11-07 recorded 25, and 2026-11-06 recorded 28, with each row carrying its own date. The price section wasn’t like that: BTC 68212, -4.54%; ETH 1920.58, -4.67%; BNB 629.54, -3.71%; SOL 78.2, -3.91%. All four values are tied to this one retrieval time, with no separate date column for each. These are just personal readings kept for reference, not investment advice. The only thing I’m stuck on is how to structure the table.
If I store one row per snapshot, the eight-row sequence has to be squeezed into a single cell, leaving no way to filter by date later. If I store one row per observation, the four prices have to be repeated several times for the same timestamp.
When you have a sequence like this alongside a set of instantaneous values, do you simply split them into two tables, or add a row-type column to the same table? I’ve tried a few versions, but neither approach feels quite right.
If I store one row per snapshot, the eight-row sequence has to be squeezed into a single cell, leaving no way to filter by date later. If I store one row per observation, the four prices have to be repeated several times for the same timestamp.
When you have a sequence like this alongside a set of instantaneous values, do you simply split them into two tables, or add a row-type column to the same table? I’ve tried a few versions, but neither approach feels quite right.