Don't know it till you see it

You can't tell a joke is funny until you hear it. You can't describe the funny qualities and work backwards to rate it.

Same for design: you know it when you see it. I need to see dozens of designs, so I can tease out what elements make sense to me. What feels right.

For Instacalc. I've tried for years to find a good, clear, cohesive design. Taking things away until only the core elements are left.

---

Expand, then contract

Make too much code

See what works

Collapse it down to that core component

Throw away the rest

Refactor that core component into a gem

keep going

I need to expand to see what's possible, then reduce it to what I need, then make it work better, then instrument and test that core inner piece.

---

I wanted to get rid of the top bar. I did. But it's too useful:

We need global nav. When we shift from "build" to "explore/discover" mode, we need to know where we are. and how to get back. navigating everything by the sidebar is too hard. The hierarchy (top down), what controls what. It's much cleaner this way.

I tried to do sidebar only, I did. But couldn't do it. The original instacalc had a top bar. So did the various permutations. I couldn't stay away...

---

Another thought: share view vs "build" view

when you are editing a calc, it looks a certain way. but when you share it with someone else, it should be the hero.

the title, bottom tagline, other things -- might be different. we want the best DISPLAY version of it. When editing you want the best BUILD version of it.