ThePtacek Ledger.

I did not build pages over a dial-up connection. I have only read about that era and looked at what survived it. But you can learn a lot from what a constraint produces, and the old constraint produced pages that got to the point with unusual speed.

When every kilobyte had a cost you could hear, the cost showed up in the editing. You did not ship a decorative hero image that pushed the actual content below the fold, because the image would have taken thirty seconds to arrive and the reader would have left. You did not load six analytics scripts, because there was no room. The medium enforced an economy that we now have to enforce on ourselves, voluntarily, against every incentive.

The constraint did the editing. When it lifted, nobody replaced the editor.

What replaced the constraint was capability. Connections got fast enough that a heavy page still renders in a second on a good laptop in a good city — and the failure mode moved somewhere you cannot see from your desk. It moved to a four-year-old phone on a bad signal, which is where a very large share of the actual web is read.

So I built this site the way I did on purpose: hand-written HTML, one stylesheet, one small script that adds a clock, a menu and some scroll reveals and then gets out of the way. No framework, no build step, no package manifest. Turn JavaScript off and every page still works, because nothing structural depends on it.

That is not a moral position and I am not claiming it scales to every product. It is a preference with a reason: I want the thing I ship to behave the same for a stranger on a train as it does for me on a fast connection with a warm cache. The old web got that for free. We have to choose it, every time, and choosing it means saying no to things that would look great in a screenshot.

The reward is a page that loads before you finish deciding whether you wanted it. That still feels like the best trick the medium has.

— Ends

Skiing is the thing I would keep if I had to give up everything else on this page. Not because I am remarkable at it — I am a Midwesterner, our vertical is measured in polite numbers — but because it is the only activity I know that punishes hesitation and overthinking in exactly equal measure.

The first chair is the purest version of that. The snow has been groomed overnight and nobody has been on it. For about one run, the mountain is a surface rather than a negotiation: no ruts, no ice scraped bare by two hundred people finding the same line, no one stopped just past a blind roll. You get to find out what you can actually do, without the terrain arguing back.

Hesitation and overthinking, punished in exactly equal measure.

Everything after that first hour is a different sport. It is a good sport — reading a chewed-up slope is its own skill, and the people who ski well at two in the afternoon are better than the people who only ski well at nine. But the early run is the one that tells you the truth, because there is nothing to blame.

The transfer to everything else is not subtle. A blank slope is a blank file. You have prepared, or you have not; you commit down the fall line, or you skid across it hoping to postpone the decision, and skidding always feels safer and always ends worse. The mountain will not let you research your way out of the turn you are already in.

What I take from it, and try to take to the desk: start before conditions are ideal, commit earlier than feels comfortable, and accept that the surface is going to get worse as the day goes on and other people show up. Get your good run in while it is quiet.

Also, and I want to be clear that this is the real reason: it is enormously fun, and the coffee at the bottom afterwards is the best coffee available anywhere on earth.

— Ends

Minnesota advertises its lakes by quantity, which is a very Minnesotan way to undersell something. The number is a fact about the map. What actually matters is the specific one you end up at repeatedly, and the fact that it does not require anything from you.

I build things for a living and for fun, which means most of my hours are spent inside a loop of my own construction: notice a problem, decide on a fix, implement, discover the fix has a shape you did not expect, repeat. It is a good loop. It is also extremely difficult to exit voluntarily, because there is always one more obviously-worth-doing adjustment, and the adjustment is always small enough to justify.

A lake does not have a next step. That is the whole feature.

A lake does not have a next step. There is no version of sitting at the end of a dock that ships better if you concentrate. That is not a limitation of the activity — it is the entire feature, and it is the only reliable off-switch I have found for the loop.

The winter version works too, and is less photographed. A frozen lake in February is an enormous flat nothing with wind on it, and it produces the same effect through completely different means: the cold is loud enough that you cannot hold a train of thought about a roadmap.

Building from the middle of the map has its costs. You are further from the rooms where things are announced, and the nearest mountain is a plane ticket. But there is a compensating advantage nobody puts in a recruiting deck: it is very easy here to close the laptop and be somewhere that does not care what you shipped this week. That turns out to be worth a great deal, and the state gives it away free, ten thousand times over.

— Ends