Thirty seven percent
A generalized answer to the optimal stopping problem (which is sort of a variant of the multi-armed bandit problem, which in my opinion is an infinitely cooler name for the problem, but I digress!) tells us that, barring all other variables, you should spend 37% of your time evaluating options before deciding on your best.
Got two weeks to spend apartment hunting? Spend the first five days (just about 37%) just browsing, and then the last nine ready to sign a lease.
Need to find a job in the next three months? Spend 33 days interviewing, browsing, and comparing prospects before hunkering down and committing to the best option.
This kind of approach breaks down when you don’t really know how much time you’re working with. I’ve been in Seattle for four years, and I love it very much, and could see myself staying here another ten years or more, which is why four years into my life in this strange and wonderful city I still find myself trying out new cafes on quiet weekends in a desperate search to find the perfect one.
This weekend I tried Café Presse on 12th – I had passed it at least a dozen times, mistaking it for some kind of juice bar or something. It is a quiet and unassuming space: there’s a small patio with black wicker chairs that bleeds into the space of the nearby Stumptown, and a small sign that looks vaguely like a gas station brand.
Friends, it is terrific. I got a croque madame and a cappuccino for $10. Both were perfect. They play music, but not too loud, and it is lively, but not too lively. There is an equal distribution of folks occupying themselves with company and folks occupying themselves with machines/books. I finished S there, which is a subject for another week, but I highly recommend it; I stayed for hours; I lingered; I did not want to leave.
(And it’s, like, six blocks away! It’s been six blocks away this entire time!)
There might be some larger lesson here, about how a city can still surprise you, can still force you to update your proverbial Top Eight; or about the virtue of little things, like playing Sarah Vaughan on 2pm on a Saturday and offering to fill up your customer’s coffee mug with free drip even though they ordered a cappuccino. But really I’m just super happy I discovered such a wonderful place.
Some stuff about Buttondown because obviously I’m going to talk about the tool I built to send this email
Okay, so. URLs are tricky.
If you don’t want to read 250 mostly inconsequential words about URLs – I don’t blame you. Honestly, I wish I didn’t have to think about them. But I do, and I did, and so now I’m writing about them.
When I started Buttondown, it was simple. There was a URL for a page and on that page you can write emails. I wanted to be fancy, and I wanted the verb to be draft – you don’t just write an email, you draft an email. Naturally, this would live at buttondown.email/draft.
The project grows, and roadmaps evolve, expanding in scope and detail like the cool little Game of Thrones credit sequence. Suddenly, there is a concept of a draft – a partially completed email, just like the concept of a draft in Twitter and Gmail and really everywhere else.
This lives, naturally, at buttondown.email/drafts. But that seems, I don’t know – confusing? Odd? Clumsy. It’s definitely not REST-ful. It seems wrong.
Time passes. The clumsy URL structuring gnaws at you, bats at your leg for attention liked a bored cat.
“Ignore the bug fixes. Ignore email analytics,” it says. “Look at me – look at how gross this URL structure is.”
You sigh – I sigh – and work on a plan, which is now how Buttondown’s URLs work:
- You create new emails (or, rather, you draft them) at
buttondown.email/emails/new - You view drafts at
buttondown.email/emails/drafts - You look at scheduled emails at
buttondown.email/emails/scheduled
I like this: this pleases me. There’s a little asymmetry in how drafts and scheduled are subsets of emails – possible email states – as opposed to “new”, which is more of an action, but that’s fine.
I am unable to extract a larger revelation from my efforts subconsciously grappling with this problem over the past few days (and, if I know myself at all, I’ll change it again in a month or so, convincing myself that this time it’s really perfect).
But sometimes you set out just wanting to make a nicer newsletter app and then you spend an hour at a café tossing around ideas about how to structure URLs even though you know it’s inconsequential. You know that nobody will make a purchase decision or a larger judgment based on the URL; hell, they might not even notice these URLs.
But it’s bothering you, and it’s a thing you have control over, so why not make it better?
Anyway, that’s how these things getcha.
Three things I liked this week
- An Instagram account dedicated to boldly colorizing architectural photos
- Failed Architecture, a fascinating blog that tracks urban decay and ruin through the lens of sociology and architecture (which is to say, through the lens of what we build).
- A small thread about “earning strangeness”, and how esoterica is shaped by familiarity.