The containers of yesterday and the content of today

There are almost no logical places to put a 60-inch tv in a craftsman foursquare house.

The house is built for light and air and an era where the home entertainment center was a piano and a victrola. Where meals were eaten at a dining table and gas lighting and coal boilers were the order of the day.

But we have a big tv that we watch together, and an electric recliner, and every bedroom has a computer putting out heat. Getting our stuff to fit in this space has taken a lot of ingenuity and is an ongoing work in progress. The built-in buffet drawers are filled with DVDs instead of table linens.

Look, we only have so much china, but a lot of DVDs

Technical content here

I’ve been listening to people talk about software containers for, y’know, over a decade. The idea of a container is an abstraction that’s handy for us. We can think of a bounded box instead of a namespace in a filesystem. But containers are not actually boxes, and they don’t fit everywhere.

Sometimes a container is more than you need for what you’re packing. You don’t need or want to put a china figurine in a box big enough for a lampshade. The figurine will rattle around dangerously, and the box will be fragile and prone to crushing, and you wouldn’t be able to fit all the boxes in a truck to move them.

It’s so dangerously easy for us to over-engineer our solutions that it’s almost hard to talk about anymore. It’s a standing joke that developers will write blog posts, once they got done building an entire static website to do so. Meanwhile, WordPress is right here. That’s a lot of effort in packaging so that we don’t have to consider the content.

But secretly marketing

When I see a technical solution (containers, web3, ML/AI) promoted as a universal solution, I admire the ambition of the people selling it. But I also wince.

Nothing is going to fit if we try to extend it to do everything. Marketing isn’t just about selling something, it’s about selling the right thing to the people who need it. I mean, it’s impressive if you can sell people something they don’t need, but they’re never going to trust you again.

I’d rather tell someone that my product isn’t what they need now, and build trust and a relationship, than have them buy the wrong thing. I’d rather refer someone to a competitor or a free product if that’s what’s right for them. I want to be in technology for a long time, and products come and go, but personal reputation is pretty persistent.

Back to the house

Buying a house is a commitment to long-term thinking. You want to do things right, because you’re the one who will probably be living with the consequences. Given our average tenure in a tech job, it sometimes feels like we’re renting, and we’ll be gone again in a couple years, but that’s an illusion. We always have to live here, with ourselves. How we arrange our stuff, do our work, and relate to others, it’s all long-term.