For a long time, publishing online mostly meant putting our work where people already were.

That made sense.

Facebook had the audience. Google helped people find things. Email reached subscribers directly. Other services handled jobs that would have been expensive or impractical for a small newspaper to reproduce ourselves.

We still use outside services, and I expect we always will.

The problem wasn't that they belonged to someone else.

The problem was expecting them to carry things they were never designed to carry for us.

A social network can be an excellent place to discover a story. It isn't necessarily where we want the only durable copy of that story to live. A search engine can help someone find a local business without becoming the permanent record of that business. An email provider can deliver a message without becoming the institution sending it.

Those distinctions took me a while to appreciate.

When a newspaper publishes primarily through other platforms, its relationships gradually become difficult to separate from the mechanisms used to reach them. Readers become followers. Businesses become impressions. Articles become posts moving through feeds whose useful life may be measured in hours.

None of those things is inherently bad.

They're simply different things.

A newspaper has another job.

Someone should be able to find the story six months later.

A business that has been part of the community for twenty years shouldn't disappear from our digital world because its advertising campaign ended Tuesday.

A church should be able to correct its directory information.

An event should exist somewhere long enough for people, search engines and calendars to find it.

A subscriber should have a direct relationship with the newspaper rather than requiring some other platform to recognize that relationship first.

So we've been moving more of that work onto our own domain and infrastructure.

From the reader's perspective, much of the change is deliberately unremarkable.

There are articles.

There are subscriptions.

There are community calendars, directories and archives.

Businesses have profiles that can remain useful after a particular advertisement has run. Print can point toward information online. Something published online can later become part of the printed newspaper. Radio, social media, search and email can help people discover what we're doing without any one of them having to contain the whole thing.

That's the part that matters to me.

The website doesn't replace the newspaper.

It gives the newspaper another place to exist.

Print remains useful for things print does unusually well. Social media remains useful for reaching people who happen to be there. Search helps someone looking for a particular thing. Radio reaches people in another context entirely.

Our own infrastructure does something different.

It gives all of those paths somewhere to return to.

That changes advertising too.

A print advertisement is necessarily temporary. It appears in an edition, reaches that week's readers and then becomes part of the archive.

There is nothing wrong with that. That's how newspapers have worked for a very long time.

Online, we can add another layer.

A local business can have a durable profile containing the information someone actually needs. An advertisement can point toward it. A social post can point toward it. Search can find it. The business can update it when something changes.

The advertisement still matters.

It just no longer has to carry the entire relationship.

The same thing happens with community organizations. A church, fire department, library or civic group doesn't need to become a continuous stream of posts merely to remain findable. Its basic presence can persist quietly until somebody needs it.

That sounds like a small distinction.

At local scale, I think it's an important one.

Most of community life isn't newsworthy every day.

The hardware store is still there on the days when nothing interesting happens at the hardware store. The church still exists between events. The town board remains part of local government between meetings. A restaurant doesn't cease to matter because it hasn't produced something sufficiently entertaining for a social feed this afternoon.

A useful local information system should be able to remember boring things.

Maybe especially boring things.

This is also why we've invested in operating more of the underlying infrastructure ourselves.

Ownership gives us options.

We can decide how long something remains available. We can change how different parts of the newspaper connect. We can preserve archives without needing them to satisfy someone else's product strategy. We can build around the peculiar requirements of the communities we actually serve.

Ownership also gives us responsibility.

If our server fails, that's our problem.

If our software is badly designed, we don't get to blame an algorithm.

If we collect information we don't need, operating the database ourselves doesn't somehow make that good practice.

Running infrastructure locally doesn't make us virtuous. It makes certain decisions ours.

That's useful because some of those decisions matter to what a newspaper is trying to preserve.

Other decisions don't need to be ours.

We don't manufacture our own printing press. We don't build the cellular network. We don't need to create our own payment rails, search engine or social network merely because those services exist outside the newspaper.

The question we've gradually learned to ask is simpler.

What does the newspaper need to be able to do for itself?

For us, one answer is maintaining the durable relationships underneath the things we publish.

The article belongs somewhere.

The author belongs somewhere.

The business belongs somewhere.

The event belongs somewhere.

The subscriber has an account somewhere.

Those things can then be projected into whatever forms are useful: a printed page, a website, an email, a social post, a search result or something we haven't built yet.

The forms will keep changing.

That's fine.

We've already watched several generations of the web arrive and disappear. There will be more.

The mistake would be requiring every relationship to disappear with the form through which someone happened to encounter it.

This doesn't make our own website permanent either.

Software changes. Servers get replaced. Domains can move. Organizations themselves can fail.

We're not trying to create something immune to change.

We're trying to keep enough of the newspaper under our own stewardship that we can decide how it changes.

Existing subscribers will gradually receive access to the digital parts of that newspaper. Businesses and community organizations can maintain useful information alongside the stories and advertisements that already connect them to readers. Archives can accumulate instead of being scattered across platforms. Print can lead into the web, and the web can feed things back toward print.

None of this requires abandoning the larger internet.

Quite the opposite.

We can use it more freely when we aren't asking every service on it to preserve something it was never responsible for preserving.

Facebook can help someone encounter us.

Google can help someone find us.

Email can bring something directly to a reader.

Radio can tell someone what's happening.

Print can sit on a kitchen table for a week.

And after all those different encounters, there should be a place to come back to.

That's what we're building.