Restarting Blogging 3rd Time , What's New This Time
Every framework decays — that's the premise, not the problem. The problem is that a how-to has exactly the shelf life of the version it was written against, and I wrote almost nothing else. Three in five of mine were for a framework that's been discontinued; the most-read one is for a framework that's still thriving, and it's stale anyway. This blog writes at the layer that outlives the tooling: what I built, what broke, what it cost — including where I was wrong.
Three in five of the posts I wrote between 2016 and 2020 were about Xamarin.Forms. That framework has been discontinued. If I stopped there, the lesson would be that I backed the wrong horse, and that lesson is worthless, because everyone backs the wrong horse eventually. Frameworks decay. .NET ships an LTS, then an STS, then another LTS. Blazor's hosting model got rewritten. Nobody writes about tooling that stands still.
So framework decay isn't the complaint. It's the premise.
The part that actually indicts my old writing is harder to dodge. My most-read post ever is a Blazor how-to from 2020. Blazor is alive and I'm still using it: this site is a Blazor Server application, and the blog engine rendering this post is one I built on it. Same framework, still healthy, still my daily driver. That post is stale anyway, because the component and hosting models moved underneath it.
It didn't need the framework to die. It just needed the next version.
That's the real shelf life of a how-to. Not "until the framework is discontinued" but "until the version it was written against stops being the version people run." Which is about a year.
Instructions versus reasoning
What I wrote was instructions. Configure this. Wire that up. Here are the eleven steps. Instructions are coupled to the tool the way a screenshot is coupled to a UI. Change the thing and the artifact isn't dated, it's wrong.
What I mostly didn't write was reasoning. Why this architecture and not the other one. What I expected, what actually happened, what it cost me to find out. That kind of writing tracks the problem rather than the tool, and problems have much longer half-lives. Offline sync, state management, verifying code you didn't write, the cost of a token. None of those got solved by a version bump. All of them are still open.
The archive did fine on attention. High six-figure view counts across a decade. What it didn't do was compound. There's no position in it, nothing anyone could disagree with, nothing that makes a case for how I think.
Why it stopped
The database those posts lived on is gone. That isn't the interesting part, since the writing survives elsewhere and I'll link it at the bottom. The interesting part is that I stopped in 2020, years before anything was lost. The archive isn't why I went quiet, so I don't get to use it as the excuse.
I stopped because the writing was a second job stacked on the first one, and I wasn't enjoying it. I tried other formats afterwards and dropped those for the same reason.
I've stopped reading that as a discipline problem. Writing tutorials about a tool is homework. It's work that exists only because you decided to publish, and when the week gets tight it's the first thing to go, correctly. No amount of willpower changes that arithmetic.
Then the format stopped mattering
Comments (0)
No comments yet
Be the first to share your thoughts on this post.
Leave a comment
No account needed — just your name and email.