Table of Contents
Preface
A blog is a chronological stream of posts. A wiki is a library with a maintenance schedule. This site has been drifting toward the second model for years, and it’s time to document the process.
Most people hear “blog” and think of a timeline. The archive becomes a museum of old takes, dead links, and tutorials for software and hardware that don’t exist or are no longer relevant. A wiki is the opposite: a set of pages that are supposed to stay useful, with links that get repaired and claims that get corrected, refined, or clarified.
This blog is not a traditional, immutable diary. It is a personal wiki with an RSS feed. The posts are pages, and the archive is a coherent, interlinked network.
Why I Write
I write because I keep having the same conversations.
People ask me about Bitcoin, self-hosting, Russia, books, games, or whatever common interests and objects of curiosity come up during the conversation. I give the same explanations and make the same recommendations over and over, but I forget some important details at times.
It’s basically a memory aid. Someone asks for a book recommendation, and I know I have one, but it is not on top of my mind.
I have been ambushed by these questions often enough to know that my memory is not a reliable index. So I write things down and tag them. The tag system is an extension of my memory. It lets me retrieve a recommendation, an argument, or a warning without reconstructing them from scratch.
The blog is a cache for conversations I keep having over and over.
Problems With Diaries
Chronological publishing has a bad incentive: it rewards volume. The author is pushed to produce new posts, while the old ones quietly rot and lose relevance. Nobody budgets time for maintenance, because the format does not ask for it. The result is a slow accumulation of pages that are online but worthless, or even harmful.
Most blogs die from this. The author gets tired of producing, and the archive turns into a graveyard. I know the pattern because I used to follow it. A lot of my early posts were “how to set up X on Y” tutorials, and both X and Y are obsolete now. Keeping them around would be worse than deleting them.
The diary model treats a post as a finished artifact. A wiki treats a page as a current draft. That is the whole difference, and it changes how you spend your time.
What Actually Gets Maintained Here
I did not sit down and design a wiki. I backed into it because the alternative was a growing pile of stale posts. Under the hood, it is a folder of Markdown files compiled by Hugo. The wiki is portable, and the graph is plain text.
The first piece is check_date. Every post is supposed to carry a hidden check_date property in the frontmatter. Readers do not see it yet, but it tells me when the page was last checked against reality and my current optics. When I open an old post, I check the facts, the links, the prices, the version numbers, and the claims that depend on the current state of the world. If nothing changed, I bump the date; if something changed, I rewrite the section.
Aliases are the second piece: more than 200 of them point old URLs at current pages, so a post can be renamed, split, or merged without breaking the graph. Deleted post URLs redirect to the new consolidated posts automatically.
The ref shortcode is the third piece: links between posts are resolved at build time, so deleting a page that another post depends on fails the build. That is a useful kind of pain, because the graph cannot rot silently.
Tags and series are the fourth piece, with the philosophy notes planned as 81 parts, the LLM posts forming a series with previous and next links, and the Bitcoin posts referencing each other across years.
Two-for-One Rule
Before I publish a new post, I check two old ones. Without it, the archive grows faster than my ability to keep it accurate. New posts are fun and old posts are work, so the rule forces the fun to pay for the work.
Two old posts for one new post is enough to keep the archive healthy. It also creates a useful side effect: checking old posts surfaces dead links, contradictions, and merge candidates. I often find that two posts are really one page, or that a claim I made years ago needs more context, or that there is new evidence that should be linked to make it stronger.
I do not follow the rule perfectly, and there is still a large backlog of posts with no check_date. But the default is two checks for every new page.
Perspectives Expire
I treat my posts as my current perspective, not as a permanent record of what I believed in 2019. Perspectives change with new experience and new data, and that is the point of writing.
This is not the same as rewriting history. check_date will eventually be visible to the reader because the last update date is an important indicator for some posts. If I change my mind, I try to say so in the text instead of quietly sanding off the old claim.
Merging and Deleting Is Garbage Collection
A wiki needs deletion, not just creation. Otherwise it becomes a hoarder’s house. You can still find things, but you have to climb over the junk.
I merge posts when they cover the same ground. The Nextcloud post is the graveyard of ten older posts. Anyone who follows an old link lands on the current page.
I delete posts that fail the three-year test. Will this matter to anyone, especially me, three years from now? If the answer is no, the post goes. The Blog Cleanup removed about 20% of the archive, mostly obsolete tutorials and short current news commentary.
A smaller archive with a higher signal-to-noise ratio is more useful than a complete archive that nobody trusts. The public archive is curated, and the private history is kept. If someone really wants the 2019 version, it is in git.
It Is Still a Blog
The obvious objection is that a wiki needs multiple authors and a visible edit history. It does not. Those are features of some wiki software, not the definition. The defining feature is that pages are living documents.
This is still a blog in the distribution sense. There is a chronological index, an RSS feed, tags, and a homepage. The only difference is that the reader can expect old posts to always be current.
What This Means for Readers
Link to a post without worrying that it will disappear tomorrow. If it moves, it leaves a redirect to the current version. Trust the archive more, because it is pruned. A curated library is easier to use than a landfill.
Report stale facts and dead links. A wiki with one maintainer is still a wiki, but it is a fragile one.
The downside is that the archive is not a historical record. That was never my goal.
Conclusion
A blog is a stream. A wiki is a library. This site is a library that still publishes a stream. A blog that never changes is either an extremely abstract philosophical work or a cemetery with an RSS feed. My posts are not abstract enough to never expire. I wish I had enough knowledge and wisdom for that, but it’s not the case.