Recently I’ve been thinking about “digital gardens”, which resemble a public, curated knowledge base or a wiki. Or maybe public, constantly changing notes. Some best examples are gwern.net by Gwern Branwen, Everything I know by Nikita Voloboev and Andy’s working notes by Andy Matuschak. Their common denominator is that they grow over the time, similar to gardens. It’s up to the gardener how will it look like.
Digital gardens naturally lean towards the concept of wikis or knowledge bases. I already have one knowledge base: my personal notes, so understandably I don’t want to repeat myself or copy-paste my notes here. I discovered, however, that continuous refinement of thoughts is beneficial and moving the notes to the blog gives them a structure which they didn’t have earlier.
Digital notes aren’t the only source of new articles. I have two other methods of storing my thoughts for future use, because I can’t imagine using any form of keyboard to quickly jot down something about a book I’m reading:
- I use a paper: I have several notebooks, each used on a different occasion.
- I record myself and leave the recording as a note to myself on Signal2 for hassle-free access on my other devices.
I collect these notes later and decide where I should put them: into my garden or knowledge base. The decision is usually straightforward, because it usually had already been made.
In the process I don’t simply copy my exact words I think them through, edit them, add a lot of context to them, sometimes remove the whole notes, because they turned out unnecessary or untrue.
These are newborn branches of my digital garden, which must be looked after with a special care. Readers, however, must be also aware that they are not looking at the final form of the article. For that, they have a special status taxonomy1.
- draft - an early stage of most of articles, usually an outline or a bunch of unfinished or poorly edited paragraphs. Articles might change drastically. Thoughts and ideas require additional verification. These pages are not included in the list of latest articles and in RSS feed.
- in progress - more-less coherent page, which you’ll be able to actually read and understand something. This step is sometimes called editing (I can’t afford hiring a professional editor, so I must do it myself). It’s not the final form, but pretty close. This status also includes always-unfinished pages.
- finished - finished piece. It doesn’t mean that it won’t receive any updates, but that I’m not actively working on it.
- rework - articles once finished, which need a big rework for some reason. Sometimes they’re outdated, sometimes they’re written in Polish and I intend to translate them soon.
Good gardeners know when to cut sloppy shoots. Removing and merging some of the older articles is necessary. Old news aren’t news anymore, some mini-series were discontinued, but look quite good when merged together (see: 1, 2). Or I’m inventing a new way to categorize things.
Unfortunately, moving/removing content breaks permalinks and bookmarks, assuming that someone have bookmarked my page.
The rationale for breaking permalinks is that I have lost hope in the web as a medium to preserve things. Everything on the web is ephemeral and should be treated as such. Projects such as Wayback Machine or archive.today do reasonably good job archiving web pages, but they are not silver bullets. They belong to third parties and these, by definition, cannot be trusted with your most precious personal data. Browsers should be doing a better job bookmarking web pages, but they don’t.
This leaves us in a place where we have to save everything ourselves or be ready for the loss. I simply acknowledge this fact. I’m trying to preserve some content, but I’m not making any promises. I’m still thinking how to backup data after my death.
Anyway, what’s the point of preserving permalinks when content of pages isn’t set in stone (see previous section)?
Linking between articles is crucial for creating a site which resembles garden. I do it manually on a best-effort-basis. I think that generated connections between pages are soulless and only intended to keep readers longer on a website. I don’t monetize my thoughts so there’s no reason to waste time readers: if there’s nothing related to the topic they’re browsing, so be it.
It’s very imperfect and cumberstone process, but that’s the part of gardener’s life, isn’t it?
Additional to hyperlinks, pages on similar topic are grouped together on the main page and with more fine-grained tags.
On the main page there is a special section which holds articles which I find the most important. Articles on this list are sorted by relevancy: a coefficient which is calculated based on few other factors:
- article importance, in scale 0-10, which I set manually;
- the article’s age: newer articles with the same importance are more relevant than the older ones, up to some certain point in the past;
- existence or lack of certain categories or tags (for example, articles written in Polish are less relevant to the general audience).