181 karma · joined June 28, 2016
Every time I save an edit to a page in ⦗ℍ𝕪𝕡𝕖𝕣𝔱𝔢𝔵𝔱: h0p3⦘, two jsons are created in a subdirectory of `~/Downloads`, an old and new copy of the edited page. That's usually the only versioning I directly use. I'm finished editing the majority of pages within a week (except for retroactive footnotes, which are themselves dated upon injection), but there are some pages I've edited thousands of times; all of it is captured, and I can archive or delete these at my leisure. Sometimes, I'll walk through my nightly zstd archives (and I've not had to go through my waterfall backups across drives yet), and even more often just use web.archive.org (since it's all wrapped into one html file). I've always wanted extremely lightweight automated versioning (not what I already do by hand, as I sometimes explicitly link together versions of a page) built into my wiki itself (none of the plugins available really do it well, imho), and I may work toward that.
I do have git repos available, and I know some found it useful. Traditionally, I keep a 2-minutely updated git repo (reset nightly), as well as a nightly. Something has gone wrong when I use it, but perhaps that's a lack of practical creativity otherwise on my part. For those participating in the mutable torrent swarm, hidden default Resilio Sync archives have provided useful versions to myself and others as well.
I'm paranoic about keeping a record of my streams of thought and the objects I grow in my garden, and I often find the most important aspects of the versioning of my thinking (and feeling) are captured by just thoroughly using and living with the tool in the first place.
I could see federated networks perfecting some aspects of self-sovereign, decentralizable identities or portable logical accounts, as well as fitting standards for sharing the data we find most salient among that federation. It's a start. I can't say something like Mastadon has succeeded here, but these amazing projects simply haven't received the funding they deserve. I could also imagine a centralized, for-profit institution creating something similar, but it would be a trade secret. I hope we find the answer before they do.
I'm not going to work on that problem until I have an underlying p2p network meant to actually replace at least the web (and, I am, of course, a huge fan of Sir pfrazee's astounding work in this sphere). I consider that an extremely tall task as it is, and, so far (and I say this with profound [[gratitude]] for those who have worked to solve the problem), nothing comes even close.
It's a pleasure to meet you, btw.