the (frequently enough) project archive

Normally you shouldn’t access this page; but since you’re here, let me explain a bit what goes on in my head. :)

My personal archive had been a million git-annex pointers (and growing!) living next to a collection of scripts in archive/* repositories since early 2024. The name was pretty striaghtforward: archive for the sake of preserving the past and the memories thereof. (and stop letting Google and Tencent have access to my entire childhood likeness (gaaaaah i’m so embarrassed. ///n///))

Despite the shared top-level directory, there has never been a collective “project archive” initiative? I would mutter that to myself on multiple occasions, but that doesn’t legitimize it in and of itself. I would flee the classroom with my MacBook to the library, attached to the power brick on my cube power strip, one old HDD with all the family photos, and one new HDD holding the replicated and checksummed files from the old one. As I was struggling to shape the identity of rapidcow.org, I was also rapidly moving data from collapsing points of failure that I was completely previously oblivious to, but couldn’t be clearer in hindsight. if I had just been a bit faster, a bit…

“Project archive” is still just a tongue-in-cheek though, eh. Besides, diary would be the only thing I was worried about three years ago, and I named that Project Perspective, after the journaling app I used itself DX… but most of it came from me getting tired of writing things on my phone as my writing exploded into long LISP-ian paragraphs of mess; and it made my thumb sore so much typing for hours long I wish I had a desktop app. That discomfort, combined with the fact that Perspective, the app stopped receiving updates and thus a desktop version is unlikely to ever happen, led to my backup project.

Even if you feel comfortable with your software’s ecosystem, comfort should not be the reason for inaction, I believe. I have been reflecting the path I chose in a relentless race against time, against software obsolescence; like packing up as much as you can from a burning house slowly drowning in the throes of fire… but this is not really meant to be. Not barring the physical longevity of digital medium itself, data have none of the temporal qualities: they don’t degrade over time, they are cheap to replicate, and deterministic hashing makes sense. (I still can’t get over how confident we can be about the contents of a file, down to each bit, just from a fixed length of hash!) The urgency of data rescue (minus physical damage and degradation) comes from the longevity of software that interprets its semantics.

Software, unlike the files they operate on, require regular tending and care—like a garden. They need to evolve as we upgrade to new hardware supported by our PC manufacturer, which demand new operating systems (this is especially true on Apple devices, where the OS version is not downgradable, and App Store developers set a minimum OS version, if you try to re-install on a device you have never obtained the older version for). Clearly, software doesn’t evolve on their own, so when the software becomes obsolete, so does the file.

Thus the integrity of a file matters as much as the software that can open it, even if some go obsolete over time. Text files basically live forever, because any decent text editor in the world can open it. You can open it with any programming language capable of I/O, then you just have to decode the raw bytes into Unicode characters given it is a well-known encoding. (With UTF-8, you don’t even have to know how to decode most of the time!) Code, which is based in text, also benefit from these properties. (It’s hard to say it still runs, but it certainly is as tangible as data gets—as words that bear meaning, as linear runs of pure unadulterated text.)

Applications out there in the real world™, which work with data far more complex than linear plain text, don’t fit well into this description. You’d think that instills a moral responsibility in them to maintain an access mechanism that outlives their ecosystem, and a level of transparency about which aspects of your work they store to ensure that you maintain access to what’s yours in a lossless, editable format, and I’d agree so too. But too often the industry is caught up with the temptation of progress by drawing new users in with shiny new features, and of profit by hindering access to your data behind a paywall or a service which is pinned to their server (looking at you, iCloud and OneDrive >n>…) And they try marketing off of that comfort? Comfort in what? Their willingness to deem your data as disposable as long as they get to push their success should, if anything, make you feel discomfort. They are selfish, despicable, greedy pieces of sh-

… Ahem. As I was saying…. This is a project trying to… rescue my data. And be able to view it as if it never got exiled from its old home. Pretty much that’s it.

My current public work in this project is something called the bridges… It may help you too! So that’s why I made it open-source :) It’s intended to be as unbiased as possible, so people who are much better at applications can build pretty UI on top of it.