Comment by ravenstine | Hacker News Reader

Comment by ravenstine | Hacker News Reader

People need to keep their own personal archives. While there are great projects like archive.org, it's becoming more clear that it's not enough. There's not enough resources for an organization(that isn't spying on everyone) to archive YouTube, or Slack history, etc., so it's up to you to back up that data if it's important to you. Culture can be easily lost to the void if we don't do this.

As I've mentioned in another HN thread, I've been using youtube-dl to back up YouTube videos that are important to me since it's increasingly obvious that YouTube is merely a profit machine that would delete all of its user's videos in one fell swoop if they made an extra buck off it. I've also begun to back up some blogs that have vast mounts of great content, but I've found that HTTrack is... well, terrible. That's not to say it doesn't technically do its job, but it's way too aggressive. I wan't something that's easier to configure and can do a better job of taking "snapshots" of blog entries while blocking ads, javascripts, being able to resume properly if my power goes out, etc.

There may be more modern projects, but there's also "Lots of Copies Keep Stuff Safe" at https://www.lockss.org

When you say personal archives, I wonder a little about what we'd lose to filtering by people who know they're building an archive for posterity, versus people in the future rediscovering things that were safely forgotten for some long span of time.

At the delta of personal archives and projects like this, I imagine something a little like SETI or folding for home archival (is this just describing IPFS?), but I think one of the challenges is how we preserve the right kind of data. I did some spring-data-cleaning this year while building a new system, and one of my tasks was addressing duplicate files. This forced me to confront hundreds of gigabytes of VM images, snapshots, and massive piles of duplicate files from various installations of Ruby and Python.

But perhaps I underestimate the value of even the duplicate VM images or language installs. Perhaps, centuries hence, researchers will hopefully sift yottabyte of data looking for all of the dependencies to successfully resurrect applications they need to experience in order to understand references to them in well-stewarded tweets, posts, or articles.

You can always help by volunteering machine time to the archive team if you want a low effort way of helping. https://archiveteam.org/

I noticed old articles I had bookmarked were disappearing so I’ve started using a self-hosted Bookstack to save a copy of everything interesting I read. I enter reader mode and copy eveeything on the page, and then paste it into the WYSIWYG editor. Everything is formatted nicely and it even creates a menu using the head tags of the sections. The only problem is, while images copy correctly, they’re still hosted on the original server so I have to replace them with an uploaded copy. Only takes a couple of minutes though. I could probably automate this with a script.

If you use Firefox, there have been several implementations of this add-on that I've used for ~10 years:

https://en.wikipedia.org/wiki/ScrapBook

Reply on news.ycombinator.com