The blog post talks about our self-hosting movies, photos, and podcasts, in nice Netflix-like interfaces. Sharing photos. That sort of thing.

You are talking about preserving intellectual independence.

Both are nice to have, but they are sort of different problems, right? Yours seems more important. And yours could probably be solved by a local copy of Wikipedia and an FTP server full of digital textbooks.

IMO one dangerous misstep we can make with self-hosting is to assume we need to start by matching the centralized services look-and-feel and polish (which is getting worse every year anyway).

▲

whilenot-dev 4 days ago | parent [-]

> one dangerous misstep we can make with self-hosting is to assume we need to start by matching the centralized services look-and-feel and polish

That's an interesting take. I think matching these services isn't a necessity, but getting a polished look-and-feels just helps adoption. Adoption isn't an exclusive scenario and everyone is free to choose and mix how they see fit.

My private collection won't ever compete with Netflix, Google or the like, and that's completely fine. It will stay a private selection of media with a strong personal preference - it ranges from research to entertainment, and also includes stuff that documents my own individual history. It'll shrink and grow as I want it, and if it reaches a scale that makes the jump from archival to hoarding work I'd simply need to reconsider my preferences.

Here's my take: The scaling issues of these tech giants won't ever reach my personal archive and any challenges with re-indexing, data analysis etc. should be completely approachable on SOTA hardware. Running anything that improves the searchability of my own archive can be run locally and in the timely intervals I prefer. To have this kinda quality approachable is a huge thing, and I can't wait until I can self-host some RAG enhanced vector search engine for a personal archive that grew overs years to take shape.

▲

econ 3 days ago | parent | next [-]

> My private collection won't ever compete with Netflix, Google or the like

You are not allowed to do that!

The reality is that with minimal effort and money you can cheaply buy some multi TB drives. Get a 100 euro computer, a 50ish internet connection, a torrent feed and a usenet sub. Netflix has very little stuff by comparison.

You can also set up a crawler with some archiving and crawl that part of the internet that is interesting to you specifically. Feed it the urls you've visited, crawl 1-3 links deep from there and you will be in awe how fantastic the search results are. If you type "banana" in the search box it will give you a surreal page full of interesting pages about bananas. (Interesting to you only)

As the collection grows and time expires you will also be increasingly shocked how many of those website are not around anymore.

I ran a dead link checker on 10k old bookmarks for a niche topic one time. I didn't pay much attention to the topic for 8 years or so. I was very excited thinking I would find all kinds of new developments. In stead it was depressing as f. Something like 50 links worked. Stuff like a link to a Google search result.

▲

movedx 4 days ago | parent | prev [-]

> includes stuff that documents my own individual history.

By this do you mean family photos and the like? I'd like to hear more about this. I'm building up a personal library like this too.

▲

whilenot-dev 4 days ago | parent [-]

Family photos, letters, contracts, receipts etc.

	▲	movedx 4 days ago \| parent [-]
		Ah OK. So basically: life. Roger. Do you use paperless-ng and the likes?