One of the things I miss most were the old Yahoo directories and Download.com. I'd go in there digging for a game and the results weren't sorted by 'relevance', so it would range from trash to some weird underdog game nobody ever heard about.
These days, search results are all weighted, especially on app stores, so you'd end up with everyone playing Among Us and never finding games like Predynastic Egypt or King of Dragon Pass.
I’m (slowly) working on an old-school manually-curated index of websites that I think have that “soul” of the old web. Many aren’t old, and lots are even using the latest in web technologies, but the common theme is “someone put time and effort into hand crafting this website for no other reason than because they were passionate about it”.
I really do think there’s a gap in the modern web for discovering stuff like this. Google only wants to show you what social networks and major entities have for your query, and SEO spam buries everything else.
lol, I have to agree with SEO spam. I tried looking for a good pasta sauce recipe lately. I remembered something recommended by Pocket, which was amazing (Essentials of Classic Italian Cooking), but it didn't even show up in the first few Google search pages for pasta. None of the results were even a variation of the superior recipes.
The top search result has 9 (!) ads, mentions the book, but butchers everything that makes the recipe good. It shows rave reviews and a rating of 4.8. It probably is better than what everyone's mom taught them, and better compared to a generic tin can sauce. But I think the old school algorithm of "best selling cookbook" does better.
Try looking for other heavily SEO-ed things and it's even more hopeless: weight loss, business advice, games.
I recently came across the following website in a HN thread complaining about bloat in recipe websites, I haven’t tried it out yet but it’s an interesting premise:
In the wake of the last "app stores are monopolies"-outrage (which happens at least yearly), I was thinking about a monopoly-less-appstore ecosystem.
In a place where everyone can start an app-store, people will need to carve out their niche. People could then start hand-curated appstores. For a niche. Put together with care and effort.
A little like-f-droid, but much more common and therefore covering a lot more niches. A "christian appstore", "socialist appstore", "appstore for old phones" and so on.
Kindof like the old "manually-created" indexes, but for modern times, in which phones are the primary gateway to the internet.
The promise of recommendation engines is that you will get that niche experience that will scratch your itch just right. I don't know how well that works in practice, but even so, it seems like it removes the chance of stumbling on something so alien you'd never consider it but you end up loving it.
I was thinking the other day about non-demand TV and radio, and how it was obviously worse in many ways than what we have now. But I spent many a happy afternoon watching a movie or classic TV show that I'd never heard of because I had no options, and really enjoyed myself.
Spot on. I’ve tried to merge the two ideas for my kids, I’ve got a NAS filled with shows, movies and music, and playlists that shuffle each so they can get some randomness and not just watch/listen to the same few repeatedly. We work pretty hard to try incorporate a big mix of content for them, and it seems to be working quite well so far. Also stops them being inundated with ads and recommendation sidebars designed to manipulate them.
I think on demand TV has a bit of a Skinner effect. You discover something awesome every once in a while, but most of the time it's trash, and you get frustrated looking for something really good. It doesn't help that a lot of Netflix productions have 20 episodes or so, of which only about 3 are really good.
Back in the day, you'd have your Simpsons and Power Rangers and just learn to enjoy it and talk about it with all your friends.
I think algorithms are built wrong too. We'd probably click on some show with a sexy model, or a rom-com twist on a zombie flick, or some reality cooking show with celebrities, or a gritty remake of a children's show. We don't really intend to watch it. But they pique curiousity and a little disgust. The algorithms register these clicks and long views as an "interest", and gives us more of these clickbait-ish shows.
I think what would work is simply narrowed down genres, say, Animation > Children > Horror (e.g. Adventure Time) or Animation > Adult > Tragicomedy (e.g. BoJack Horseman). Past that, just arrange at random, alphabetically, and not even by popularity.
> It doesn't help that a lot of Netflix productions have 20 episodes or so, of which only about 3 are really good.
Yeah this was really obvious to me in some of their Marvel stuff. Overall enjoyable enough, but only if I binged it while doing something else. Pacing and a lot of dead air time was otherwise a huge problem.
These days, search results are all weighted, especially on app stores, so you'd end up with everyone playing Among Us and never finding games like Predynastic Egypt or King of Dragon Pass.