#Mozilla begrudgingly implements tracking protection by default, after being shamed into it yet again by Safari, the DuckDuckGo browser, and Gnome Web (just the ones I know who implemented it first) and talks a good talk while still having Google as its default search engine and continuing to get hundreds of millions from Google/Alphabet, Inc. – the largest surveillance capitalist in the world. (Mozilla Corp is ~entirely funded by surveillance capitalism.)



@aral I think, we should decentralize the search.
We need a p2p general-purpose search engine with our own crawlers, written in something native and efficient (like Rust) so that anyone can install and run their own copy, and, importantly, with clean and easy UX plus the name snappier than Google.

DuckDuckGo, good as they may be now, are still a central organization.
Sometime, google had "don't be evil" motto too.

@rugk @aral searx is just a shim for Google et al. That's not what we need. As well as StartPage and DDG.

One of the projects that get it kinda right, is YaCy. But it's Java, and I'm not installing it. And, damn, the name is awkward. "YaCy it" doesn't have quite the same ring to it as "Google it".

@drequivalent @rugk @aral Searx is inspired from from the front part of Seex, a dead project that planned on decentralizeing crawling/indexing/search and failed. The maths were really hard...

@drequivalent @aral We have that in Yacy now., but I'm not impressed with it.

I came to the conclusion that to get effective peer-to-peer search we need to move away from general purpose search. That the two are at odds.

I'm very interested in addressing the pain points of making this a reality!

I adressed YaCy. It's on the rght track, but it's not perfect.
How exactly would you "move away from general-purpose"? Ok, you may move. I may move. Who am I kidding, I won't. Nobody will and no amount of convincing will make them. Pulling up a phone, quickly entering a question and getting an answer is wicked sick, nobody will ever turn that down. I doubt you would.
I also highly doubt that control and surveilance is some inevitable evil to go with it

@alcinnz and also mind you, I'm not talking about building stuff just for me or you. We are what they call "Internet natives", we know the tricks of the trade of being online, Proxies, VPNs, TOR, we can get around just fine. But that doesn't solve the systemic problem. I'm talking about building stuff that is sufficiently safe, secure, private, and which an average Joe can use. Yes, the masses. Those who don't know a file from a pile. Your mom and dad and the guy that serves you drinks every Friday. We gotta take them away from big G and get them on our side which means we have to give them what they have with Google but better. Only then things are starting to get interesting. @aral

@drequivalent @aral To oversimplify for the sake of tooting, there's a couple aspects to it.

First let's acknowledge that this isn't actually a new thing to people. Many will jump on Amazon rather than Google for product searches, as doing so reduces the haystack being searched and allows the search engine to be better tuned to it's purpose.


@drequivalent @aral That single search box is a major convenience to most people, and should be kept. But who says it must Source it's results from a single provider? Web standards exist that lets us do much better than that.

Fundamentally though the only thing that'll convince to switch is convenience, and it's my passion to see if we can achieve a better experience by making this switch.

And you're totally right to ask these questions, as I've got a high hurdle I'm challenging to leap over.

@drequivalent I fully agree - up to that "Rust" thing. The more I look into it, the more I think that the whole P2P thing only will succeed as soon as it is *actually* P2P and so eliminated hosting for arbitrary non-tech users, in other words getting it to run mustn't be more complex than, in example, installing an app on a smartphone or a web browser on a desktop PC. Maybe that would even work...?

@z428 @aral You know that Rust compiles into a machine code binary that you just run, do you? With things like Java, you need a virtual machine and also it's wasteful when it comes to resources. We need the leanest shit we can get.

@drequivalent Yes. 🙂 Actually I like Rust a lot. And we could go on for ages arguing about why Java sucks and why loads of people (including "my" business environment) still use it. But my point was different: We need the leanest shit we can get, but we should *very* much focus on "being lean" on a wholly different level. My ideal end user couldn't care less which language is used to implement this - it should be as easily accessible as somewhat possible, no matter the language used. 😉

Sign in to participate in the conversation

Русская нода социальной сети "Мастодонт", части Fediverse - всемирной федерации социальных сетей. Зона общения, свободная от рекламы и шпионажа, теперь и в России.