It is, though the numbers are depressing.

A site like Reddit grows in its daily active users more than 15x (!!!) entire MAU base of Lemmy. “Reddit migrations” are barely a margin of error for them.

This is excluding, liberal, assumption of bot counts.

Even better, it’s now a nice database of who companies and governments can go after when they want or need to!

The Internet is just a bunch of servers my dude.

Someone has to pay for them, and all the other infrastructure around them. And with a large part of the world being on the internet a significant portion of their day the costs for even the most efficient centralized services running “at scale” (see: hundreds of millions MAU) are astronomical. In the tens of millions to hundreds of millions of $ annually, just an infrastructure, never mind human resources.

Almost none of these companies survive off of donations. Wikipedia stands out as one that does mainly because they host static content, which is incredibly cost efficient to serve up., and even then their costs are pretty astronomical (there are some debates around their costs of course).

Federated services have an asymmetric scaling problem. A linear growth in users results in a exponential growth in infrastructure costs. While centralized services tend to be almost the entire opposite of that and usually see logarithmic infrastructure costs against linear user growth. Where infrastructure costs are more efficient as their user base grows.

Federated services don’t benefit from running at scale, the more they scale up the less benefit there is to scaling. It’s a really shit situation to be in.

This is why the internet is largely just cyber feudalism. Because the only ones that can afford to host large scaled services for their users are the ones that are making money off of it. And that’s for centralized services, never mind decentralized services which are unbelievably more expensive to host.

I’m coming at this from the standpoint of an engineer, I don’t have answers or solutions, but the first thing we have to do in order to start figuring out solutions is to recognize the problem.

What most people in this thread don’t realize is that what you’re seeing right here is the problem with federated services in this day and age.

Federation protocols and systems just are not mature enough to scale.

Yes you will essentially always have to abandon ship anytime any federated service scales up it’s user base. It will always be entirely unaffordable and unobtainable for randoms to host their own servers because the compute storage and networking requirements will far exceed what most can’t afford.

As an aggregate federated services are always more expensive to host then centralized services. And that cost scales less efficiently than centralized services. Meaning that with linear user growth you get exponential cost growth, and the barrier for entry follows.

Which means that all federated services have to have centralization in order to scale. In their current form.

This is a really tough problem to solve and is going to take a lot of time and money to build good solutions for. Time and money that… You guessed it, is largely funded by profits not donations.

And now we have looped back around.

Let’s not mention the abysmal performance for servers. Making it largely infeasible to scale.

It’s not the solution, not even remotely close, unfortunately.

Except that relay nodes often get out onto proxy lists.

Which means you now get to solve capchas for absoluty f-ing everything now.

I mean, the business model works? They make money, they pay staff, and they are growing.

I don’t know what you’re talking about, people have price sensitivity of course. You are projecting yours onto “everyone”, is it not a successful business?

There’s a niche they cater to, if you are not that niche then you are not that niche. Doesn’t mean the niche doesn’t exist.

$10/m is unlimited searches though…

And yeah, searches are actually quite expensive. There’s a LOT of infrastructure that goes into making something unique with your own search engine that isn’t just a wrapper over Google.

The actual compute cost per search, in 2024, was $0.0125. Kagi states they want to keep Costa below $0.015 per search, but their search partners are a major expense.

That ofc ignores all the supporting infra, devs, support…etc that goes into making it all possible.

This is the kind of conversation, healthy, back and forth, and conceding instead of doubling down as we learn more that I wish was more common on the internet these days.

Bravo, really.

They’re definitely stretching themselves too thin, but as long as I get better and more relevant, cleaner, no advertising search results for my knowledge work and research. With my privacy in tact.

Then I’m continuing to pay them for a product I find to be superior than the alternatives.

What advertising?

Every single time with red comes up there’s always this FUD. You, specifically, don’t miss any opportunity to make mention of this. Across Lemmy, which is rather suspicious. Helping the Russian war effort? That’s a pretty big leap here.

Why?

Imagine a search engine aggregator aggregating search engine results from multiple sources for aggregation. The more indexes they support the better the results are going to be for everyone, I don’t see this as a problem for data aggregation.

Why should data aggregation give any sort of shits about geopolitics?

Regardless, the topic of this post, fediverse search, is part of their own search engine anyways afaik

Cool, lots of information provided!

Oof, hell no. That’s some Facebook level cancer right there when they removed downvotes.

It’s just a form of white washing that makes the same people who made up being offended by “black lists” and “master branch”.

Edit: Y’all do realize the irony of exercising your ability to downvote a comment that is defending your ability to downvote?

This is the way.

I do this. PFsense DNS resolver, and have loopback enabled.

DNS for all the domains points at a reverse proxy (Caddy) that handles valid HTTPS termination. So all my services have valid HTTPS certs, and devices on my network can access them normally.

The issue isn’t really the OS in many cases.

The barrier is getting a battery replacement. Which many devices make exceptionally difficult, to the point where buying a new cheap phone is nearly equivalent in price as paying someone to tear down your phone and rebuild it to swap the battery

Yep. You essentially summed up my point.

There’s a difference between data display for academia and data display for the general public.

The general public is generally not well educated on understanding the data that’s presented to them. Big change in line up or down regardless of scale means big change. It could be from 100 to 100.8, but if the scale is zoomed in then that could be presented as a +80% change.

And often is and sometimes with the axes removed and shown on the news specifically to be manipulative.

I really don’t understand why I’m being downvoted above… This was literally part of my grade school education on identifying and avoiding misinformation. And later on, around how the general public understands data visualizations. They are largely understood at a glance and taken at face value without reading the axes.

This is a easy way to push misinformation. Not by actually pushing real misinformation but by taking advantage of the general public’s tendency to not read it carefully.

Which is manipulative. Which is why it’s taught in some places as part of the standard educational curriculum…

Way too much for sure.

Just the business internet to get the foot in the door for a static IP 5x’s the cost of my Internet.

It’s actually cheaper to just have DC IPs and proxy through hosted containers. Which is kind of crazy.

Negative aspect is that DC IPs aren’t treated very nice.

No I think someone saying Google it is still worse.

The former is being intentionally unhelpful.

Your example is being unintentionally unhelpful.

Intentional malicious behavior is far worse than negligence.

Using compressed axes to display data was literally “How to identify misleading statistics 101” in middle school for us…

It seems fine to you but for the majority of people it’s misleading most people look at the lines and the relative distance between them to make judgment calls. Not literally the entire point of graphs, to visually display information.

This is a well-known effect and is taught in pretty much every major curriculum.