[d0gbread]

You might not be wrong but I wouldn't mind seeing some actual data on the topic (admittedly have not looked). Google feels like a cheap example since their non-ads business model is to throw mud at the wall, and the graveyard is full of tech that did often make it into core apps. I've been on Gmail since forever, which stands as a Google-backed example of my previous comment sans some serious but statistically insignificant drama, which has prompted some to migrate and still have email.

Much of the early internet also isn't there anymore, and running a modern web application that hosts images, gifs, and videos at scale is obviously a different beast than static html where a marquee tag is as fancy as it gets.

[BaseballPhysics]

I would claim the early web diminished primarily because the users left, not because the services died off and left users stranded (though I'll note that a lot of early services--IRC, Usenet, message boards, blogs, etc, are still around, though reduced in size and scope).

As for services dying, I picked Google because it was easy. But the graveyard is endless: Myspace, Friendster, and Geocities all spring to mind, just in the social space (and since you don't want to talk about Google, I won't mention Google+ or Orkut).

That's not to say the early internet was somehow perfect. My point is simply that a service being commercial offers no guarantee of longevity, and given the need to extract profit or die, with no way for a community to take over an unprofitable service and run it for their own benefit, I'd claim the opposite is true, especially as interest rates have risen and cheap funding has become more scarce, thereby placing a lot more pressure on those services to monetize or die.

> and running a modern web application that hosts images, gifs, and videos at scale is obviously a different beast than static html where a marquee tag is as fancy as it gets.

Except that's not what a Lemmy or Mastodon instance has to do. A small instance might only need to serve dozens or hundreds of users, and it only needs to host and serve the content they subscribe to.

I'll take that over one massive single point of failure.

[d0gbread]

I'm certainly not advocating for a single point of failure, but rather a monetized, competitive space that competes on user experience. More akin to music streaming than movie streaming.

I think Usenet and forum numbers probably pale in comparison to Reddit type platforms. Centralization is easier for mass appeal

[BaseballPhysics]

> but rather a monetized, competitive space that competes on user experience. More akin to music streaming than movie streaming

I'll be honest, I don't know what this means, and comparing a service like Spotify to a social network like Lemmy doesn't make a ton of sense to me.

> Centralization is easier for mass appeal

Which is by definition a single point of failure.

As for mass appeal, I'd argue the issue isn't centralization, it's network effects. I.e. I'll use the service people I'm interested in use. If you have a lot of fragmented communities, that becomes a problem, and thus Reddit dominates. And it's the problem federation is specifically intended to solve, by stitching those individual communities into a larger network.

[d0gbread]

I mean to refer to centralization as a spectrum so I don't intend to say centralization means one option.

What I mean by comparing music streaming to movie streaming is that for all intents and purposes you can feel like your Spotify, YouTube, Amazon, Apple, whatever music libraries are pretty comprehensive, whereas that is absolutely not the case with Netflix, Max, etc.