RE: Are we rebuilding the centralized web, minus tracking?

> That [one central place] simplify a lot of things and make browsing this space faster.

IMO it's both good and bad.

A central place encourages:

1. Frequent polling, to see if there's anything new

2. Frequent polling, to see if your posts have many "likes"

3. Frequent polling, to see if you have many "followers"

4. Competition for attention, which is generally bad for an inclusive community that promotes free expression, and also leads to ...

5. ... short and shallow content optimized for maximum buzz and short attention span

6. An "A-list" of users popular enough to discourage others from sharing unpopular opinions or participating in general

(see Shirky's essay, IMO awesome and still very very very relevant)

For me, the slow pace, the cherry-picking of content to consume and the rich, sincere content are the most important features of geminispace and the small internet in general.

But, I agree with many things @Skyjake said, so I'll give one example. Centralization is a great thing in many cases, and I don't think that "centralization is bad, period". I've always hated those monolith vs. microservices debates among colleagues. In my current job, I've re-architected our core backend component to be what can be described as a "multifaceted monolith": one component that does A when it runs with -a and does B when it runs with -b: one gigantic codebase with many features and heavy code reuse between them. One deployment of the product consists of many "microservices", which run the same container image but with different flags that tell them to do just one thing, and they're scaled independently of each other. Everything works better and everybody's more productive now. Why am I telling this irrelevant and uninteresting story, from the irrelevant viewpoint of a developer? Because a central place makes life easier for the one in charge of keeping the lights on and for its users. Centralization can be achieved using something like the "federated timeline" thing you can find in your Mastodon instance of choice: there are ways to implement a "Geminispace agora" that feels centralized and convenient, although it's decentralized under the hood. I'd love to see something like this in Bubble. And I'd love to see a federation protocol based on TOFU and Gemini subscriptions, as a super minimal alternative to ActivityPub.

> Aggregators brings back the feeling of the early web with "planet" type of website (or, in more "recent" technology, the [indieweb]).

IMO aggregators and subscriptions are great features of Geminispace and gemtext, and I prefer them to social networks. It's great to have "subscribable" pages in Bubble, so one can avoid the Bubble "interface" and "follow" users by subscribing to their feeds.

My fear is that social networks, as opposed to aggregators and other things that usually don't "sort" content and "rank" people, can "speed up" the small internet and take away much of its appeal. Maybe this early web charm will go away if we replay Web 2.0 over Gemini: as I said before, I think social networks can encourage discussion (at least, about focal concerns of most geminauts), but at the same time, they can be dangerous to minority groups, non-English-speaking groups, niche sub-communities and introverts.

(It's no secret that I'm working on my own "social network" thingy, a "decentralized aggregator with ability to reply" kind of thing, a very different vision and approach compared to Bubble: I don't know if it's possible to build a "slow" social network with intentional limitations, that's fun and useful but still non-addictive to use. I don't know if there's any demand for something like this, considering the success of Station and Bubble, but I hope that Geminispace has a free spot on the slow-fast/decentralized-centralized scale for something like this. It's an experiment: even if it fails, I find it fun and maybe we can learn a lesson or two about this community and find ways to improve "infrastructure" like Station and Bubble.)