[rglullis]

Problem: you are looking for a way to get rid of the annoying issues of the modern www. What is the solution that solves this with the least amount of work?

A) Develop a whole new transport protocol that does less than HTTP, develop client applications that use this protocol, convince a sufficient number of people to use this protocol, at least to the point where the majority of your activity happens there?

or

B) Install a handful of browser extensions that block ads and other nuisances on the modern www, and have it working right away?

[bayindirh]

Option “B” implies a cat and mouse game, which you can never win.

You can’t win a game designed and implemented by a mega corporation which is specially made to earn them money and protect their monopoly by being reactive and defending all the time. Instead you have to change the game and play with your own rules.

That’s option “A”.

[rglullis]

> Instead you have to change the game and play with your own rules.

That only works if you can convince the a substantial part of the participants to also play your game.

It's very easy to create an alternative internet where we can take away the power from incumbents. The hard part is creating all the activity that is taking place in the current one.

"Oh, but I can mirror the parts I want from the current internet into the new one!"

Not without playing into the same cat-and-mouse game.

[bayindirh]

Who says I'm trying to pull in everyone from the old internet to the new internet (Gemini)? If the people I care comes along, that's enough for me, and it's up to them.

For example, I switched to Mastodon, and follow people who I really want to follow are already there, plus I met a ton of interesting people, and was able to see real forms of people I followed before, so I have updated my views on them.

> "Oh, but I can mirror the parts I want from the current internet into the new one!"

Personally, I see Gemini or other protocols as equals to HTTP/S. For example, my blog is already text in most cases, has a full content RSS feed, so, also publishing a Gemini version is not mirroring what's on the web already, just adding another medium to my blog.

If I was pumping a 3rd party site I don't own from web to Gemini with a tool, then you'd be right, but publishing to Gemini is not different than having a RSS feed in my case.

[rglullis]

> For example, I switched to Mastodon (...) and was able to see real forms of people I followed before, so I have updated my views on them.

Isn't that strong evidence that it is possible to have a "human-scale" web built on HTTP, and consequently that there is not much benefit in restricting yourself to a protocol that is designed to be limited?

> Personally, I see Gemini or other protocols as equals to HTTP/S

Except they are not. Maybe it can do enough of the things that you care about, but Gemini is (by design!) meant to do less than HTTP.

> publishing to Gemini is not different than having a RSS feed in my case.

Again: if all you want is to be able to publish something in a simple format, then why should we care about the transport protocol?

I get the whole "the medium is the message" idea, I really do. I get that people want a simpler web and I look forward to a time where we have applications developed at a more "human scale". But I really don't get why we would have to deliberately be stripping ourselves of so much power and potential. Talking about Gemini as the best solution to the problems of the modern web is like saying we should wear chastity belts to deal with teenage pregnancies.

[bayindirh]

Yes, but it's important to understand that limitations are moved to Mastodon "layer" in that case. It takes careful deliberation and restraint to keep something tidy. Mastodon does this by limiting its scope and organizational structure. We as humans like to abuse capabilities. So, to keep something tidy and prevent (or realistically slow down) rot, you need a limit somewhere. Putting that limit to humans vs. the protocol is a trade-off.

In that scenario W3C doesn't put any brakes, Mastodon puts brakes on development, organizational structure and scope, and Gemini puts brakes on the protocol. So, it's the most limited but hardest to abuse in a sense.

I probably worded my "I see them as equals" part of my comment wrong. I know Gemnini is a subset of HTTP, it's more Gopher than HTTP, and that's OK by me. Moreover, that leanness is something I prefer. See, the most used feature on my browser is Reader mode, and I amassed enormous amount of links in Pocket just because of the reading experience it offered.

> I really don't get why we would have to deliberately be stripping ourselves of so much power and potential.

Because power corrupts and gets abused. A friend of mine told me that they now use Kamal which makes deployment easy. How it's deployed? Build a container -> Push to registry -> pull the container on the remote system -> runs the container -> sets up and runs a proxy in front of that container to handle incoming connections.

That's for a simple web application...

I mean, I push files to a web server and restart its process. I'm not against power, I'm against corruption, and given human nature, restraint is something hard to practice, and that's if you want to learn and practice it.

> Talking about Gemini as the best solution to the problems of the modern web is like saying we should wear chastity belts to deal with teenage pregnancies.

I never said Gemini is the only and the best way forward. Again, for me It's another protocol, which offers a nice trade-off for some people sharing a particular set of values. It's like a parallel web like BBSes or public terminals (e.g.: SDF).

Being an absolutist benefits no one. We should learn, understand and improve. We can have multiple webs, and we shall be free to roam them in a way we like. I'd love to use my terminal to roam some text only web with my favorite monospace font and terminal theme, but I like to write comments here and think on the replies I get, too.

I find myself preferring a text-only, distraction-free web more and more, and naturally evolving my habits and personal infrastructure in that way, but I'm not carrying a flag, shouting about end-times and preaching stuff as savior. I'm not that person.

[rglullis]

> it's important to understand that limitations are moved to Mastodon "layer" in that case.

Mastodon may be my preferred social network nowadays, but it's despite the prevalent philosophy from the development team. It's also arguably the reason that the Fediverse can not manage to grow to more than 1 million MAU.

>Because power corrupts and gets abused

The solution to this is not to get rid of power and keep everyone in the same small crab bucket. It's to make access to the powerful tools as universal and ubiquitous as possible.

> I push files to a web server and restart its process.

Your friend not being sensible enough to know when to use a tool vs when to keep it simple is not a problem of the tool. Also, talking about deployment methods seems so orthogonal to the discussion that I am not sure it makes sense to carry this conversation further.

[sneak]

Not really. You could have tinyweb/oldweb sites identify themselves with a meta tag, and have a browser that only browses those. A opt-in, web-within-a-web. And turns off js, cookies, and images.

You don’t need another transport protocol.

[Gormo]

How do you stop users who aren't using the custom browser from accessing these 'tinyweb' HTTP sites? How do you prevent content scrapers and search indexers from accessing them? How do you suppress direct incorporation of 'mainstream' web content into 'tinyweb' content?

If your goal is precisely to create an parallel ecosystem that's "airgapped" from the mainstream web, and you're already going to have to develop custom clients, content formats, and server-side configuration to implement it on top of HTTP, and engage in lots of development work to imperfectly isolate the two ecosystems from each other, why wouldn't you just develop a parallel protocol and start with a clean slate?

[rglullis]

> How do you prevent content scrapers and search indexers from accessing them?

How do you that with Gemini?

> If your goal is precisely to create an parallel ecosystem that's "airgapped" from the mainstream web

There is no way you can have an air gapped network with public access. The moment this "parallel ecosystem" showed any content that hinted at something lucrative, you will have people creating bridges between the two networks. Case in point: Google and USENET.

[Gormo]

> How do you that with Gemini?

You keep it isolated from the ecosystem in which all of those things are taking place.

> The moment this "parallel ecosystem" showed any content that hinted at something lucrative, you will have people creating bridges between the two networks. Case in point: Google and USENET.

The whole point is to minimize the chance of that happening -- by limiting mainstream appeal, keeping it a niche, and avoiding Eternal September -- and to maximize the friction of bridging these two ecosystems. And so far, they've done a fairly good job of it, since Gemini has been expanding for six years without any indication of any of this starting to happen.

[rglullis]

> and to maximize the friction of bridging these two ecosystems.

There is no friction. It's trivial to write a program that can scrape a Gemini network.

If there is no one pulling data from Gemini servers yet, is not because it's difficult do it, but merely because it's still too small to be relevant.

[sneak]

Why would you need to? The big web existing doesn’t hinder or harm the existence of the tinyweb.

[bayindirh]

We have Kagi Small Web and Marginalia already, if that's your aim.

[_Algernon_]

The benefit with A is that it also removes higher order effects of the modern web. You may for example remove adverts by installing an ad blocker, but that wont change the incentives that advertising creates (eg. clickbait, engagement maximizing, etc.). With A you can guarantee that the content is not shaped by these incentives.

[rglullis]

> With A you can guarantee that the content is not shaped by these incentives.

Without those incentives, you will quickly find out that there will not be much of an Internet out there.

If you don't believe me, check how many people are on YouTube talking about Open Source, when PeerTube exists and already can reach millions of people.

[_Algernon_]

The internet and web existed for a long time before everything became infested with advertisement: Hobbyist bulletin boards, Wikipedia, the blogosphere, etc. These had enough content that a single person couldn't consume it all in a lifetime.

[rglullis]

That internet was also only interesting and valuable to a fraction of the people who use it today.

And if you don't care about that and you are thinking from what you might get out of it: an internet where 99% of the content is crap but universal will end up with more valuable content than a neutered internet that can prevent the emergence of crap, but is tailored to appeal only to 1% of the people.

IOW, no one cares about reading all of Wikipedia, and Wikipedia would never reach the size it has if it was something only for a handful of individuals obsessed about their particular hobbies.

[selfhoster11]

Good. Not everyone has to be invited.

If the global Internet is a train in the 70s, Gemini is meant to be the non-smoking car at the very tail end. Those who cannot tolerate smoking (malicious practices surrounding user attention and agency), then they are free to make their way there and sit with other non-smokers. You won't have as good a conversation as in the smoking cars, but if the smoke really, truly bothers you, it's a space where different cultural rules apply by design.

[rglullis]

You may think that way, but to me a more apt analogy is that global Internet are the regular trains of today and Gemini wants to be a bizarre version of the quiet wagons where people can only talk if they don't use the letter "e".

[uamgeoalsk]

A movie theater where 99% of the movies are crap but universal is less interesting to me than a well curated cinematheque.

What I'm wondering is why do you feel the need to dismiss people's niche interests like this? You've even been rude about it in other answers, calling it "masturbatory".

Not everything has to be for everyone. Not everything has to be for you.

[Gormo]

> Without those incentives, you will quickly find out that there will not be much of an Internet out there.

Well, there is plenty of interesting content on Gemini. If you're OK with having 50% fewer needles in order to get rid of 99.999999% of the hay, then it's a win.

[shakna]

Considering "B" is becoming less possible, thanks to Google dropping Manifest 2, and going out of their way to enforce a lot more tracking, "A" looks like a lot less effort - you don't have to fight FAANG.

[rglullis]

Chrome is not the only browser out there. Firefox is still a good browser. If you depend on Chromium: Brave is keeping Manifest v2 and their ad-blocking extensions work out of the box.

[shakna]

And HTTP is not the only protocol out there. Plenty of others exist. Like Gemini, that has multiple browser implementations.

What's your point, exactly?

[rglullis]

My point is that the choice of protocol (much like the browser) is not a relevant factor if your goal is to be able to participate in the www without dealing with the issues.

We can have all the upside of an http-based web, without dealing with the downsides. The converse is not true. A Gemini network is by design limited in functionality, which is a downside that can not be mitigated.

[Gormo]

> My point is that the choice of protocol (much like the browser) is not a relevant factor if your goal is to be able to participate in the www without dealing with the issues.

Right, but that isn't the goal of Gemini. It's goal is to create a distinct ecosystem, not to participate in the existing one with marginally less annoyance.

[rglullis]

Even worse! This makes the whole proposal even more misguided.

Different ecosystems only make sense when we have distinct populations that might be as well considered different species.

[fiverz]

It's not FAANG anymore, it's GAYMMAN now

[zzo38computer]

In some ways, A is easier, but not in all ways. Each has its own difficulties.

These are not the only possibilities, though; a third possibility might be:

C) Make a simpler set of features which are compatible with some parts of WWW and implement that.

However, you can do two or all three things if you want to do; you are not limited to doing only one thing. I think all three of these (A, B, C) have their own benefits, so you don't need only one.

[uamgeoalsk]

What's more fun? Definitely A.

[rglullis]

You are not solving the stated problem. You are just admitting that working on a new protocol is a masturbatory, "the journey is the reward" kind of exercise.

[uamgeoalsk]

I'm not aiming to solve the stated problem, I'm having fun with gemini.

[Gormo]

The answer is "A". Perhaps some people are avoiding saying this too explicitly because it might sound a bit elitist, but I'll put how I see it as frankly as possible for the sake of clarity.

Gemini is not trying to solve a technical problem with the web. Is trying to solve a cultural problem that arises from the web having become a mass medium, in which every site's focus gradually erodes under pressure to optimize to the lowest common denominator.

Creating a new protocol from the ground up, and requiring users to install a distinct client to access it, isn't just about keeping the software aligned with the project's goals, it's about creating a sufficient threshold of thought and effort for participation that limits the audience to people who are making a deliberate decision to participate. It's about avoiding Eternal September, not about creating a parallel mass-market competitor to the web.

It's not about blocking the annoying ads, popups, and trackers, just to access sites where the content itself is full of spam, scams, political arguments, LLM slop, and other assorted nonsense, and instead creating an ecosystem that's "air-gapped" away from all that stuff, filled with thoughtful content produced deliberately by individuals. It's about collecting needles together away from the hay.