I agree with the author, but from a different direction. The Gemini community has this ethos that stripping away all the styling etc. makes the internet about "content" again. Setting aside for the moment that the definition that content = text is unnecessarily reductive (and insulting to other forms of content - is a beautiful animation not content? is an interactive map of a country showing how covid is affecting it not content?), in my experience - people who care deeply about content are doing everything except bikeshedding about internet protocols.
The people whose content I most respect are writing into a blogger setup that hasn't changed for decades, or into Wordpress's editor that feels like it actively fights you when you want to write. They like making it look like they want it to, and they like having analytics or getting paid for their content with ads or subscriptions. And on the other hand, barring one notable exception I haven't found anybody in the gemini space whose content I found interesting in itself. (Of course, that's just personal opinion - you might find many more people in the gemini space interesting.) And the moment I realized that a significant portion of the content on the gemini space was about implementing a gemini reader, it hit me that it was pure solutionism. Gemini is a protocol for techies who want to bikeshed about tech!
(I could also go on about how the idea that stripping content of styles or """fluff""" is the only way for readers to get it on their own terms. I use stylus to override the font on every website I visit - it didn't require reinventing the wheel for that!)
> people who care deeply about content are doing everything except bikeshedding about internet protocols.
> Gemini is a protocol for techies who want to bikeshed about tech!
I don't agree with this on the matter of fact -- the spec has not meaningfully changed in at least a year, no major companion specs have been added since "gemlog", and the mailing list is currently down. While I may have agreed with you in late 2020, The only thing being produced on Gemini now is "content", and there's more of it than ever!
"Bikeshedding about internet protocols" doesn't mean bikeshedding the gemini protocol specifically. The Gemini protocol is stable, I'm sure, but what I mean is that people are more interested in discussing the tech especially vs http or whatever.
Well, it still doesn’t seem to have escaped the narrow community of “geeks who are into tech-related things”. I’d really like to see things that I don’t already see on places like HN (because for these things I’ll just use HN).
I don't think gopher/gemini is geeks only, and the overlap with HN is certainly small. It's very popular in some communities such as the tildeverse where less-techie folks write on gemini as well. Last time i surfed on gemini (like a year ago) i was surprised there was so little technical content.
The great thing about the web when I came along with was all the character, good, bad or otherwise that personal sites, even some corporate sites that people build had.
GeoCities came along and things could be ugly, but it felt genuine.
Even just background choices or ugly gifs…
I miss the personal wonkiness and folks not afraid to make ugly sites.
The race to ultra minimalism feels as cold as a Facebook profile, maybe more.
https://search.marginalia.nu/ made the interesting observation that a big chunk of that amateur OG internet never stopped or went away, Google etc. just stopped linking to it and it went invisible.
Without numbers, it's hard to say how much of it is left. However, I do believe that a big chunk left and mere tiny bits of it are left. And of course, archives of old and now dead things..
The vast majority of people on the internet are not ever exposed to that kind of web (and most of those who were have moved on). How does someone even find out such a thing exists? How does someone figure out how to participate, if their entire web experience comes from facebook-twitter-youtube-instagram era?
In the 90s, the original amateur internet was something you'd be inevitably exposed to if you did anything at all on the web. Geocities was a thing, ISPs often offered free space for hosting personal home pages. I saw normies, total non-geeks make personal home pages because that's just the kind of thing people found on the web and wanted to try. My older sister had her own homepage. She's not a programmer and not a geek. Her site was a part of some webring, linking other sites made by teenage girls..
"Social media" was forums, likewise hosted by individuals, often in conjunction with their personal home pages, linking to other such sites...
It's only a recurrent pattern because the capitalist machine is destroying creativity/leeway. Just like nature produces tomatoes (and other fruits/vegetables) of every color, but good luck finding a tomato that's not red in a supermarket.
Abolish capitalism and its central enforcement militia (the State) and suddenly diversity will be brought back to life. Without legal/physical threats (police/tribunals) and indoctrination (schools/media) to persuade the weird folks from being weird, we'll see more and more unique stuff.
I noticed this with restaurants. Old location (Australia): you drive to a McDonald's and get a quarter pounder. New location (Europe): you walk down the street and there are heaps of little unique eating places and small chains.
What causes the difference? I don't know. It might be down to city planning causing people to walk more and reducing the friction of entering a random store to get lunch. It could be something to do with real estate laws. It might just be cultural expectations.
Point is, this "big chain culture" is not universal and not inevitable. It's something we humans created in certain areas.
It has to do with the economic system and local urban planning. Many cities in Europe have refused permits for Mac Donald's to open shop, and that's a good thing. Because once settled, these megacorps who pay less than 1% taxes will take away all the business because they can have a bigger variety of products/services that a small shop can't afford, at a price driven by economies of scale that a small shop can't afford, treating employees in an illegal manner that a small shop can't afford to go to court over, and avoiding taxes through loopholes which small shops can't afford to know about.
If you can, don't let chains setup store near your home. If you can't, seriously consider getting involved in sabotage operations with people you trust. That's how Google ended backing off from Kreuzberg (Berlin) a few years back. Mac Donald's also had a few famous burnt-down "restaurants" here in France, but sabotage doesn't have to involve flames. A pack of sugar down the concrete-mixer will do the trick. If the shop has opened already, anything to block the locks will block business for at least a few hours.
It's David vs Goliath but if you've got support from your neighborhood you can win. Just don't ever think police and politicians are on your side.
Plausible but I think it's more generic than that. Evolution also filters out a lot of "creative" designs. Probably not as much as finance based markets I guess.
I agree with your critique of capitalism, but beware the naturalistic fallacy: nature is similarly efficient and ruthless.
Wild tomatoes are red, tiny, and barely useful as food. The myriad heirloom tomato shapes and colors you’re referencing were all carefully sought after and stabilized by human gardeners.
> The myriad heirloom tomato shapes and colors you’re referencing were all carefully sought after and stabilized by human gardeners.
They've been groomed for generations, yes. But they've not been stabilized. When i talk about the variety of tomato colors/tastes, this is something natural that will happen over a few generations in your garden.
Plants will often borrow taste from other surrounding plants, as for color i have no clue what triggers them to change, but i've seen with my own eyes after a few years, new generations of vegetables starting to change colors (not uniformly across the entire garden). This of course is not possible with trees (eg. apples/oranges) as you would need to wait several generations of trees (that's a long time), nor is it possible when you plant seeds from the supermarket every year to replace last year's plants, as the commercial seeds are almost bit-by-bit copies of one another and will yield the same tasteless tomatoes bypassing the natural circle of evolution to your local environment.
If you can, i strongly encourage to go talk to local farmers and borrow a few seeds. There's some amazing stuff out there which you won't find in commercial gardening shops. At least that's the case here in France, where the government for many years made it illegal/criminal to share or sell "peasant crops" (which a judge ruled is legal only a few years back).
> They've been groomed for generations, yes. But they've not been stabilized.
You must be using a different definition of “stabilized” than the horticultural sense, because it is absurd to suggest that tomato varieties are not stable.
> If you can, i strongly encourage to go talk to local farmers and borrow a few seeds. There's some amazing stuff out there which you won't find in commercial gardening shops.
I’m a member of a seed saving exchange and active in my local growing community. Horticulture is my primary hobby.
That's part of what's frustrating me about this! Usually non-profit projects/protocols are extensible, like the web and the Internet. This, despite the fact that for-profit corps try to appropriate/destroy the free/extensible platforms just like Microsoft/AOL tried with the Internet in the 90s.
So here, we end up in a situation that because silicon valley moguls exploited an extensible protocol (by hijacking the standardizing committee to their profit) we adopt a self-defense reflex of making everything minimalist to reduce the attack surface.
I'm interested in low-tech/ecology in general. For me, the equivalent in the physical world would be if you're not happy with a house using too much resources/energy so you decide to live without walls or floor, with a simple ceiling over the head. Sure it does the job of protecting you from rainfall, but does it fit all the other functionalities we expect from a home? I'm much more interested in a clay-based (or other local materials) housing, personally.
> Gemini does not prevent you from having any of those things.
I strongly disagree. By willingly limiting the markup, gemini actively tries to prevent you from doing these things. In contrast, what's wrong with the web is everything JavaScript and some early design issues with CSS, but HTML is good (although it could be simplified/improved).
xHTML more specifically is easy to parse and extensible so that different clients can implement more features. We need more HTML less JavaScript. In this sense, going the gopher route is a step backwards in my view, because it means for most of my practical needs (such as a simple user-submitted form) i need to stick with the web as we know it while i would be more than happy to take part and contribute in a privacy-friendly/minimalist subset of the web (as a standard i mean, i already do that on my websites).
I respect your disagreement, but I feel like you're disagreeing against the wrong point here though.
The main difference between gemini (the protocol) and http, isn't that a gemini page need only contain text and be deprived of all other kinds of context. The gemini protocol does not stop you from interacting with specific filetypes or other protocols, any more than http does. It's a protocol. It's up to people implementing clients to the gemini protocol to decide how to handle such externalities.
E.g. the lagrange browser downloads images and displays then inline in the page, which I like. I haven't tried audio files, but I imagine they could probably work similarly if necessary.
Along similar lines, the gemtext specification is not intended as some sort of viable html alternative. It has completely different design goals.
So no, the point isn't about limiting the kind of content you can interact with. The point is, by enforcing a degree of separation between what is "content" (i.e. text), and what is "externalities", this makes simplicity of text (both in terms of presentation and production) the main driver, leaving externalities up to implementations (which may be as simple as encouraging actual external use, including an http browser for http content). More importantly, this allows the user to be in control over how these potentially abusable externalities should be handled, rather than ceding that control to the webmaster, whose agenda may not match yours as a user.
And some (including myself) actually like this separation.
> E.g. the lagrange browser downloads images and displays then inline in the page
That's true, but it's based on opportunistic guesses, not semantics, which prevents many use-cases (including alt-text for accessibility of images). For example, we could do the same with video, but what about alternative soundtracks/captions? HTML has <video> <audio> and <caption>. Building a browser based on auto-guessing which link is related to which one and what the relationship is between those is... hazardous in my view.
> this allows the user to be in control over how these potentially abusable externalities should be handled, rather than ceding that control to the webmaster
I love that argument! But that was also the premise for the web, and i don't understand why we're not enforcing it there. CSS was supposed to permit user stylesheets, and JS empower users to script their interactions with the server. I would argue the problem with the web is it has moved away from declarative to imperative model where the server dictates your rendering. I can see how gemini addresses this by removing extensible declarative elements altogether, but i don't think that's the only way.
Personally, i think adding more semantic elements in HTML (eg. no-JS web components) would make client-side sheets more realistic while allowing the webmaster to propose a specific stylesheet for their recommended UI/UX. Removing the semantics is not gonna help empower clients, except when it comes to plaintext content. So, adding a new content-type over HTTP? Sure. Just throw in some CommonMark: it's well-specified and more or less complete (depending on your use-case). Adding a simpler protocol for delivering content? Why not, but it's not that easy to future-proof: HTTP is arguably more complex, but has some interesting properties such as content-type/encoding negociation or decoupling from transport security (for .onion/.i2p/DANE or any other future technology).
Just to be clear i like gemini very much and i've been wanting to support gemlog to my blog for a while. I just think there's room for some other technical foundations to explore the political promises of gemini.
> Bottom line is, if you agree that the modern web has become an awful place, let’s work on changing that for everyone, instead of abandoning it like a bunch of billionaires trying to escape to a different place, before this one collapses.
The implication being that the web can be changed? I don't believe that, I believe the web is broken and it's only getting worse by the day. I'm not going to discourage you from trying though.
> If you don’t like how modern websites track their users and flood them with ads, then don’t do that on your website [..]
> If you don’t like JavaScript, don’t use it [..]
Sorry, these are not answers for me as a user. When I browse the gemini space, I'm guaranteed to not be flooded with ads, harmful javascript, etcetra. And I'm also guaranteed not to have to play whack-a-mole with ublock/umatrix trying to make the actual fucking content somehow load. That's where the value is! "In theory, someone could make a nice website. Some actually do" is nice but it does nothing for my browsing experience in general, because the next link is going to be crap again.
Disclaimer: I haven't (consciously) heard of this Gemini thing before, but:
> I'm guaranteed to not be flooded with ads, ...
I guarantee this will only be true until it becomes popular enough to be a juicy target for the more - uhm - "business-oriented" people, it's not because of a restricted feature set.
Creating a technology which is intentionally uninteresting to business is an interesting angle of course which I can get behind ;)
Now tell the business people you can't display images and can't tell apart bots from "real people" on Gemini. They'll much prefer to keep their ads on the web. Being uninteresting to business is indeed a feature not a bug, but i agree with much of the author's criticism of gemini reinventing a square wheel: it's fun but a modern web subset would probably be easier to learn for many people while keeping backwards-compat with the rest of the world.
> Now tell the business people you can't display images and can't tell apart bots from "real people" on Gemini. They'll much prefer to keep their ads on the web.
100% not true. They will simply pay less per impression for their ascii art ads.
Sure, but there are plenty of users that are out there with different needs than you. Plenty of people out there don't deal with text well. One of the best parts of the web has been the ability to share images and simulations. Gemini space has no 3Blue1Brown and no Khan Academy. You can't watch a master cabinetmaker hand carve some dovetails. Gemini can't display textual math because the only thing text/gemini supports is UTF-8 text. There was talk in the early days of having screen readers and such but it's mostly gone unfollowed up on. There were a lot of debates about other accessibility things but the minimalist faction ended up winning over the pro-accessibility faction.
Gemini is only useful as a user if you're a software hacker who doesn't like math or images or videos or other media and have no accessibility needs. And it's borne out on Gemini space itself, as that is the core demographic of Gemini. To me Gemini feels more like a clique than an open space, where the text-loving cool-kids hang out. As far as internet projects that I feel are more inclusive (which isn't a high bar compared to Gemini), I find Usenet, Yggdrasil, and Secure Scuttlebutt better. There's a lot less waxing and waning about good and evil (and how the Geminauts are good, of course) on those places too.
The fact that Gemini is not what I want does not invalidate the opinion that the web is broken, and Gemini provides answers for that in its niche. Likewise, other people wanting other things doesn't invalidate anything. We don't exactly have to choose one technology and solve all the world's problems with it. I'm ok with Gemini being what it is and I'm not pretending it is (nor should it be) everything for everyone. It's not much for me, but it's something.
That seems more like an artifact of it not being popular than a feature of the protocol. If companies wanted to, they could just run native advertisement on Gemini just like they do elsewhere, just disguise the ad as regular news article and call it a day. Nothing stops them from putting banners on top of every page either or how about showing you an ad text before you are allowed to visit the actual content? You could even have a server-side wait time before you are allowed to move on from the ad.
Gemini just feels like rolling back the Web to an earlier stage without actually adding anything meaningful. If it ever got popular, it would devolve just the same as the Web already did, as it's nothing more than the same thing with a different paint job.
My personal take on this is that the core of everything wrong with the Web is the request–response nature of the protocol, as that allows to do a lot of crooked stuff on the server side, either intentionally (e.g. monitoring user activity, showing different content to different people) or unintentionally (e.g. causing broken links by servers going down). If the client gets plain-text or HTML really doesn't make a difference.
Another big issue is that publishing on the Web is far too complicated and expensive, which is why nobody is doing it anymore and instead relies on Facebook, Youtube and Co.
Something like IPFS feels like a much more valuable attempt as fixing the Web, as IPFS gets away from the request-response and turns the Web into a persistent data structure, that can't just be meddled with at the server side. It also makes publish substantially easier, as hosting is no longer tied to any single company, multiple people can host the same content and you don't have to pay and maintain a DNS record.
The part that is still missing with IPFS is a better replacement HTML. Stuff like pagination, image galleries, shopping carts, comments, etc. really need to become part of HTML, instead of stuff people hack together with Javascript or server side scripts. And how come we don't have an <advertising> tag yet?
Efforts in simplification should really happen on the users side, not in the protocol.
> That seems more like an artifact of it not being popular than a feature of the protocol. If companies wanted to, they could just run native advertisement on Gemini just like they do elsewhere, just disguise the ad as regular news article and call it a day. Nothing stops them from putting banners on top of every page either or how about showing you an ad text before you are allowed to visit the actual content? You could even have a server-side wait time before you are allowed to move on from the ad.
The problem that people have with ads on the web is cross-site tracking and targeted ads. Gemini protocol does indeed prevent that by restricting one request to just one response. Non targeted ads wouldn't be a privacy concern, or even an attractive option.
> The implication being that the web can be changed?
Yes - the slice you visit can be changed for you by using the right user-agent (as suggested in TFA).
> When I browse the gemini space, I'm guaranteed to not be flooded with ads, harmful javascript, etcetra
Why can't gemini space be defined by "Content-type: text/gemini" sites linking to each other and/or a limited functionality browser that doesn't render content you don't like?
> Yes - the slice you visit can be changed for you by using the right user-agent (as suggested in TFA).
It's hard to call it the web if your user agent doesn't support the web technologies.
> Why can't gemini space be defined by "Content-type: text/gemini" sites linking to each other and/or a limited functionality browser that doesn't render content you don't like?
As far as I care, it could be.
The end result would be more or less the same: people would have this new format and a new bunch of applications that aren't compatible with the web. I'm not really interested in bikeshedding minor implementation details of that level.
I'm not yet sure how I feel about Gemini (I like inline links), but this post makes me think it's starting to gain some traction. It's past the "first they ignore it" and "then they laugh at it" phases and now into the "then they fight it" phase.
Given that the mainstream stack we all are dependent on has foundational problems as OP acknowledges, language like "solutionism at its worst" feels counter-productive. Surely anything Gemini does or doesn't isn't as damaging as status quo. So why not let a thousand flowers bloom? Criticism like this feels like the People's Front of Judea (https://www.youtube.com/watch?v=a0BpfwazhUA)
I guess it depends on how big of a "they" is required -- and if this is an attack, not harsh laughter.
That said, I think Gemini is intentionally designed to be ignored; an enclave or ghetto on the Internet, depending which connotations the speaker wants to have. Its success or failure will depend on its ability to have a self-sustaining community, which is interesting mainly in that discovering Gemini requires finding it via the web at the moment. I suppose there might be a Gemini-IRC connection, too; we will see if Gemini has IRC's longevity.
If you think the web sucks, stop equating people (not companies) building alternatives with Google and Facebook. I think OP is the Judean People's Front while Gemini is the People's Front of Judea. Or something. Who cares?
The way I see it, constant complaints about the web led to Gemini, so the constant complaints about Gemini will inevitably lead to... something else.
Unfortunately, that something else will also be minimalistic, separatist and elitist, just less so than Gemini. It will probably use a strict subset of Markdown and be primarily text-focused and won't support scripting. But rather than "a thousand flowers" blooming we'll just have Gemini and "Gemini lite," because that would satisfy the needs of 99% of the potential audience for such things.
> language like "solutionism at its worst" feels counter-productive[...] why not let a thousand flowers bloom?
Because resources are not infinite. Gemini sucks the air out of the room.
> Sometimes an adage is trotted out that goes roughly like this: Welp, it's not perfect, but it's better than nothing!¶ And sometimes that's true. It's at least widely understood to be true, I think.¶ What I don't see mentioned, ever, is that sometimes "it's better than nothing" is really, really not true. In some cases, something is _worse_ than nothing.
> I think the real tragedy of Gemini is that although its existence is a reaction to a genuine problem[... i]t's another instance of a bad solution to a problem making that problem worse. This happens when bad solutions capture the attention of people concerned with and/or affected by the problem, and then they divert resources away[...] instead of allowing those resources to be better put to good use
Yeah, I just axiomatically disagree with this sort of control-obsessed, zero-sum thinking. Good ideas have always had to compete with other new ideas in addition with incumbent old ones. There is always an organizational challenge in addition to the technical one. If whatever you want to happen is threatened by Gemini, it's not good enough. If you want to wait until everyone else stops trying new ideas, please say hello to the heat death of the universe for me.
I don't know what you mean by "control-obsessed", and I'm not a zero-summer generally, but zero-summability is relevant to what's going on with Gemini. A person who spends 250 man-hours working on Gemini is not going to get that back. Likewise, a project that would have benefited from that person's enthusiasm had Gemini not existed is not going to.
> If whatever you want to happen is threatened by Gemini
I don't think you're conceptualizing the criticism correctly. This is not "B is better than A". This is a case where B never comes along because A exists. The person who would otherwise catalyze B's existence and success doesn't, because they think that A is sufficient and/or anyone they might talk to about B would just respond, "I dunno; isn't that what A is for?" You can see this with Mozilla, for example. (As a former Mozillian, that's what I had in mind at the time I wrote the linked blog post.) I have become especially sensitive to this after my experience between 2006–2013 and seeing the contrast of that time period vs Mozilla's role over the last 10 years—which is basically a black hole that keeps people from effectively organizing anything that resembles the early days of Firefox development. I recognized something similar after moving to Austin in 2014 and signing up for lots of volunteer events that were by-and-large just organized to be ways for affluent young professionals to feel like they're doing good by burning their attention surpluses, whether or not any of those events were actually a worthwhile use of those resources. See also:
> I don't think you're conceptualizing the criticism correctly. This is not "B is better than A". This is a case where B never comes along because A exists.
This risk is omnipresent in the universe. We never know what possible futures we cut off when we make a fork in the road. You don't ever step into the same river twice. However, this risk also applies to the projects you like. If we all worried about this we'd not do anything. Rhetoric like "Gemini sucking the air out of the room," "person who spends 250 hours working on Gemini," "it's like voluntourism," -- I could just as well replace Gemini with triplescripts throughout your comments.
No, you could only do that if the basis of my comment were about generic, FOMO-driven hand-wringing, where A is unbound, so substitute any A and the criticism remains true. That's not what we're talking about. The criticism involves the observation that Gemini, specifically, is bad.
If Gemini specifically is bad, you should be able to argue that without making these other arguments that apply to any new project. Convince people they shouldn't care about it. So far I haven't found that side compelling. If you concentrate your energies there, I might.
They can't be applied to any new project. I don't know why you're ignoring this, even with the clarification using the well-understood concept of free vs bound variables. "These other arguments" are a direct response to the question you posed <https://news.ycombinator.com/item?id=30068014>.
My position is something that I know you already agree with. You can't write, "Surely anything Gemini does or doesn't isn't as damaging as status quo" and think that my argument is unsound—because it's the same concept.
> If you concentrate your energies there
The last thing that we need is _more_ energy being lost to the Gemini sinkhole. That's the whole point!
I don't understand. What clarification? I don't see any mention of free vs bound variables anywhere in this thread or your links. I meant energies in describing the technical shortcomings of Gemini, and concentrating all the energy you're already willing to expend, not adding more. If you're tired of arguing, so am I. I'll step away now.
> > This risk is omnipresent in the universe[...] If we all worried about this we'd not do anything. Rhetoric like "Gemini sucking the air out of the room," "person who spends 250 hours working on Gemini,"[...] I could just as well replace Gemini with[...]
> you could only do that if the basis of my comment were about generic, FOMO-driven hand-wringing, where A is unbound, so substitute any A and the criticism remains true. That's not what we're talking about.
Gemini devs could have contributed to something else? Sure. They didn't, though.
Historians avoid counterfactual arguments because they lead nowhere. Perhaps the timeline without Gemini is worse. Perhaps it's better. Not possible to say. Best to just live in the moment.
> Gemini devs could have contributed to something else?
Not the argument I'm making. I'm observing that others' energy that gets captured (or more accurately, spent; i.e., fails to be captured for use elsewhere) as a consequence of Gemini's gravity is wasted. It's necessary to make a distinction between Gemini's originators and people sucked into its orbit.
>it certainly does not justify completely replacing existing infrastructure and standards that humans have mutually agreed upon.
It does, it makes sense. These standards led us to everything that has ruined the old web, it led us to corporate controlled Internet where few major websites aggregate 80% of the content. It led us to scroll-jacking, bot-driven, located behind firewalls and captchas websites where genuine users can't even log in or post anything.
It is why we have to use browsers that are more complex than entire operating systems just to support everything that can be done over HTTP. It led us to monopolies on web browsing - how many pages of web standards you have to implement and comply with if you want to make your own browser, just to visit google dot com? And at any time and moment all your efforts can be interrupted by security researchers who would tell you that your browser is insecure and now you have to comply with hundreds of pages of security standards as well, all because of how much can be done over HTTP. You can't even host a simple website over just HTTP without somebody telling you that you have to use HTTPS, even if you don't need it, because why not? You'll have the green icon!
And because of such monopolies, most websites also turned into "shells" that try to ruin your experience as much as possible so you go and download their app where they have more control without being restricted by browsers. And this, in turn, led most people going off the general web and staying inside their apps.
Sure, you can make your app that throws away most of HTTP-based bloat, but then someone forks and makes it support some of the bloat because they like it that way, and then someone else adds a bit more, and we are back at it again. You can't just "go back" and strip everything from HTTP, entire thing has to be done in entirely different way where there is no possibility of repeating history, because there is no other road history of HTTP can walk by.
> These standards led us to everything that has ruined the old web
That’s the entire contention, though. I don’t think the standards really had anything to do with it.
The only thing that even comes close is maybe DNS, since it ties web pages to specific organizations, so a page only continues to exist as long as an unbroken chain of custody keeps it there. But Gemini doesn’t even address that (ask IPFS and FreeNet how it’s going).
But mostly, browser vendors created the standards. And most of the current browser vendors became dominant browser vendors by using some other business to gain an edge in the market: Google used their ad business, sure, but before them it was Microsoft using their operating system to push their browser on people, and they largely just ignored the standards.
> entire thing has to be done in entirely different way where there is no possibility of repeating history
But does Gemini actually do that? What stops your client from making ad-hoc extensions to gemtext?
No amount of writing, or lack thereof, can stop evil Gemini browser vendors from just ignoring what the text says. You need something else to make sure that power players don’t just ignore the rules you set out.
You’re entirely right; the sole reason for Gemini is to exclude people. It’s more than solutionism, it’s elitism. It’s some people hiding their content to the quote-unquote “cool” kids and claiming it’s a “better” solution just because the primary clients are terminal based.
Which is why nothing of value is lost by ignoring it, IMO.
Eh, it would be really nice to have back a Web where clicking a link couldn't result in loading a page that's tracking your mouse movements while you're on it.
IOW a web that's safe and user-controlled, again. Just a damn hypertext document browser that doesn't also bundle a spying suite and hand control of it to anyone by default. We should have realized when people started putting up troll pages that broke your computer simply by using JavaScript's normal capabilities, that we needed to seriously rethink including a scripting language with all kinds of ability to act against the user just because they followed a link, as part of the Web. We didn't do that, though, and instead we got gestures at everything.
I'm not sure Gemini's the right solution, but that would be a nice thing to have, which we do not have now.
> Eh, it would be really nice to have back a Web where clicking a link couldn't result in loading a page that's tracking your mouse movements while you're on it.
It's still possible if you use uBlock, or enable JS only when you need it. You don't have to throw the baby (gemtext missing inline links!) out with the bathwater.
> It's still possible if you use uBlock, or enable JS only when you need it.
Possible, but painful. Take HN's official search engine, for example: https://hn.algolia.com/
I open it. "This page will only work with JavaScript enabled" sigh Accept JS from hn.algolia.net and the mangled cloudfront domain (and make sure not to accept ravenjs.com). Possibly cdn.jsdelivr.net if I want to spare a click later, because I don't know if it's useful (it's not). Reload. Now the results are blank. Ah, it makes XHRs to telemetry.algolia.net, a mangled algolia.net domain, and three mangled algolianet.com domains. Think for a second. Accept the latter too (actually, either is enough, but I don't know that). Reload. Oh, it works now.
Luckily there isn't a privacy consent popup I need to reject (or block with uBlock).
It's exhausting to do that almost every time you visit a new website. And with https:// URLs, you don't know ahead of time how many rounds of accepting you'll need to do before opening.
I’m not that familiar with Gemini, but implementing search functionality via Gemtext is impossible right? Would the appropriate comparison not be pages that are similarly static to Gemtext?
I’m not convinced that an entirely new protocol is necessary. How is the Gemini experience different from just browsing all webpages with JS disabled, or if you want to go even further, using a web browser that just doesn’t implement JavaScript?
I think abut it kind of like this (as someone who has never used Gemini):
There are all sorts of uses for cryptographic proofs; e.g. you can sign a document and prove that you were the one who signed it, or you can do a bunch of double-SHA hashes and prove ownership of a Bitcoin, you can prove via a SSL certificate that the content that claims to come from somedomain.example actually does, etc etc.
But there's no way, via currently existing protocols, to deliver a document with a proof that the document does not contain code designed to track the user or that the document will display properly without running arbitrary scripts.
Gemini is bikeshedding, yes. But it solves this social problem - a problem of expectations. You can't make a webpage with tracking scripts and deliver it over Gemeni because there's no point. No one who received the document would actually execute any of the scripts. So Gemini succeeds at creating a little insular community where tracking and web analytics are not just forbidden, but impossible.
You can do that without making intentions verboten altogether. I'm happy to see someone's weirdly styled homepage or tumblr blog, and Gemini strips too much away for it to be worth it.
> It's still possible if you use uBlock, or enable JS only when you need it.
I've been blocking JS for a damn long time but it just breaks so many things. And if you selectively block JS with umatrix, that also breaks things sometimes.
I gotta say I'm really fed up with playing whack a mole. I'd like to find a corner of the internet where things just work and don't have antifeatures.
the second definition you didn't list is "the attitude or behavior of a person or group who regard themselves as belonging to an elite", which is the meaning used in this case. I don't agree that's elitism either, just wanted to point this out
Gemini really doesn't exclude anyone. It's not exactly difficult to use a Gemini to HTTP proxy and you don't lose anything except for access to some weird experiments with the client certificate features.
I mean, ignore away, but I don't think its fair to say that the sole reason for Gemini is exclusion of people.
Meh, everyone doesn't have to cater to the lowest common denominator all the time. Besides, I can't say I find the geminispehere very elitist or excluding after browsing it on and off for a year or so. There are no gatekeepers apart from accepting the technology itself.
Gemini folks maintain Gemini for themselves using their own hardware and their own time, so I think they can do whatever they think its best for them.
If they want to strip HTTP and remove all cool features it has, it's their problem, not ours.
If they want to treat content as text only, again, it's their problem, not ours.
If they want to publish their content as Gemini only, it's their content, not ours.
If they want to bikeshedding about internet protocols, it's their time, not ours.
Now, about Gemini folks being elitists and intentionally excluding people with their protocol, take a minute to think about what tech companies are doing.
Tech companies and many other companies that rely heavily on tech are making our society dependent more and more on online services, selling all kind of products and services that need an internet connection to work, even for basic features.
In a world where half of population does not have internet connection, and geek folks feel excluded by a protocol just because they don't like it, how can we name what tech companies are doing, if it is not elitism and exclusion?
"tech companies" are not a protocol, nor are they a network built on that protocol, so the comparison here is specious at best. HTTP doesn't ban anyone, nor does TCP/IP nor does HTML, the internet or the web.
It seems to me like the author is experiencing how the web itself feels if you try to step off the browser upgrade treadmill or use an older computer. A website refuses HTTP and you don't have recent TLS certs? Tough luck, pal. This page dynamically loads content with JS? Buzz off, guy using Links. Can't pay your online bills or join a video chat because you aren't using Chrome? Get with the times, grandma, and surrender to the new flesh!
There will come a point- and for some things we're already past it- where there is exactly one viable web browser, and it is controlled by a company that sells ads. This is why Gemini exists.
The post is a broad, and IMO, valid criticism. Though I'm not sure having a different protocol is a _huge_ problem. If Gemini ever got popular both backends and clients would add support.
I think the real issue, and the OP touches on it, is the text/gemini format. There's a ton of content that would enrich Gemini (and that fits Gemini's mission) but that simply isn't usable in text/gemini. Specifically, I'm talking about inline-links (though that isn't the only issue).
I wrote a server in elixir, and tried to convert Wikiversity's Introduction to Programming (1) both manually and automatically. I found the output unusable and couldn't come up with a case where I'd ever prefer/recommend it.
> There's a ton of content that would enrich Gemini (and that fits Gemini's mission) but that simply isn't usable in text/gemini.
That's kind of what I've concluded too. I like the mission of gemini (and I believe the web is fundamentally broken and unrepairable), but I do feel like text/gemini flings way too far in the opposite direction. It's not a new pattern, I see this all the time: thing X is too complex and bloated and user-hostile, let's make a suckless thing Y. Y is minimalistic to a fault.
Making something that's too minimalistic or too complex is easy. Finding the right balance and making something that is small and simple yet still capable of covering wide use scenarios is an engineering challenge..
Since I agree with gemini's mission, I'm not going to complain too much about it. It's not exactly what I want, but if it encourages people to write content and deliver it in a form that is guaranteed not to have web's problems, I'm all for it, because I do stand to benefit from it.
I don't often use Gemini, but here's the thing that I think this misses and that Gemini does well:
If I click a link in a Gemini browser, I know for sure that it will not start loading and running a ton of resource-intensive scripts. I know that it will not start autoplaying audio or video. I know that it will display fine in plain text, every single time - so it works in my terminal, in Emacs, on my phone, etc. And I know it will be fast.
Yes, I can use NoScript, uBlock Origin with comprehensive lists, etc. and start to approach that level of simplicity... but every other link I click will likely be broken in weird and wonderful ways. In Gemini, they all work (provided the page still exists, of course). I think there's some value in that.
> If I click a link in a Gemini browser, I know for sure that it will not start loading and running a ton of resource-intensive scripts. I know that it will not start autoplaying audio or video. I know that it will display fine in plain text, every single time - so it works in my terminal, in Emacs, on my phone, etc. And I know it will be fast.
With a `text/gemini` content type over HTTP/1.1 you could just use cURL, wget or fancier stuff like w3m/lynx and get the exact same benefits but would not need to install yet another internet facing tool that may even get things wrong security wise (even with simple stuff a possibility).
Gemini's most distinguishing feature seems to be that it is considered feature complete. If everything goes as planned, even decades from now building a usable gemini client should be easily doable. Completely unlike web browsers which started off with fairly simple html but picked up a lot of bloat over the last few decades. This alone might make Gemini worth it.
I think the author is analyzing Gemini from a technical / engineering perspective, and of course, it doesn't make any sense.
I think Gemini lack of compatibility with the Web is its main feature, and I think people using Gemini want to create a separate new "community". The spirit is "we lost the web, it was taken over by corporate interests, let's accept it, leave it and create our new thing" (you can replace "corporate interests" by "normies" for some people I'm sure).
I think this is all fine, creating new communities/clubs/whatever when you are not happy with existing ones is good, maybe you'll be alone in it, maybe not, good for you.
I think the main problem I see here is that, if I understood properly, a Gemini link was shared on a web page. I don't think it is okay. You want to have your own private club, you got it. But please leave other clubs alone. It's perfectly okay that you don't want to communicate with me. But then don't.
"Why couldn't one simply build on top of existing HTTP infrastructure, throw away all the baggage and instead implement a new Content-Type, which existing browsers then could parse?"
Because escaping the annoyances of the "modern web" may require escaping from the "modern browser". The changes users want to "modern browsers" will never be made. The vendors of these programs do not answer to users. They answer to web developers and advertisers. These are large, complex, insecure programs usually controlled by organisations that seek to profit from online advertising. The advertising focus leads to complex web pages. Not the type of simpler pages that some users want. (NB. Not "most", but "some".) Gemini, because of its limitations, allows users to retrieve resources without the need for one of these "modern browser" programs. If a new Content-Type was added to HTTP, what is the likelihood that other parties outside the "modern web browser" cabal would write small, simple, alternative browsers, e.g., aimed only at this Content-Type. Look at the market share, i.e., available selection, of web browsers. It is not diverse.
Whereas, writing a Gemini client is dead simple. A Gemini browser cabal where users have only a few choices and they are each controlled by corporations is unlikely.
Asking web developers to "please make simpler web pages, thanks", when the "modern browser" allows for complex pages and integration of advertising is not a succesful course of action. Most of these web developers answer to advertisers or to employers who answer to advertisers. They are not going to ditch the user annoyances, they are going to seek profits. Gemini seems to address this problem by making advertising difficult. Without the "modern browser", the possibilities for advertising are limited.
Similarly, asking web users to "please use a text-only browser", e.g., Links, when so many web pages try to use "modern" browser features that enable complex web pages is probably not a successful strategy for many users either. As a long-time Links user, that strategy has worked for me, though.
The only complaint I have about Gemini is the absolute requirement for SNI. Not every IP addresss will necessarily be hosting multiple Gemini sites. Under the current protocol, even addresses hosting only a single site must require SNI. That makes no sense. It serves no purpose. It should be optional not mandatory.
>A Gemini browser cabal where users have only a few choices and they are each controlled by corporations is unlikely.
This is the key point. It is like the idea that you don't need the fastest person to escape from a chasing bear, you only need to be faster than the slowest person.
A user on the conventional web, is the slowest person here. And thus the success of Gemini browser is dependent on the existence of the slower person, that is, a large number of users available for exploitation, on the regular web, so that the business will leave the Gemini users (and the likes) alone.
HN actually implements this idea by remaining minimal. But even this place got overrun by shills these days, which means such piecemeal strategies will not work. So it makes perfect sense to do all the way and use a different protocol altogether..
Many 'rules' (like one request -> one file) create limitations in Gemini that prevent it from being capable of web-style corporate behaviour, so even if business isn't content with regular web users and doesn't leave Gemini users alone, their options and the damage they can do is hampered (unless co-opting the entire movement).
I think the protocol being designed around ensuring it's difficult for sites to exploit visitors is a core goal that gets lost in the blinding document minimalism that's hits you when first encountering Gemini, OP didn't seem to get it.
> if business isn't content with regular web users and doesn't leave Gemini users alone, their options and the damage they can do is hampered
I am afraid you are limiting the scope of "exploitation", as enabled by Internet, to those things done by Ad companies like Google family.
But I was mostly referring to the use of Internet for all kinds of exploitation. Think of a seller on amazon putting fake reviews in the web, or deleting actual negative reviews. They won't bother with the Gemini users because there is enough victims on the real web. But if all users are using Gemini clients, then they will be writing fake reviews in Gemini domain.
That is smaller it remains, the less penetration by big business and selfish interests, and thus the more valuable it is, for the person who wants to find real, genuine information.
This also applies to real discussion. People are going to speak freely in Gemini, because there is less chance of encountering an army of shills.
Keep an eye on sites that sell gemini identities with "reputation". Once those start popping up, you ll know its time to move on to something different.
> Not the type of simpler pages that some users want.
And yet most of these users don't use lynx...
> Only complaint I have about Gemini is the absolute requirement for SNI. Not every IP addresss will necessarily be hosting multiple Gemini sites. Under the current protocol, even addresses hosting only a single site must require SNI. That makes no sense.
If you're only hosting 1 site, isn't the privacy leak negligible because there is 1:1 mapping from ip to domain so an attacker caneasily determine it.
Besides, in a system playing fast and loose with pki, its more like a do not disturb sign than an actual lock.
"If you're only hosting 1 site, isn't the privacy leak negligible because there is a 1:1 mapping from ip to domain so an attacker can easily determine it."
What if it is a company serving advertisers, not an "attacker".
Not sure why this myth of effortless, reliable translation from IP to domain name in "real-time" exists amongst HN commenters. Show us who is doing this for the purposes of advertising and how it is worth the effort and can be relied on. Even if this were possible, it still does not justify sending a plaintext hostname over the wire when it is not necessary to retrieve a page.^2 Not every site requires SNI (yet "modern" browsers send it anyway).
Tell us how to reliably^1 translate any IP to a domainname at the same speed as one can sniff SNI, and with no extra effort. Assume the use of TLS1.3 so one cannot simply examine a plaintext certificate sent over the wire for a CN or SAN. Then show us where and how this is routinely being done by various companies selling online ads or ad services.
1. PTR will not suffice
2. This is like arguing that because it is theoretically possible for a third party examining traffic to discover some private information through a process that requires referencing additional sources, the user should therefore broadcast the information with every request, despite that it serves no purpose for the user to do so.
Sniffing SNI is common practice.^3 Performing effortless 1:1 mapping from IP address to domain name for the entire www in real-time, for advertising purposes, is not. As long as browsers send SNI with every request, there is no need to do that, even if it were possible.
3. Encrypted Client Hello (ECH) will eventually prevent it. See https://defo.ie
There is no harm in Gemini making SNI optional instead of mandatory.
> Show us who is doing this for the purposes of advertising and how it is worth the effort and can be relied on.
Why would an advertiser do this? Advertisers are typically in leauge with site operators. Site operators just tell them this data (maybe with rare exceptions like superphish). Advertisers don't do this because they don't need to.
Advertisers are not the adversary tls is meant to thrawt. You don't use the lock on your door to thrawt the person who you invited in and opened the door for.
> Not every site requires SNI (yet "modern" browsers send it anyway).
Its difficult to tell if a site needs it or not at the stage where you send it. The current solution seems to be custom dns records (e.g. what ECH is doing last time i looked)
> Tell us how to reliably^1 translate any IP to a domainname at the same speed as one can sniff SNI, and with no extra effort.
The traditional answer is to just sniff the dns traffic.
Otherwise (e.g. if using DoH) just create a db of popular sites you care about. This is not trivial, but still quite easy and the easiest part of the attack being discussed by far.
Not an advertiser necessarily but any entity or person that can "monetise" the data collected. The collector might use the data itself, it might license, sell or transfer the data, it might provide services that rely on the data, who knows. Some users may not want to voluntarily share this data when they derive no benefit from doing so. We do not have to guess all the possible ways, besides locating the applicable TLS certificate, that the data might be used before we can honor the user's wish that this data not be sent in plaintext where it is not needed for choosing the certificate.
AFAIK, sniffing SNI is already used for the purpose of censorship by some countries. This has been published. It would be ignorant to think that this is the only purpose for which such data might be used, or that any purpose would always be non-commercial and unconnected, directly or indirectly, to web advertising. As the use of DoH increases, sniffing SNI would seem an easy substitute for sniffing DNS.
1. A real-time list of every domain visited by a user.
"You don't use the lock on your door to thrawt the person who you invited in and opened the door for."
For some users, advertisers are not an "invited person". What is more, companies like Google have attempted to force the use of TLS for every site, even ones where, in the user's or site operator's opinion, TLS is not needed.
"It is difficult to tell if a site needs it or not at the stage where you send it."
But this is not an argument for sending SNI by default, even where it is not needed.
A TLS proxy can be configured to distinguish sites that need it from sites that do not. This is what I do. The default configuration is to not send SNI. This makes sense because the majority of sites I visit do not require it.
As such, from where I sit, the solution chosen by modern web browsers is to prioritise websites that use CDNs that depend on SNI. The side effects for users of indiscriminantly sending SNI, i.e., sharing every domain the user visits in plaintext on the wire, are not as important as reducing costs for those websites using TLS and CDNs. Arguably, SNI is for the benefit of websites and CDNs at the expense of users. (Hopefully ECH will obviate this tradeoff.)
"Otherwise (e.g. if using DoH) just create a db of popular sites you care about."
According to this answer, 1:1 mapping is not an equally easy alternative to SNI. Sniffing SNI is "trivial" and works for any https site, whereas 1:1 mapping through a database is "non-trivial" and only works for "a selection of popular sites [one] cares about". SNI makes the task of monitoring a user's web use easy. If SNI is not available to sniff, then the task becomes more difficult. This is the point.
Sniffing SNI is easy. The theoretical 1:1 mapping alternative proposed by HN commenters is more difficult. This is the point. What is easy and reliable for all www sites versus what is more difficult and unreliable for all www sites. The point is not what is possible^2 and what is impossible. That is the red herring diversionary argument tactic that HN commenters defending gratuitous SNI like to use.
2. It is possible to avoid DNS altogether and to only send SNI when it is required. I have been doing this for years. Gather bulk DNS data and load the data into a forward TLS proxy that stores domain:IP addresses mapppings in memory and does lookups in real-time as requests are received.
I find it ironic to read an English blog post deriding something for being inaccessible hosted on a domain containing Katakana, something most English keyboards would not be able to type. I cannot tell if the author is Japanese, but I am not sure it matters since the content appears to be exclusively English.
> And it's easy to set up on Windows. I can toggle English and Japanese input with Windows Key + Space.
It's virtually impossible to set it up if you don't use a QWERTY layout. I have a horrible pile of hacks that more or less work for now, but are being removed in windows 11.
I think my comment still stands. Without extra configuration, there is no way to type the characters directly from an English Keyboard. And most users don't have that configuration setup. Even if you can type in romaji, it still requires the knowledge of what combination of latin characters correspond to to a Katakana character. I don't think you should have to enable international keyboards to access a domain.
My Firefox browser shows the punycode as well, potentially something with your configuration? Only Japanese characters that Firefox renders is the name of the tab, URL itself is using the romanized characters.
The author isn't considering how lack of features can be useful. Just like the lack of gotos in a language allows you to make many assumptions about a program, Gemini lacks just about every mechanism for fingerprinting and tracking that's become an inherent risk of accessing web pages, hugely reducing surface area for attacks.
Even lack of styling with every page looking about the same can be considered a feature, everything looks familiar and we aren't inconvenienced by bad design decisions.
I went down the Gemini rabbit hole last week (https://taoofmac.com/space/blog/2022/01/17/0800, and even coded my own minimalist server for it) and tend to agree that both its protocol and markup are oversimplified technical solutions in search of a problem, although I wouldn’t cast it so harshly.
It does feel like escapism/subculture of sorts, and I do believe that retreat from the Web is one of the underlying motivations, but (maintaining a relatively minimalist Markdown-backed site myself) I agree that there are no real technical reasons for it to exist (even Markdown support is something that is relatively easy to agree upon across browsers).
What I do like about it is the Lagrange browser, and the extremely legible reading experience—-I consume most of my content via RSS feeds and Reader Mode on websites I visit frequently, and I like the uniformity and predictability of the layout, as well as the lack of ads and distractions.
I’d be happy enough if Lagrange’s unfettered visual style became a sort of “forced Reader mode” in standard browsers…
Admittedly I up voted this not because I agree with it, but because I want more people talking about Gemini, both for and against it since it leads to healthy and constructive debate on a new technology.
I wondered the same thing... why invent a whole new thing? Making a web browser capable of displaying the Internet circa 1999 is not trivial but it is far, far, far easier than building a modern one.
It's about building an ecosystem. Its way easier to build a community of users and content which display on an 80 char terminal when the protocol only supports it and every link links to more content which follows the same restrictions.
IMO Gemini is not a solution but a toy. Nothing wrong with that, its fun to have toy protocols and communities, but it isn't useful to many.
I think because the inevitable result of that is that you constantly land on broken pages, just by following links, even if you start in some kind of enthusiast area that caters to such a browser.
You could mitigate that by only linking to pages that implement the HTML subset deemed valid. And instead of linking away to another domain, you could syndicate those pages on a centralized domain so to prevent loading dangerous or slow content on the open web. We could call it something like Accelerated Mobile Pages.
I have sincerely not idea. I skipped over any https talks because it was not fun to look at and assumed folks would have a dependency. There is a list of client/server in a bunch of mainstream languages, did those people had to implement TLS? That sounds like work.
The distinction between devs and users is less sharp when everything is as simple as the Gemini ecosystem. And that is precisely what they want: technology that is resilient and does not (and will never) depend on a caste of specialists and large corporations such as Google, like the web does.
Nah - Gemini FORCES people to use a very strict small surface area. Building on top of existing tech you would need to trust people to do it right, and that ain't something you trust.
It's similar to how, here in HN comments, you cannot embed a giant 300MB javascript advertisement monstrosity. Or even just an image.
If people could do that, then the space would be destroyed by malicious actors.
What if we eliminate HN and do the communication P2P? Where author hosts own content. How can I as author still guarantee you as reader that I won't do that?
I think if you want to further limit yourself than Gemini yet maintain some of the semblance of the existing modern web, you can go the Beaker Browser [1] route.
Yep there would always be the temptation to say "well I need just one cookie" or "well I need just this bit of javascript" Gemini doesn't give you those options.
It’s also make the protocol considerably easier. I read it in two small settings. It’s surprisingly digest and make you want to start implementing a server or a client for it.
I think it's only a matter of time before Gemini does that. The moment it's actually used by people who aren't interested in the underlying tech, someone will fork the browser and include either video support or scripting of some kind. Or they'll say, maybe it should display PDFs too - and that will bring the JS engine in as well.
I think this is an interesting argument, but my question is why can't it just work in the browser I have with an extension? No matter what people say, modern websites with JS and HTML and such aren't going away. I can't just stick my head in the sand and only use this special browser since I use too many things that make my life better in meaningful ways that use the traditional protocols. Many of them wouldn't be possible using Gemini (I assume, I haven't dug in much).
That means if I want to use Gemini I would have to add a whole other browser to my computer and switch between the two while browsing which is super annoying. At that point I just won't bother.
What is the end game here? Do they want all websites to switch to Gemini? Do they expect people to use two browsers? Are they saying that anything that can't be built on Gemini should be a Native App?
I use Firefox. When I click on a gemini:// link it opens it in Lagrange. Seems pretty seamless.
If you're not going to install Lagrange, I'm doubtful that you're going to install an extension either. This isn't about you personally, it's a statistical judgment about the market segment you're signalling yourself to belong to. People reluctant to install native software are also unlikely to install extensions.
Having a separate app provides some additional protection compared to an extension that can look over your shoulder at everything you browse. Lagrange runs in a separate process entirely, and it can only snoop on the Gemini websites I visit.
Me, I go to Gemini (and Mastodon, and the tildeverse, and plan.cat, and twtxt, and, and..) precisely because it has no hope in hell of adoption. If someone's posting there they're intrinsically motivated[1]. That's the stuff I want to look at. When I'm in small pools like these I'm less legible[2] to marketers.
> It feels like the people working on/running Gemini infrastructure don’t want to actually solve the issues with the modern day web and instead just wanted to be different,
Many of Gemini's adherents can be described broadly as left-anarchists or left-libertarians. Their assessment of the problem is that the web has been captured by large, centralized corporations. Their alternative is human-scale technology and human-scale communities: a protocol so brain-dead simple that anyone with a bit of technical know-how can write or run a client or server, and anyone can write gemtext. It's a philosophy that users should have a close relationship with and understanding of the tools and platforms they use, which should be built and run for free (one can think of many real-world organizations analogous to this). Could you build these tools and communities on HTTP? Sure (and there is a wide overlap between "small web" communities on HTTP and gemini), but it's a more radical separation to build an entirely separate "place" for them.
> Bottom line is, if you agree that the modern web has become an awful place, let’s work on changing that for everyone, instead of abandoning it like a bunch of billionaires trying to escape to a different place
This is a bizarre analogy, because, unlike the other alternatives that the author presents (e.g. blockchain), there is no money or corporate interest in Gemini. But I do agree that Gemini is, to some degree, escapism, it does not confront the institutional problems of the web. But those problems won't be solved by any protocol -- not Gemini, not Dat, IPFS, not some future imagined perfect protocol, and certainly not blockchain. They are solved by politics: challenging the institutions that control the web and its infrastructure and the policies that they make. Gemini is, to me, at least a nice reprieve from the web as it is, and a demonstration of what it could be, unsullied by the drive for profit and domination that fuels it today.
> Many of Gemini's adherents can be described broadly as left-anarchists or left-libertarians. [...] Their alternative is human-scale technology and human-scale communities: a protocol so brain-dead simple [...]. It's a philosophy that users should have a close relationship with and understanding of the tools and platforms they use, which should be built and run for free (one can think of many real-world organizations analogous to this). [...] Could you build these tools and communities on HTTP? Sure [...] but it's a more radical separation to build an entirely separate "place" for them.
I used to consider myself a left-anarchist, though now I might better be considered a post-left anarchist nowadays. More radical, perhaps in PLA style, is to do away with this silly idea of "human scale"; we have computers, so let's do some computing, and get something interactive like the Web, but from an axiomatic design that is implementable in your lifespan, and ensures privacy and security properties from such axioms.
"Many of Gemini's adherents can be described broadly as left-anarchists or left-libertarians."
Many? How many? Which capsule? Who? How were they identified? What were the numbers for centrists and people who swung the opposite way politically? How was it defined?
I don't want to speak for any individual, but it's pretty clear if you just go through the blogs/social media feeds of any of the major figures in the community. I didn't say it's the only perspective, but it's obviously prevalent
>This is a bizarre analogy, because, unlike the other alternatives that the author presents (e.g. blockchain), there is no money or corporate interest in Gemini. But I do agree that Gemini is, to some degree, escapism, it does not confront the institutional problems of the web.
I don't think the analogy is super bizarre. The Left-anarchist milieu, building its own separate spaces, unschooling their kids, living in their Californian suburb and going to Burning Man has always been more thoroughly bourgeois than even the worst cutthroat entrepreneur.
I can imagine a 40 year old retired tech guy living off his stocks raising his own chickens, self-declared left-libertarian socialist going "Oh, you use HTTP? That's nice. Corporate? I'm beyond it. I blog on Gemini". At least big corporate makes stuff ordinary people can actually use.
> I don't think the analogy is super bizarre. The Left-anarchist milieu, building its own separate spaces, unschooling their kids, living in their Californian suburb and going to Burning Man has always been more thoroughly bourgeois than even the worst cutthroat entrepreneur.
I would not describe myself as a left anarchist, but I think that this is an extremely uncharitable analysis of a much more diverse and serious group of people that don't really have anything to do with the things you listed.
> Many of Gemini's adherents can be described broadly as left-anarchists or left-libertarians.
That's interesting. People that I know (not personally, but the names I recognize) that are the most vocal about Gemini are self-proclaimed socialists. They want complete elimination of corporations and capitalism all-together, so ostensibly it might appear as 'left-libetarian' but quite the opposite. If we put on the liberatarian lenses, ideas of gemini are pretty cool but like many liberatarian ideas, they're impractical and often rooted in more emotion-than-substance. I admire the clean-slate approach sometimes because it gets rid of the cruft that we've built up over the years. It allows new tooling to be made with fresh eyes and hindsight. Think of it like the internet shedding off snake-skin.
The practical engineer in me says "We need to reinforce robustness, but also allow mad-scientists to do some wild experiments".
you know, it is possible for a simple piece of technology to ideologically appeal to more than one named region of whatever arbitrary political identity matrix one chooses to view the world through.
Having a new protocol has value for the same reason HAM does. You hope that the character of the content itself will be interesting enough to be it's own subculture.
Unfortunately it looks somewhat nontrivial to use and unnecessarily different from everything else without offering any true new features. All it has is the content and culture.
It's also yet another privacy at all costs project. Which is fine, but I feel like that's all anyone does anymore and they've forgotten all other innovations.
Still, the long form Web1 content and old forums were wonderful. Is there anything I should be looking at? What's the coolest Gemini content? How do gemini users like to communicate?
Communication, there's an informal way of addressing other gemlogs using 'RE: <title>'. It's fraught with issues, as I have discussed. There's IRC (#gemini on tilde.chat) and Usenet (comp.infosystems.gemini) now that the mailing list (itself a potent source of drama) has gone to the great bitbucket in the sky.
Since there are apparently a few users here: What server would you recommend to get started? Any small tutorials on setting up a gemini site? Any links to gemini sites (maybe your own) you want to share?
> The question here shouldn’t be why not to use a subset of HTTP and HTML, but rather, why not build on top of HTTP with a different markup layer other than HTML. We have APIs using HTTP with JSON instead of HTML, for example.
As you can read in the Gemini FAQ, the reason that Gemini is not compatible with HTTP is that it should be possible to know that when you browse Gemini content, you will stay in gemini content.
If Gemini were compatible with HTTP, then when you make a Gemini website, you know that many people who use it probably are using HTTP browsers, so it will be easier to mix Gemini with HTTP links which means that Gemini users will be more exposed to the type of content that they were trying to avoid by using Gemini.
If Gemini instead is incompatible with HTTP, it will be harder for site owners to put a HTTP link on their Gemini site because they don't know how the user's client will handle that.
Maybe the best solution could have been something that makes it easy to get in but not as easy to get out.
I mean something like it could be possible to have a HTTP link to a Gemini site that would work in all browsers, but that Gemini site could not have a HTTP link.
I just don't know how you would make that.
Is there actually a good way for Gemini as a protocol to make linking to an HTTP site harder? I don't think the protocol actually limits the type of link on a Gemini page, and if I click on an HTTP link in an app my OS will just pop it up in a web browser without a problem. Yes I'd know the app changed, but I'm not sure that would really help guard me from following a link to the current web
It is common for Gemini clients to represent HTTP links differently from Gemini or Gopher links, so the user can make an informed choice before clicking.
Same as how browsers compete with new HTML tags and such - CommonMark + few extensions could be standardized) with others left to the browsers to decide.
HTML inside markdown is anyway legal as per the standard, so anyone could use <table> for anything complicated anyway.
Such an implementation would also lead to users having more control and say in how the web pages look (similar to Gemini), by letting you read most content across most websites in fonts, styles, and colors of your choosing. This aids a lot in accessibility as well.
It's not prevented by being its own protocol, only delayed. HTTP was pretty simple at the start, then it got popular and more features got added to satisfy more users. There is no law of nature or humans which prevents Gemini from being hijacked by others later if it becomes popular enough. And if it stays niche, then the current (developer-heavy) community can keep it as simple as they want.
There's something to be said about limits. Interesting things happen when you set limitations. The Oulipo group, made up of writers like Perec and Calvino, would intentionally set limits in order to spur creativity. What's the harm in having a web protocol like Gemini attempting to do the same thing? The modern web is bloated, captured, no longer uncharted land waiting to be explored. Why not refuse and set off into the waters of Gemini? What do you have to lose?
I just wrote a "Gemini is a little gem"[1] inspired by the OP article. I think that most criticism of Gemini kinda misses some important aspects of the ecosystem and ethos behind the protocol.
My understanding was that making it difficult to connect to, in order to keep out people who wouldn’t/couldn’t meet the bar, was an intended benefit and not an accident (whether that’s a good idea or not is an exercise for the reader, but the premise of this article seems to assume that it wasn’t the intent.)
What author hints at is a view of
Gemini community as ideologically monolithic
Internet Amish, refusing technology and
standards because they're opposed to
modern state of things: the opposing(author)
view is monoculture of Hypertext Web
subjugating all protocols will eventually dominate due user-friendly
and flexible nature of web stacks.
Its obvious Gemini isn't a tech solution
but a rebellion against corporate web
dominance that intentionally cripples
web experience for greater user control, but author tries to compare it
with familiar web sub-formats that are
characteristics of content: Gemini
isn't about content, its a form of
interaction that prescribes certain
limits that Internet Amish will feel
comfortable with, without introducing
too much overhead(like Tor and I2P).
Indeed, one could at least argue that gopher was a simpler protocol, gemini makes short work of that, too, and leaves it to be only an incompatible and less capable https protocol with some other convention for the markup.
I entirely agree that there's no good reason not to use HTTP[s] as a transport for textual information, which is kinda what it's made to do anyway..
I'd argue against a gemini content-type, I'd get as far away from the association with gemini as possible.
text/feather is a better content-type, it makes it clear that it's text, with a feather-light structure imposed on it.. That structure should probably be very little more than support for line breaks and hyperlinks.
OP is correct. HTTP/1.1 requires the Host header (1). The client must send it, and the server must reject requests without it. Some implementations simply don't follow the standard. For example, you can test this on twitter.com, which appears to be correctly following the specification.
OP is only correct if using HTTP/1.1 and not 1.0 or 0.9.
I’ve done plenty of telnets using 1.0 without the need for the host header. Just depends on if the server is a dedicated server or running a bunch of hosts (and the default host isn’t the one you want).
My apologies I was being way too pedantic. You're obviously right about the standard and that some implementations do not follow the standard. I'll even admit complete defeat since I must have skipped over this part, "this server accepts." I was really just objecting to the phrasing of "bare minimum" largely because I'd encountered implementations that do not match specification. But also, the part, "for me to request the website from it" does not quite match the language of the rfc. Therefore, my argument has been reduced to "screaming into the void" is a way to request the website from a server.
You can build a minimal browser that uses HTTP and only renders special markdown files. You can use the Accept header to send special markdown to these minimal browsers and minimal HTML to normal browsers. A new protocol is unnecessary to accomplish the goals you mentioned.
It pretty much would. You couldn't link to the HTTP-Markdown-Web from the HTTP-HTML-Web (normal browsers don't and won't support Markdown) and you couldn't link to the HTTP-HTML-Web from the HTTP-Markdown-Web (the Markdown browser doesn't and won't support HTML)
(Almost nobody would vary the response based on the Accept header. Besides, if they did, you might as well just set an X-No-Ads-Please header and send HTML in both versions)
I don’t think you fully understand what I’m proposing.
You create a set of conventions that “Gemini compatible” websites conform to. This will allow your site to work with small Gemini-only browsers. Those conventions mostly amount to sending markdown for URLs when the user agent sends “accept: application/x-Gemini-markdown” (or whatever the mime type would be). Web servers send the markdown to Gemini user agents while they send normal html to browsers that don’t send the Accept header.
To signify that the site is a “Gemini site” at the URL level you can create a convention that the url must end in “.gemini” e.g. https://foo.com/blog.gemini
> (Almost nobody would vary the response based on the Accept header. Besides, if they did, you might as well just set an X-No-Ads-Please header and send HTML in both versions)
I don’t see how sending a different file based on an accept header is any more technically unrealistic than creating an entirely new technology stack.
Your "Accept: application/x-Gemini-markdown" might as well be "Accept: text/html-without-ads-or-scripts-please" and then there is no need to write an additional document parser.
Well the html format does allow for JavaScript, images, etc. Not to mention it allows for arbitrary document hierarchy. it would be a large cognitive burden on users who are publishing content to remember what is the gemini-supported subset of html that is allowed. It would also be hard to standardize the subset and prevent the subset from growing.
What Gemini got right is that simplifying and limiting the publishing format prevents bloat.
That's a massively more complex implementation for no benefit.
Also it's definitely not enough to set a HTTP header. Need to use a different schema so that the restraints on content can be expressed in the URL of a link.
A key feature of Gemini browsers is to visually distinguish between links to Gemini:// URLs and links to HTTP:// URLs.
> That's a massively more complex implementation for no benefit.
It’s barely complex. It’s a standard dispatch on the accept header, all major web servers handle this out of the box. Also it has the significant benefit of working with existing browsers. On what basis were you making that statement?
> Also it's definitely not enough to set a HTTP header. Need to use a different schema so that the restraints on content can be expressed in the URL of a link.
You can accomplish the same URL transparency by using a file extension convention. E.g. “https://foo.com/my log.gemini”
> I meant that an implementation of HTTP is massively more complex than an implementation of Gemini.
The subset of http necessary to support Gemini-style requests is really not complex at all. I’ve written basic http servers in hours if not minutes.
> The extension shouldn't be `.gemini` it should be `.this-is-not-malware-i-super-pinky-promise`
If you are someone who uses what I’m calling “http-Gemini” because they are worried about malware then use the stripped down “http Gemini” browser. For the rest of the world who is already comfortable with potential malware in their links, there is no change. The control stays in the user’s hands.
Gemini is so silly. The logic is basically as follows:
- people are using web features in a way we don't like. They could do it in the way we like but that's unpopular.
- instead lets make it impossible to do things that we don't like in our system. But people have to voluntarily use our system. For some reason people will use gemini voluntarily even though they refused to voluntarily use the web in a way we approve of in the first place.
- People are using web features in a way we don't like. They could do it in the way we like but that's unpopular. Some do but it's impossible to divine it ahead of time before you follow a link and see what horrors lie behind.
- Instead let's make it impossible to do things that we don't like in our system. People have to voluntarily use our system. By virtue of having their content in our system, our users are guaranteed a comfortable experience.
I went down the Gemini rabbit hole 6 or 8 months ago and didn't really understand the solution it offered, but you're really missing the point here.
The idea isn't that consumers are using the web in a way that is disliked. The web is run by and for large corporations legally bound to do what is best for shareholders. The web allows those corporations to take advantage of consumers in ways that the average user doesn't understand and often wouldn't knowingly consent to, and in the process opens the door to entire classes of vulnerabilities that wouldn't otherwise exist.
Gemini may include a few unpopular assumptions and limitations, but a consumer never has to worry about any content on Gemini being vulnerable to script injections, malicious spyware,
or site silently mining crypto on your hardware. Tracking is still technically possible, as mentioned in the OP article, but in a very limited sense and only really at the level of page requests.
Gemini may be an over correction, but it at least starts (or continues) the conversation of whether a limited feature set is the most effective way to fix so many of the problems on the web today.
I think that this premise is wrong. The modern web does not suck. FB is usable by my grandma (at 89 she posts semi-regularly!), but there's no way she could understand how to set up an IRC client, how to telnet into servers, or how to write markdown.
Of course, there's a lot of bad and a lot of baggage that comes with companies like Facebook and Google, but it's important to keep some perspective: there's also a lot of good. Gemini would literally make the web unusable for my grandma, and probably my mom, too.
If your grandma can install a web browser, she could install lagrange. If she knows how to navigate to 'facebook.com,' she can navigate to a gemini URL. If gemini got enough user share, lagrange or something similar would come installed on a computer by default. Gemini is not all that hard to use. From the user perspective, it is just like using Chrome, but the websites are more basic.
It's not that they have a plan to abuse people for money. It's just that being opposite of boring gives them profit. Outrage brings engagement. Engagement brings time on the platform and ad clicks.
If that is the extent of this line of reasoning then I guess I am in agreement with dvt. That doesn't seem "bad" to me in the least. I thought cr__ was sharing new information regarding the often criticized study I linked above.
Let's be honest, 99% of what your grandma has to tell you could be put in an email with no intermediating parties. The web is totally unnecessary for that.
> The modern web does not suck. FB is usable by my grandma
I'd say the fact that your grandma has to use FB instead of publish on the Web itself is exactly why it sucks. What most people are using isn't really the Web, but apps that happen to (ab)use the Web. Many of which don't even use the Web for the app, but just redirect you to a mobile app instead.
Now I don't think Gemini is actually fixing any of this, but I do feel that the Web today is fundamentally broken and the lack of self-publishing is disheartening, though understandable given that the Web makes it far from easy.
I think a lot of the hostility people have towards Gemini comes from the frustration they have when they consider the absolutely massive scale of the modern web. Can a community of FOSS activists really fit the square peg of the modern web into the round hole of user-friendliness? Probably not. And I think on some level they understand this. Will governments around the world force user friendliness on major tech companies? Maybe, but probably not to the extent that satisfies us.
Browsers are a natural monopoly. The reason is scale. Modern browsers are just too fucking big. Bob from Seattle isn't taking down Google with his neato "classic browsing experience". Nobody uses Bob's browser.
So how do you rid the web of its surveillance scripts, telemetry-filled browsers, bad UX decisions, unnecessary bloat, manipulative algorithms and lowest-common-denominator pandering? Well, Gemini says you don't. This is where people get red in the face. How could you possibly give up the good fight!? Just visit websites that aren't trash! Well, sure. I do visit blogs that treat me well, but there's a reason most of the web is garbage. Because garbage makes money. Individual consumer decisions make up an aggregate of consumption behavior that shifts the market towards certain development patterns. "Surveillance capitalism" could also be called "you get unlimited entertainment for free, and in exchange I keep the receipts" capitalism. That's a pretty good deal for most people. And that's why the business model keeps getting copied. It's scalable, popular, practical, and yields the highest number of consumers of your product. It would literally be irrational for a business NOT to do this. When you block 30 trackers on someone's recipe page, you're taking a square peg and trying to force it into a round hole. That isn't a recipe page, but a tracking page that happens to have a recipe on it. Making it "work" is contrary to its intended purpose.
So Gemini basically asks the question "what if we made a protocol that is architecturally predisposed to being user-friendly?" Gemini is a thought experiment in making a tool that's for a particular thing, for a particular group of likeminded people. What if instead of accepting dependancy on a browser duopoly, the protocol allowed anyone to hack up a browser in an afternoon? What if the protocol was designed to make surveillance capitalism an impossible business model? What if instead of trying to turn the corporate web into FOSSweb and bringing everyone into web utopia, developers created their own web that caters to user-friendly experiences for the niche minority of people who actually care about all that high-minded moralizing?
So is Gemini useless? Maybe for you. But I found interesting blogs and I use the Lagrange client, which is very pretty. I enjoy reading peoples' thoughts on philosophy, economics, science, environmentalism, all kinds of things. And for the first time in decades, I feel like I'm actually surfing the web again, as I browse feed aggregators and click on random links, digging through post after post on all kinds of cool things. And it feels like a relaxing break from having to worry about what scripts may be on the sites I browse. I just click and go.
Will Gemini show me 4k videos of puppies being carried around in strollers at the park? No. Gemini will never fill that part of my browsing habits. But does it have to? Consumers have become so domesticated in their mindset that one tool has to do it all. Once the tool turns out to work against you, you find yourself choosing between living in modern society or becoming a hermit. There is value in letting a hammer just be a hammer, and not trying to tape a saw on top of it.
The people whose content I most respect are writing into a blogger setup that hasn't changed for decades, or into Wordpress's editor that feels like it actively fights you when you want to write. They like making it look like they want it to, and they like having analytics or getting paid for their content with ads or subscriptions. And on the other hand, barring one notable exception I haven't found anybody in the gemini space whose content I found interesting in itself. (Of course, that's just personal opinion - you might find many more people in the gemini space interesting.) And the moment I realized that a significant portion of the content on the gemini space was about implementing a gemini reader, it hit me that it was pure solutionism. Gemini is a protocol for techies who want to bikeshed about tech!
(I could also go on about how the idea that stripping content of styles or """fluff""" is the only way for readers to get it on their own terms. I use stylus to override the font on every website I visit - it didn't require reinventing the wheel for that!)