Poor Man's Web
zserge.com
zserge.com
IMO, protocol is largely a distraction from the good stuff.
The question I’ve been wondering lately is, “why don’t most people I meet have personal websites?”
1. For non technical users (and there are lots of them) learning enough HTML and figuring out how to put it somewhere accessible is too high of a bar to overcome (forget about updating it ever). It's quicker and easier to just establish a presence on some social media site.
2. For technical users who are capable of setting up a website, it's easier to just go where people already are (i.e. social media).
3. It's anecdotal, but I rarely see anyone browsing the web these days. Most web usage seems to consist of endlessly scrolling through Facebook looking for something to interact with. A link to somewhere else might get clicked on, but the user always goes back to the newsfeed once they're done looking at the link.
Exactly. People on HN overestimate today's use of the web. Regular people barely know what a website is at this point or have any idea what a browser is or does. They only know it as a place to ask The Google something, but even that has changed with The Google's app/widget and voice activated assistants. Sure those things open up web views, but that view may not even have a URL bar, so as far as the user knows, they're in "the app", which is kind of true I guess. That "Chrome" or "Safari" app is just kind of a funny thing that opens up sometimes. Most users aren't opening up the browser and typing in URLs. A fraction of them are, but that fraction is diminishing, and I would guess that fraction mostly reflects desktop users. All this is to say that, if you want to create a blog today and not spend effort to heavily promote it and use SEO tricks (that will immediately go out of date), you can basically forget it if you want anyone to read it.
Isn't it obvious? It's simply easier to use a social media platform that makes it easy to publish without having to know how to set up a server or mess with frameworks or code, and that makes it more likely that people will read what you have to say. The utility of personal websites has been replaced by services which solve both the problems of ease of publishing and ease of discovery and networking.
Most people don't have personal websites because getting an audience of users to a personal website is like asking a mountain to come to you.
Another proverb is: "Meet people where they are at"
There are exceptions to these proverbs but they require exceptional people or exceptional circumstances. The vast majority of people are not trying to be exceptional with respect to the kinds of things they would put on their personal websites. Very few individuals are willing to put the effort into "moving mountains" to build an audience on their own website.
The answer from my experience is simple: because most people that I meet are not programmers (at least not in a strict sense), and don't know many programmers.
On the other hand: most programmers and many STEM degree holders that I meet do have their own websites.
... because they're all posting on social media platforms big and small, just like everyone else.
I think that valuable follow-up question would be "Should they have one?"
I think there's currently a discoverability problem. I love small websites, blogs, passion projects etc and they give me 1000x more joy than social media posts. But I still don't really know a good way to discover them.
Hacker news is honestly probably the way that I find most small web stuff, I like kagi's "small internet" filter a lot too. But as a whole, the likes of google etc are really geared up towards funneling me into looking at posts in social media sites rather than anything else.
But even with business websites, the two aren't equivalent. If a shop has Facebook as their only web presence, for instance, then they are completely invisible to me.
1) search name
2) install app
3) open app
4) tap through sign-up. Maybe type your name or other familiar details.
5) Nothing to do here because it already imported all your contacts, if relevant.
6) Post/message. You did this all from your phone. It was free. Everyone you care about automatically knows you’re there and how to reach you in the app. Tap camera icon to post pictures or video.
If you have trouble with any of it, ask literally anyone for help.
Personal website:
1) ???????
Because it's still nearly impossible for a non-tech person to create one. Facebook/Tiktok/Instagram offer a lot of what people would want from a personal website for free and give you connections to your friends/family/strangers
The bigger question for me is, if it was dead simple and all they had to do was -say- pay $10/mo, would most people actually choose to have a personal website?
I think the issue comes down to most people not seeing enough value to have a personal site to pay for it. And if they don't want to pay, the free DIY options are either a site like Medium or the technical hurdles they have to jump through to get it up and running manually
I’m sure this is where people chime in with their survivorship bias (“you’d be surprised how many people would read your site”) or wildly different value sets (“even if no one reads it, it can help you grow!”). But some of us just want to work as little as possible, and then go home and play with our dogs, or go for a run, or binge watch Netflix, or read a book, or pop an edible and make some music, or pen overly grumpy comments on HN. A personal website doesn’t just bring me no value, I’d go so far to say it would bring negative value.
Why would you have to do this? You could just update it if/when you feel like updating it.
I've had a personal website for decades, and I've always just updated it when I had something I wanted to share (code, usually). I don't bother with artificially trying to "create content" for it at all, let along with any sort of regular cadence.
My website is collection of reference material and software, not a blog (even though it contains actual blog-like articles and creative writings). Since my goal for it is not to maximize audience, things like update cadence are irrelevant.
But it serves some other value too. I had a subdomain point to my home PC so my friends could connect for gaming. And my hg repo on there. And a DB I sometimes use because I'm lazy and don't want to run one locally. Worth the $5/mo
You're putting the cart before the horse. Having a personal site doesn't have to be any more complicated than minting (ideally permanent) URLs for stuff you're already creating—a namespace that third parties don't control. I'd argue that if you're trying to "create content for it", then you're already off on the wrong foot. This is another reason why I strenuously argue against the Smol Web clan who loudly advocate for personal websites as a sui generis medium[1] for (a specific, technofetishistic type of) personal expression—because people in turn respond to it the way you are here, and then their conception of a personal website is compromised exactly like this: that it's something that they don't actually want.
You probably created some kind of document (or other artifact) for distribution to people within your sphere at some point in the last week or three. Why shouldn't it be citeable by URL? That's what TBL's original vision of the Web was really about—not the false/misplaced nostalgia for Geocities-style gewgaws that insist upon themselves and motivates so much of what comes to the fore and dominates these conversations when the subject of personal websites comes up.
Roughly in order: Blogging (as a concept), assholes, Google, social media.
I think the first real hit to personal websites was just the concept of blogging. Before blogging was a popular concept there was rarely an expectation someone's website would see constant updates. Just because they shared pictures from their vacation to Spain there was no assumption you found someone's travelogue that would see constant updates. You could "finish" a personal website.
There were of course plenty of sites that would see regular updates but they were free to set that expectation themselves. There's a conceit to blogging that it's something a person just keeps doing. A blog can never be considered complete. All of the organization is temporal and most of the time so is the navigation. If we find a blog that hasn't had updates in some time we've been conditioned to assume it's dead.
A follow on effect to blogging as a concept was the change in web page authoring software. It went from being designed to build a finite website to build continuous stream that is a blog. Organization of sites went from trees to lists. Tags and categories in most blogging software are just filters for temporally sorted lists of posts.
Because the conceit of blogging is constant updates, local first software was insufficient. It gave way to the likes of WordPress that could be accessed remotely so posts could come from anywhere. This helped kill the market for website authoring software.
Then of course came assholes. Even when you'd decide to start a blog, the Web 2.0 features that helped connect you to others like comments and *backs got abused by spammers and malware. Having a blog meant becoming an unpaid administrator and moderator or disabling all dynamic feedback mechanisms. In the worst case assholes turned your blog into a security liability.
Google adjusted their algorithms to favor recent content. This helped destroy the old style home pages because they became invisible to almost everyone. Unless a site had some very specific content that happened to tickle Google's algorithms just right its results would be on the 400th page of search results after a thousand blogspam links.
Finally social media just became the replacement for blogs and homepages. As an end user there's very little administration that needs to be done. Posting is quick and easy and doesn't ask for formalisms like document titles (naming this is often actually hard) on ontologies. The server administration is also handled by professionals so shitheads flooding your "site" don't rack up huge hosting bills. Content hosting is also handled with the same ease as other types of posts.
I'm not saying social media in inherently good or blogging is bad. I just think they contributed to killing the idea of having a home page.
I know these were deliberately left out for "simplicity" reasons, but then they have a fairly unsimple mechanism for user login/auth that totally undermines their claim for making clients little more than slightly-augmented-terminals, which for me is a smoking gun for these things being missing purely on an ideological basis only and an attempt at controlling how people use it. That's fine, it's their thing and they can do what they want with it.
There is a growing trend of people simply choosing to use XHTML Basic (1). This is a "stripped down" version of HTML originally intended for early phones, PDAs, set top boxes etc - it has the critical missing features of Gemini, but forgoes some of the more "advanced" features of the modern web people have issues with. I'd highly recommend people simply target that instead.
I personally would like to retry a version of the script tag (possibly more like the object tag) with native support for WASM that would make it easy to run scripts in any language... which is closer to the original vision before Javascript "won," and possible without requiring browsers to support the languages. Also maybe things like an <include> tag for HTML, and other things that fell by the wayside of early HTML. I'd like to see something like HolyC, with native support for text, hypertext, drawing and code.
But 99.9% of people seem to just want a web that's basically whitepapers with links, maybe images.
Form submit etc to send network traffics are all available, JS is only needed when you wanna do something cute like network call (AJAX) or change dom elements without refreshing the page. And come up with kluge such as allowing js to change the route with in the domain.
Footnotes: 1.https://1stwebdesigner.com/pure-css-navigation-menus/
1. very poor choice of text size and/or color
2. imposing stupid width/min-width
3. sticky headers
I don’t really see any way to get around corporations on the internet, unless I’m missing something.
I don't have a static IP, but it only changes on the order of weeks or months. So a dynamic DNS configuration with a 5 minute TTL works fine for me.
I have no idea if web hosting is in my ISP's TOS, and I also don't care.
2) I have a fixed IPv4 address, and a dynamic IPv6. So far the latter also seems to be fairly stable, and I host my webpages on both IPv4 and IPv6 these days.
My server is a ThinkPad X240 with VMware ESXi which in turn host Ubuntu for SSH, Web, etc... And a Pi-hole VM just to block the "wost" of the internet.
For your second point, Dynamic DNS is super useful. I've had a good experience with noip.com/ddns.net, but there are plenty of other options out there.
If you want to run code instead of just keep a static website, then you may need to get a cheap server from a multitude of companies that offer that (Linode, Digital Ocean, AWS etc) but I find that you can almost always get away with static sites which require next to no work and will stay up as long as those companies are still functioning (or don't shut down the free plan).
> gem:my.domain.com/hello
# gemeni:|bike.domain.com|shed
Or even better, everything is little endian to allow autocomplete from previous entries by the user in the shell so that once the user hits the domain keyword the shell history parser assumes the user wants to stay on the site
# gemeni:|shed|bike|domain|com
I’m not sure how feasible this as I’ve never written an internet or shell language
Forward slash (/) in DOS, while dash, tack, or hyphen (-) are used for UNIX for command line flags or parameters. The / in UNIX was for the path. The backslash (\) is for DOS paths. (Caveat: I started with UNIX in the mid-80s.)
[0] https://datatracker.ietf.org/doc/html/rfc3986#section-3.2
Didn’t know about it, here is the video of that talk in YouTube:
https://www.youtube.com/watch?v=8JOD1AQGqEg
There is a poem-style intro titled “Flatscreens / a song about code” until the talk proper starts at 5’11”:
https://www.youtube.com/watch?v=8JOD1AQGqEg?t=311
Edit: I’m watching it and it is done in an entertaining style but is a bit loose on some details I know about, such as implying that the IBM PC development was started in the early 1970’s when it was in fact started in 1980. Still worth watching so far.
Edit 2: finished, it is definitely worth watching. It goes into “The Twilight Zone” of a very interesting alternative history. And does a good job reporting on the Gary Kildall / IBM missed deal for their PC OS.
It's limited to those that are primarily interested in text (and the occasional image). Which probably prevents it from ever going mainstream (and thus an eternal september). What it's generally sacrificing is interactivity.
That seems like it's shooting more for the early internet. Which, yeah, has quite a bit of overlap with people who can code, but not as much as you're implying. And I do remember quite a few echo chambers on the early internet, but I remember a lot less of them without algorithmic feeds. Just by chance, you'd end up running across a bunch of people who disagree with you on a lot of things.
Sure, but being smart enough to implement one usually also means being smart enough to use something of the shelf. But few of the available clients are very accessible to a non-programming audience. It is basically a "natural" filter mechanism.
> What it's generally sacrificing is interactivity. That seems like it's shooting more for the early internet.
There is a very rudimentary form mechanism, resembling a bit HTTP 0.9, where you could only do get requests with query parameters, which might allow to simulate a bit of interactivity, but I with the protocol being set in stone that would forever remain a hacky simulation (if you only have a hammer...)
? There’s a very nice one that runs fine on my iPad, Elaho.
$ echo "feed" | nc typed-hole.org 79
julienxx@nein.club toots:
thread 'main' panicked at src/main.rs:52:39:
called `Result::unwrap()` on an `Err` value: reqwest::Error { kind: Request, url: Url { scheme: "https", cannot_be_a_base: false, username: "", password: None, host: Some(Domain("nein.club")), port: None, path: "/api/v1/accounts/01C2ETC44GDGZ27S9WKZMSHKWV/statuses", query: Some("limit=5&exclude_reblogs=true&exclude_replies=true"), fragment: None }, source: hyper::Error(Connect, Ssl(Error { code: ErrorCode(1), cause: Some(Ssl(ErrorStack([Error { code: 167772294, library: "SSL routines", function: "tls_post_process_server_certificate", reason: "certificate verify failed", file: "../ssl/statem/statem_clnt.c", line: 1889 }]))) }, X509VerifyResult { code: 18, error: "self-signed certificate" })) }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
I'm not sure why it's "poor man's" web, other than being minimalistic (text and images only.) I thought the phrase meant something less desirable to use, less capable. But this could be genuinely pleasant!
I will keep repeating that in its quest for simplicity Gemini threw the baby out of with the water.
Basic styling and inline images are invaluable.
I've taken to writing my markdown and other documetns simiarly. Cory Doctorow does something similar.
https://pluralistic.net/2023/12/20/em-oh-you-ess-ee/#sexytim...
We don't need Gemini. We can get the feeling of the Smol Web by just choosing an effective subset of the features of HTML.
I don't think that's working at all. Their website is so unapproachable bad, that it fails in selling me reasons why I should even care about this or read further. Letting something standing on its own only really works well if you have a small amount to deliver. Any slightly lengthy text will just bury you in a desert of letters.
Is this a bad thing?
Books are a "desert of letters", with little to break the text up outside of chapters, sections, and paragraphs. If you broaden the scope a bit, you can added illustrations and photos. People have been reading books for generations. While many books do break that mould, many books continue to follow that tradition.
You're forgetting pages. The fact pages physically limit the visual bounds of all the letters helps people read them. The words from page 6 aren't going to come into view while you're reading page 4 but scrolled down a little too far.
Depends on the purpose. For a Webpage, which is a collection of short texts, to sell you on something, it is bad.
> Books are a "desert of letters"
Depends on the book. A phone book would be a desert of letters, I don't think many would enjoy reading them. Something like a novel, would be a forest of chapters, full of trees with letters arranged in a meaningful way, leading you on a road toward a goal. But a webpage is not a novel, it has an informative purpose, and is full of little small texts of equal value.
Even the most boring books often have things that Gemini purposefully omits: from styling to inline illustrations to diagrams to insets, asides, footnotes, tables of content, just tables, typographic marks etc. etc. etc.
So what do you make of books, then?
I don't think people "keep trying to make Gemini happen" in the sense that you mean. They're not aiming to replace the web. They've got a cozy little community that likes the 'smol', text-based web. And while 90% of people might think they're crazy, there are others out there who would like it too if only they knew it existed. Posts like this make the community a few individuals larger. I think that's the goal.
This webpage is not a book. It has a different purpose. With a book, I know what it contains, where it's leading to, usually they have an abstract for this.
I have nothing against Gemini and the people in general, I'm just saying the limitation is not working well for every type of text. Pictures and a bit more structure would be useful for the boring informative texts.
Depends on the definition of Webpage and book.
> There are webpages that have interactive 3D models (which physical books can't).
Pop-up-Books?
Have you ... seen books? You can't compare a medium as varied as books to a purposefully limited and primitive medium of Gemini.
I feel that if there were some way of promoting a Javascript-free web, everything the world needs is already there in all modern browsers.
I suspect that between HTML5, and indeed XHTML, and CSS and all the many modern image formats and so on, everything important that almost any site needs could be done using these tools and no JS at all.
And the result could also be interpreted and rendered successfully by much smaller simpler browsers, along the lines of Netsurf and Dillo, which are 10% of the size of a full dynamic-content browser or less.
The question is how.
A contest? Make the richest website you can that uses no Javascript, Typescript or anything else, and win a prize as well as promotion?
What I want is less use of Javascript and related scripting languages across the web. I'd like more websites that don't need it.
I am merely proposing a mechanism for getting people to explore what can be done without it, and how it might in fact be easier, more fun, and more maintainable.
I am not trying to create a reason to make script-free websites. I think they are their own justification.
I am trying to propose a method for making them visible and showing webdevs what can be done and why it's worth trying.
Gopher never really went away. A few enthusiasts were keeping it alive. Those enthusiasts realized that Gopher had a number of shortcomings, so Gemini was created to address those shortcomings. It was not created to address the shortcomings of the web. (At least not directly. Indirectly one could argue those enthusiasts kept Gopher alive due to the shortcomings of the web.)
As for styling and inline images: in a sense, Gemini offers styles to a limited degree. Those styles are tied to the structure of the document, while the appearance is left to the software rendering the document. Even though inline images are considered a faux pas, I seem to recall Lagrange offering that feature. Again, we are dealing with the rendering software making the decision rather than the author. Since the end user chooses and configures the rendering software, it is the end user who has control (rather than the author).
From my view in Gopherspace, Gemini is a better fit for the Gopher+TLS thing people keep trying to do which is both incompatible and doesn't square with Gopher being an ultra-light protocol, and thus most appealing to those people who thought Gopher would be the "new smol web" but found it's more its own thing. In particular, Gopher's signature strong menu-document hierarchy that a lot of us Gopher nerds like doesn't have any true parallel in Gemini.
I started gopher.floodgap.com back when it was gopher.ptloma.edu in the late 1990s largely as a historical preservation because I remembered all the cool stuff you could get there. Back then the Web hadn't metastasized to the Tetsuo blob it is today, so that obviously wouldn't have been the reason. I can't speak for the later adopters, but Gemini doesn't scratch my Gopher itches fully (see also https://oldvcr.blogspot.com/2020/11/a-gopher-view-of-gemini.... , my notes on this from 2020).
And thank you for gopher.floodgap.com. While I had some exposure to Gopher in the mid-1990's, most of my exposure was through Floodgap (and SDF) in the early 2000's.
Everything Gemini does, you can do it in HTTP. In fact, a subset of HTTP+HTML would be even simpler than Gemini (mostly because of a lack of TLS) and still compatible with all the modern web stack. Simplicity isn't the main goal of Gemini. Exclusiveness is.
The small web is not for poor men and women. It's for the rich. Those who are rich in culture, intelligence, curiosity, engagement...
The existing "web" is for the technologically poor. It's a chavvy, shallow plastic shit-filled trench, a ghetto where thieves and pimps run free, and good ideas die like dogs - to somewhat paraphrase Hunter S. Thompson. I can't see anything of value left there, and cannot imagine how anyone would equate leaving it behind with being poorer.
Nah, that probably will never work.