The Internet is the most powerful communication tool ever created, yet the infrastructure underpinning it has been leveraged by the tech industry to extract our data and exploit our attention. The Human-Centered Computing Foundation seeks to change this dynamic by creating an alternative architecture for the web. As an approved participant in ICANN’s Applicant Support Program (ASP), we are officially launching our campaign to secure a new Top-Level Domain (TLD) dedicated entirely to ethical, human-centered technology.
To explore our vision for how a human-centered TLD would function, read, download, and share our full initiative overview below:
It is somewhat ironic that the Human-Centered Computing Foundation decided to publish a treatise in PDF format rather than just HTML.
that’s your contribution to the discussion?
PDF’s spec has been free of charge since 1993, and it’s been a public standard since 2008. What exactly is your concern?
We want it to be as easily readable by the most amount of people possible. We are, after all, a human-centered organization.
The principles behind this effort are admirable, but I’m concerned about the practical implementation. Who is funding this? How will one person per domain be verified? Is self-hosting actually required, or can one use an external host; in the case of the former, how would it be verified, and in the case of the latter, what exactly does this tld accomplish that isn’t accomplished by free tlds, tlds like .me, or services like duckdns? Open source software clients as described are a major undertaking; why keep them under the umbrella of this tld?
I respect the goals here, and I want this to succeed, but this is a massive project, and without a clear plan I’m afraid this will peter out even if the tld is granted.
PDF is commonly recognized as the preferred format for documents that you want to maintain the same look and feel for everyone on all devices.
How to get domain?
You have a bad usability when you publish in PDF format – https://www.nngroup.com/articles/pdf-unfit-for-human-consumption/ – There are exceptions, like stuff which needs to be printed out, or saved on the local computer, but otherwise HTML or even raw text – see the RFC’s – is preferable over PDF.
Right download the PDF. Just cleaned up a client that followed those three words. PDF is not more accessible. HTML is.
By relying on the PDF format, the Foundation is not just failing to practice what it preaches, it is actively perpetuating the exact toxic, resource draining, and anti-human dynamics it claims to fight. If the HCCF wishes to be the vanguard of an ethical web, it must recognize that publishing large PDFs is an exercise in digital hypocrisy.
I honestly do not get all the fuss about pdf vs html.
Let’s accept this critic and move on, no point to continue underlining it…
I don’t get it – it’s a domain, therefore only a question of ns lookup. the “pamphlet” (is it really, if it’s only one page?) mentions the namespace but then throws a bunch of other stuff out (VPNs, email, etc). who implements those? how is markhahn.self any different from markhahn.ca? and what does “One Person, One Domain” even mean, given that there are multiple people who have practically any given name…
Good idea, but the PDF is one page? Not much of a “pamphlet”. No planned price, no information about where to get one.
Also, agreeing with everyone else, nobody wants to download a PDF. Say it right on the page. Even more silly given it’s one page with no actionable information.
Pdf is more human. Html is readable by bots.
Your claim is commonly recognized amongst web developers as the preferred excuse used by inexperienced or boss-induced designers, who are willing to sacrifice the adaptivity of HTML for a pixel-exact design placed inside an inflexible presentation rectangle that will adapt to no view rectangle it ever meets.
Also easier to bundle some malwares into it 😉
Alt-F4’d because I don’t want to leave my browser or view a non web native document in a pdf.js shim. I’m posting because somehow the nerds forgot about the end users.
Which pdf based web do you prefer?
“Follow us on…” and a list of walled gardens where you have to submit to tracking by big-corp. Shameful… Where are your accounts on decentralized media??
you sound like a bot
I would switch my site to this. It was breckyunits.com but I failed to renew it last year after like 17 years and now some entity has covered it in casino ads.
You can read it on github still. Gets decent traffic. Ad free and no analytics and works fully offline. Designed the language from the ground up to be simple and human.
Actually — you do.
is that a goddamn em-dash??
We are in the very early stages of this, the ICANN evaluation process (which will take years) has not even begun yet. The goal of publicizing this… brochure if you will is to raise awareness and garner support which we can show to ICANN during our evaluation. The brochure has links to donate, follow us on socials, or sign up for the newsletter where we will be sharing more details as we go along. Any of those things will help us out!
PDF documentation has several major drawbacks compared to HTML documentation: pdfs are significantly larger than equivalent HTML pages, increasing bandwidth usage, storage requirements, and download times. Rendering pdfs consumes considerably more CPU and memory than HTML, especially for large documents or on mobile and low-powered devices.
PDFs are a common attack vector for malware and phishing because they can contain active content such as JavaScript, embedded files, and interactive forms, and vulnerabilities in PDF readers have historically been exploited by attackers.
PDFs are poorly suited to the web, making deep linking, navigation, indexing, and sharing specific sections much less effective than with HTML. Their fixed page layout provides a poor reading experience on smaller screens, often requiring constant zooming and scrolling instead of adapting to different display sizes.
Updating PDF documentation is cumbersome, typically requiring the entire document to be regenerated and redistributed rather than updating individual pages.
PDFs are generally less accessible than well-structured HTML because accessibility depends on correct tagging, which is frequently missing or incomplete.
PDFs are difficult to version and review efficiently since they are primarily binary files, making change tracking and collaborative workflows less effective than with HTML-based documentation.
Overall, PDF documentation prioritizes a print-oriented document model over the interconnected, searchable, lightweight, and responsive nature of the web.
Nice initiative! 🙌
No word on how or when to reserve a subdomain?
I personally think both PDF and HTML documents are much too fancy; you should provide simple text documents. There’s also markdown and TeX, Word,…
I take it back. PDF is perfectly cromulent for now. It’s early.
I wonder how many languages need to be supported 🙂
Loved the entire idea, just found it funny that it’s a 1-page PDF instead of having this in the website…
“dedicated entirely to ethical, human-centered technology.”
How are you going to enforce this? Why can’t my AI bot that I host locally use the .self TLD?
Plans like these will fail for simple reasons
Who gives free .me domain? Because if you are speaking of promotional offer from namecheap for students, then your vision excludes the vast majority of the population.
can’t wait for the .self TLD launch party! I assume the invitations will be faxed to us as a 1-page TIFF file to ensure ‘maximum compatibility?
lol, lmao even
Use of PDF and Follow us with corporate social networks… right, but what about selfhosted services?
This would make it very easy to scan for vulnerable homelab setups…
Seriously, is this some kind of joke? We’re thirty years into the web as the world’s largest platform for reading text online, and sharing information on a file designed for printers? I don’t even have a printer.
john.smith.self?
How is that ever going to work?
go…..your.self
This pdf is not validating with iso standards.
ok…
Would *.your.self count?
.me domains are not free, of course, and I wasn’t saying otherwise. What I meant was that there are both 1. free domains, and 2. domains expressly for personal usage (i.e. .me). Having both combined in the .self domain would be nice, but it lessens the value proposition when both free and personal domains already exist.
People are so distracted talking about the choice of PDF. It’s fine. The problem is that the document is not interesting, the HCCF website is not a good demonstration that the organization understands what people-centric means as you have to struggle to even find any real person named on their social media or website. It seems to use soulless stock photos and lacks self-expression, which does not walk what it talks.
That is all in addition to the general lack of detail about how the finances are supposed to work. A lot of companies have excess capital to offer public goods and many times those offerings come with many challenges that you need to have the resources to face, which many companies do.
How are you going to moderate the TLD to make sure that every subdomain adheres to the right principles and spirit of the free service? How are you going to solve the trust issues of people needing to provide both their full name, address and their always-current IP address consistently to a non-profit with limited resources and questionable security practices compared to a company like Google or Cloudflare who have the funds to hire outstanding security researchers with active persistent investment to maintain that posture?
This is a problem I’ve spent some time thinking about and do have my own thoughts on solutions, but in some ways the feeling we are trying to solve comes from a filtering issue. One problem is that, there are actually more personal websites than ever, but over the decades search engines have deprioritized them in search results not exclusively as a result of companies vying for spots, but because of stalking and other forms of abuse. The reduction in the personal feeling of the web is coming from multiple directions and some of the logic behind that change is reasonable.
You can make a new Angelfire or Geocities, but Github and Cloudflare are already similar aside from the missing directory component. Web directories used to be everywhere and you could volunteer into a directory. Online phone books also used to be a charming part of the web. Now phone books and web directories are mostly not gracing people’s screens anymore.
Even with a directory and solving the hosting, you still face the e-mail hosting problem, where major e-mail providers like gmail already solve so many critical issues around e-mail. Then, if a user is hosting personal information on their site and has their full name in their e-mail address, for many people it will become an e-mail address they don’t want to use on every website as it will simply make it easier for companies they don’t particularly trust to profile them.
I say this because, .self might have a spirit of personal technological sovereignty, but handling the VPN and e-mail already makes that largely performative while posing as many risks as they solve. So the larger perceived mission is to make the web feel more personal, which requires actually creating connections since there is no lack of personal websites in practice, only connections.
There are TLDs of all kinds and I have my doubts that simply a well intended philosophy will be enough without addressing the angles the problem stems from to start with. It’s more of a verification, technology, security and abuse problem than a lack of will for people to express digitally in a personal way.
We need to rethink home internet routers, processor security, search engines, browsers, smartphones and verification.
Cloudflare solved many issues already from a personal or security perspective, but not in a way that increases the connectivity between people. I’m not opposed to HCCF making an attempt, but even with help from for-profit companies (which could be an inevitable requirement) a transformative impact is doubtful without a holistic approach.
Unless you reframe your goal to simply make it easier for people to self-host their websites and ditch all of the larger vision. That’s been done with some success and if it could inspire more people to learn about hosting or how the internet works, then it might do some good in other ways.
The only thing a top level domain would do is give attackers more confidence that self hosters have a home set up
Self hosting is not something I want people to know I’m doing.
amen
Thanks to everyone who pushed back on the PDF. I am a screen reader user and was not able to read the PDF. Sadly, your decision to use .pdf means that this project is, at this point, Human Centered for only certain definitions of human.