The old internet shows signs of quietly coming back

I feel like this article doesn't address the reason why people moved away from the "old" Internet --- most people only care about their _content_ not the particular technology that backs it.

In the old times, you could only host content if you were really good with the technology. The people who had websites were people who knew how to set up and run a web server, and to compose the HTML content that was being served. Similarly, communities who ran BBS forums, mailing lists, IRC, etc. all needed at least some people who could set up and manage the technology for them.

Centralized services (Web 2.0) changed that. Now all of a sudden you didn't need to run your server --- you could just type some text and click some buttons to put up a blog on a shared service (myspace? blogger? wordpress.com?). You didn't need to run your own forum or mailing list or chat rooms, you could just put your community on an existing service that provided that (Google groups, FB/Whatsapp/<insert your IM> Chat). Need a forum? You have reddit. What is Twitter if not just a collection of mailing lists that you can openly subscribe to but not post on? Centralization allowed the vast majority of people to do what they wanted to do while "outsourcing" the technology part to somebody else, because most people don't care about the technology they care about their content.

The problem with the modern decentralized web (3.0?) is that we haven't focused on solving the real problem --- making it easy to manipulate _content_. If anything, setting up a Web 3.0 presence is _harder_ than it was on a centralized service, and as long as it remains harder, the new "old" Internet will always be a niche.

Its the data centers that are going to be responsible for end of web2, when they figure out they can undercut web2 and offer a continuous federated open data storage system and virtual machine. Advertisers will pay them for access directly. Google, Meta, would just be third party indexing services.

And I think the heart of it is that most people working in this space want that. They want the niche back.

Today I see some people spend more effort tailoring their content to a specific platform than I spent inside FrontPage (yes, I know) and uploading the files to my ISP-hosted website. Which, unlike some of these platforms is still up, despite me not having touched it for two decades and having changed ISPs like a dozen times since. (But maybe I just got lucky with that ISP ?)

For now. My company kubelt.com is fixing this.

Your white paper isn't loading for me. Is there somewhere else I can read it?

At best that could be only marginally profitable. Such a service would have to be a commodity in order to be interoperable. If it could have high margins the index owners such as Meta would keep it in-house. As long as they control the services users want, they control the market.

Your prediction could come to pass only in the case of the open backend being tied to a massively popular service that could compete with those giants

How's this?

http://ipfs.io/ipfs/QmNfXy5uCQjEWabnAsvJJPGkKNc6wSZu9M9w2cNx...

I fully agree. Technical people often get lost in the weeds. I recall someone trying to extoll the virtues of a self-hosted YouTube alternative (the name escapes me). To him it was clearly superior to YouTube. So I installed the solution, and spent an hour trying to set up and subscribe to the required indexing services. I gave up and went back to this person and explained that no one is going to spend even 10 minutes messing around if they just want to see funny cat videos.

I've built my career on trying to explain this human element to technical people, and if anything, my job security only improves year over year. I see this same attitude in the Linux community: a willingness to work around UX issues because of perceived benefits in other areas. This is why Linux remains such a niche OS for consumers. No user should ever, for any reason ever, have to open a CLI to install a program. But try arguing that on any distro forum and expect a world of hurt.

Whilst we can debate whether open source is better or worse code, noone debates whether open source has better or worse ux/ui. Clunky UI is what we expect from open source projects...if web 3.0 could create incentives for this phenomena to change.. It would be a lot more interesting

> In the old times, you could only host content if you were really good with the technology. The people who had websites were people who knew how to set up and run a web server, and to compose the HTML content that was being served.

I don't see what's particularly difficult with hosting a static or mostly-static site, and why you'd have to be really good with it (many early content creators weren't). With Apache shared hosting (where your domain, certificates/acme, and basic htaccess are setup for you), this is a matter of putting a couple of HTML and optional CSS files into your document root via ftp/sftp back in the day, or scp, or WebDAV. And HTML is really not that complex a format (or at least used to be before responsive/mobile web), being based on SGML which was designed as structured text format that can be created with a plain text editor in a way that's being taught in secondary school for decades now, and tips available in magazines etc.

This is not true at all in all areas, especially not for tools aimed at devs. PostgreSQL has way better UX than any commercial database I have used.

Latest is on GH: https://github.com/kubelt/whitepaper/blob/main/kubelt_whitep...

404 Not Found - That repo doesn't exist, or maybe is private

I think spam, random DoS attacks, and hackers killed the old intenet. Not many people want to keep up with moderation and software security updates.

I think you're still missing OP's point. Sure, using ftp/sftp or scp is pretty easy for a software engineer, or technology enthusiast. However, that doesn't mean a content creator is happy to learn it, since as OP said, they care about the content not the actual tech. If they have a choice between using something that requires scp and writing HTML versus a graphic environment that is WYSIWYG, then the choice will very likely be the latter.