For months, Dave Winer has been talking about a different kind of web. One designed around the needs of people who write and people who read, rather than platforms that want to control everything. On November 6, he showed how far he has come. His new web is beginning to take shape, and some of it is already available to explore.
I have been following Dave Winer’s experiments with what he calls the Web of Markdown for some time. The idea appealed to me immediately. A simpler web, built around ordinary text documents, where reading an article doesn’t require downloading half a software application first. But when I read his November 6 update on Scripting News (http://scripting.com/2026/11/06.html), I realised that he is working on something considerably more ambitious. He isn’t simply building a browser for Markdown documents. He is bringing together several technologies he has developed over the years into what could become an alternative way of publishing, discovering and reading information on the internet.
Winer has several names for his project. Sometimes he calls it the Web of Markdown, sometimes the Writers Web, and more recently BonBonWeb. Apparently, he hasn’t quite decided what to call it yet. But the underlying idea is becoming increasingly clear. He wants to create a web where writers can concentrate on writing, readers can decide how they want to read, and nobody is forced to depend on a particular platform or software company. Most importantly, he wants every component of this new environment to be replaceable. If you don’t like one application, you should be able to use another without losing access to the content or the people you’re connected to.
That sounds rather attractive to me. And the interesting thing is that Winer is no longer just describing his plans. He has a working prototype.
A first look at BonBonWeb
The best way to understand what Winer is developing is to try his BonBonWeb demonstration (https://demo.bonbonweb.org/). It looks like a relatively simple browser, with an address field and a place where documents appear. You can navigate between pages, follow links and read articles. At first glance, it doesn’t look particularly revolutionary. But there is an important difference between this environment and the browser you’re probably using to read this article.
BonBonWeb is designed around Markdown documents rather than conventional HTML webpages. Markdown is a simple text format that allows writers to create headings, paragraphs, lists, links and other basic formatting without using complicated software. The files remain readable in almost any text editor, and they can be displayed by different applications without requiring a complex publishing system. Winer wants to use these ordinary documents as the building blocks of a new web, connecting them through hyperlinks just as the original World Wide Web connected HTML pages.
You can also open Scripting News through the BonBon demonstration (https://demo.bonbonweb.org/?url=http%3A%2F%2Fdave.writersweb.org%2Fscripting) to see what his own writing looks like in this environment. The experience is deliberately uncomplicated. You get the article, its structure and its links, without the elaborate interface that often surrounds content on modern websites. It feels more like reading a document than interacting with a website, although the ability to follow links means you can still move freely between different pieces of information.
There is one slightly confusing aspect to the demonstration. You’re effectively running a browser inside another browser. Winer acknowledges this in his announcement, explaining that the full version of BonBon is intended to be a separate application. It already includes familiar browser features such as bookmarks, browsing history, an address bar and the ability to view a document’s source. As of November 6, the full application wasn’t quite ready for release, but Winer suggests that it won’t be long.
What if readers controlled the appearance of websites?
One of the most interesting aspects of BonBonWeb is the way it separates content from presentation. On today’s web, the publisher generally decides how an article should look. The website determines the fonts, colours, column widths, backgrounds and countless other visual details. Sometimes that results in a beautiful reading experience, but sometimes it produces pages that are unnecessarily difficult to read. And even if you dislike the choices a publisher has made, there is often little you can do about them without installing additional software or using your browser’s reading mode.
Winer wants to reverse that relationship. The writer remains responsible for the words, headings, links, lists and other elements that give the document its meaning. But the reader decides how those elements should appear on the screen. In the BonBon demonstration, you can experiment with this by clicking the A icon next to the address field. It opens a menu with five presentation styles: Basic, Book, Classic, Mono and Newsprint. The content doesn’t change when you select another style. Only the way it is displayed changes.
I particularly like this idea because it gives readers a degree of control that has gradually disappeared from much of the web. Perhaps you prefer a layout resembling a printed book, with comfortable typography and generous margins. Perhaps you like something that looks more like a newspaper, or maybe you simply want a very basic presentation without any decorative elements. In BonBonWeb, those preferences belong to you rather than to the publisher.
Winer makes an interesting distinction here. He doesn’t want writers to lose the formatting possibilities that make longer texts readable and expressive. Quite the opposite. He specifically mentions headings, bold and italic text, hyperlinks and lists, pointing out that many Twitter-like systems have historically restricted these basic features of web writing. He wants writers to have those possibilities back, while leaving decisions about visual presentation to the reader. In other words, writers should control what they say and how their text is structured, but they shouldn’t need to become graphic designers as well.
There is another consequence of this separation. In BonBonWeb, the content isn’t allowed to dictate the browser’s CSS or execute its own JavaScript. The document is there to be read, not to turn itself into an application. That may sound like a technical detail, but it could have considerable consequences for ordinary internet users.
Think about what happens when you visit a commercial news website. Before you can read the article, you might encounter a cookie banner, a newsletter invitation, several advertisements, a video that starts playing automatically and perhaps a pop-up asking you to register. Much of this has become so familiar that we hardly question it anymore. Yet what is the benefit to someone who simply wants to read a few paragraphs?
A browser designed specifically for reading Markdown could eliminate much of that additional machinery. Documents would generally be smaller, pages could load more quickly, and the browser wouldn’t need to execute scripts simply to display an article. There could also be privacy and security advantages because website-supplied JavaScript is a common mechanism for tracking and other unwanted behaviour. Of course, this wouldn’t make browsing automatically anonymous or completely secure, and a simple HTML page can offer many of the same advantages. But BonBonWeb would make a quiet, predictable reading experience the starting point rather than something users have to create for themselves.
Much more than a Markdown browser
What makes Winer’s November 6 announcement particularly interesting is that BonBonWeb isn’t intended to operate in isolation. He explicitly connects it to several other projects he has been developing, including FeedLand, RSS.chat and textcasting. These technologies might not mean much to ordinary internet users, but together they help explain why Winer believes he is building something bigger than another browser.
FeedLand (https://docs.feedland.com/gettingstarted.md) is Winer’s system for working with RSS feeds. RSS allows people to subscribe directly to websites and receive updates whenever something new is published. Instead of relying on a social media platform or search engine to decide which articles you should encounter, you choose the sources you want to follow. FeedLand builds on that principle by providing tools for managing and discovering feeds. It’s part of an approach to the web where people maintain their own collections of interesting sources rather than depending entirely on algorithmic recommendations.
Then there is textcasting (https://textcasting.org/), Winer’s approach to making text publishing work across different applications and services. The idea is that writers should be able to use the tools they prefer and distribute their writing without becoming dependent on one particular platform. If publishing systems agree on open formats and ways of exchanging information, a document doesn’t have to remain confined to the application in which it was created.
Winer also connects BonBonWeb to RSS.chat (https://this.how/rsschat/), a project exploring conversations built around RSS. In his November 6 announcement, he explains that comments are planned for BonBonWeb and points to a demonstration of RSS.chat (https://demo.rss.chat/) to illustrate the approach. The functionality hasn’t yet been connected to BonBonWeb, but the intention is clear: readers should eventually be able to do more than simply consume documents. They should also be able to respond and participate in conversations.
Imagine how these pieces might work together. You write an article using your favourite Markdown editor and publish it through an open publishing system. Other people discover it through RSS, read it using BonBon or another compatible browser, and perhaps respond through a commenting system that isn’t tied to a large social media platform. None of these activities necessarily has to depend on software from the same company. You could replace your writing application without losing your documents, choose another feed reader without abandoning your subscriptions, or switch browsers without changing the content you can access.
That is the larger ambition behind Winer’s project. He doesn’t want to create another closed platform where one company controls publishing, discovery, presentation and interaction. He wants an environment in which these functions are separate and the software performing them can be replaced. In his announcement, he is emphatic about this: every part should be replaceable. There should be no silos.
This is an important difference from the way most social media platforms operate. If you publish something on a particular service, that service usually controls how your content is displayed, how people discover it and how they interact with it. Your followers and conversations are often closely tied to the platform itself. Moving elsewhere can mean leaving much of that behind. Winer’s approach is intended to avoid this dependency by building on open technologies that different applications can use.
Whether the complete system will eventually deliver all of this remains to be seen. The November 6 announcement shows a working browser prototype and describes how Winer wants to connect it to his other projects. It doesn’t mean that every component is already finished or that a fully functioning alternative web has suddenly appeared. But for the first time, the various pieces are beginning to form a recognisable picture.
My own little experiment with a Markdown browser
All this is particularly interesting to me because I recently built a small Markdown browser of my own. Mine is considerably more modest than Winer’s project. It can browse Markdown documents stored on my computer’s SSD or available through my local network, but it doesn’t connect to the public internet. It is essentially a personal tool for navigating and reading my own collection of documents.
Even so, building it has given me a taste of what a simpler browsing environment can feel like. You open a document, follow a link to another document and continue reading. There is no elaborate application interface competing for your attention, and the files don’t depend on the browser continuing to exist. If I stop developing the application tomorrow, the documents will still be there. I can open them in Obsidian, a text editor or another Markdown viewer.
This is something I increasingly value. I use Markdown for much of my writing and note-taking because I want my words to remain accessible independently of the applications I happen to use today. The format is also particularly convenient when working with AI tools, including locally running language models, because Markdown documents are relatively easy to search, process and reorganise. But the most important advantage has nothing to do with AI. It is simply that my writing remains in a format that I can understand and use without depending on a particular software company.
Winer is now taking that same principle and applying it to a much larger environment. Instead of using Markdown simply to store and navigate personal documents, he wants to make it possible for people to publish those documents on the internet, connect them through links and participate in a broader network of independent writing.
Could this really become an alternative web?
There are good reasons to be cautious about how far the project might develop. BonBonWeb is still experimental, the complete browser hasn’t been publicly released, and the commenting system isn’t connected yet. There are also practical questions about how publishers will make their Markdown documents available, how people will discover interesting material and whether enough applications will support the same conventions to create a genuinely open environment.
The question of adoption may be particularly important. A browser becomes useful when there are interesting places to visit, and those places need people who are willing to publish. Winer seems optimistic about the possibilities. In his announcement, he even suggests that existing social networking software might eventually adopt the approach. That would be an interesting development, although it is obviously too early to know whether this will happen.
There is also no obvious reason why BonBonWeb would need to replace the conventional web. I enjoy the enormous variety of personal websites, including those with unusual designs, experimental layouts and creative uses of HTML and CSS. Some websites are works of art in their own right, and I wouldn’t want to lose that diversity. Other websites are interactive applications that need technologies far beyond what Markdown can offer.
Perhaps the most attractive possibility is that the two approaches could coexist. A personal website could continue offering its familiar HTML pages while also making its articles available as Markdown. Visitors who enjoy the original design could use their conventional browser, while those who prefer a quiet, consistent reading environment could open the Markdown version. The content would be the same, but the reader would have more choice.
This might also make independent publishing more accessible. People who simply want to share essays, notes, discoveries or personal reflections wouldn’t necessarily need a complex content management system. They could write in a text editor, publish Markdown documents and let the reader’s browser take care of the presentation. For many personal websites, that might be more than enough.
Perhaps the web needs a little less technology
What I find most appealing about Winer’s project is that it brings together several ideas that have been circulating around the independent web for years. Open publishing formats, RSS, personal websites, portable content, reader-controlled presentation and software that doesn’t lock people into particular services are not new concepts. But combining them into a coherent environment that ordinary people might actually use is a much bigger undertaking.
There is also an interesting detail in Winer’s announcement about how the software is being developed. He mentions using Claude extensively to experiment with BonBon and its visual presentation. AI-assisted programming has made it possible for individual developers to build and test software ideas much more quickly than before. In this case, it is helping an experienced developer explore an alternative to some of the complexity that decades of software development have introduced into the web.
I find that rather amusing. We are using increasingly sophisticated AI technology to help build software that deliberately does less.
And perhaps that is exactly what makes the experiment worth following. Much of the technology industry seems to assume that progress means adding more features, more automation, more personalisation and more complicated interfaces. Winer is exploring a different possibility: what happens when we remove some of those layers and concentrate on the basic relationship between people who write and people who read?
In his November 6 announcement, Winer describes this project as potentially his last major contribution to the kind of world he believes in. Coming from someone who has already played such an important role in blogging, RSS and open publishing technologies, that is quite a statement. It also helps explain why he seems determined to bring his various projects together into something larger.
I don’t know whether BonBonWeb will become a widely used alternative to conventional browsing. It may remain a relatively small experiment, or it may inspire other developers to build compatible applications and services. Perhaps it will eventually become part of a broader movement towards simpler, more independent ways of using the internet.
But I like the direction in which Winer is thinking. A web where writers can concentrate on writing, readers can decide how they want to read, and the software connecting them doesn’t have to belong to one company.
A web where links connect documents rather than trapping people inside platforms. Where the content matters more than the machinery surrounding it. And where changing your browser, publishing tool or feed reader doesn’t mean abandoning your digital life.
For now, I’m going to spend a little more time exploring BonBonWeb (https://demo.bonbonweb.org/). Because this is one of those experiments where the most interesting question isn’t what new technology we can add to the web.
It’s what might happen if we finally started removing some of it.
Explore Dave Winer’s new web
If you’re curious about the project, these are some useful places to start.
- Dave Winer’s November 6 announcement — The original post in which he brings together BonBonWeb, FeedLand, RSS.chat and his broader vision.
- BonBonWeb demonstration — Explore the experimental browser and try the different reading styles.
- Scripting News inside BonBon — See Winer’s own writing displayed through the new environment.
- RSS.chat demonstration — An example of the approach Winer intends to use for comments.
- RSS.chat project — Background on conversations built around RSS.
- FeedLand — An introduction to Winer’s RSS-based system for following and discovering publications.
- Textcasting — The broader idea of publishing text across different applications without becoming dependent on a single platform.