ACE, Atom Syndication Format, and Atom Activity Stream

[sjehuda]

Mauve, you would love this idea.

Utilizing Atom Syndication Format as a cross-protocol document standard.

It is currently implemeted at https://journal.woodpeckersnest.eu/ where every page is Atom Syndication Format (different shapes and special forms are realized by different XSLT stylesheets.

gemini://woodpeckersnest.space/~schapps/journal/2025-08-10-an-analyses-of-the-benefits-of-ace.gmi

gemini://woodpeckersnest.space/~schapps/journal/2025-07-29-announcing-ace.gmi

It works wonderfully.

We can navigate from Atom documents that are posted over eD2k, to Atom documents that are posted over Gnutella, to Atom documents that are posted over Gemini, to Atom documents that are posted over IPFS.

View source code of https://journal.woodpeckersnest.eu/posts/2025-05-05-the-garden-metaphor-of-worldwide-finance/?xml to see the navigation instructions.

<link href="../2024-11-05-xmpp-as-the-internet" rel="navigation" title="proceed" type="application/atom+xml"/>

We might want to utilize XLL (XLink), but "rel" of atom:link seems to suffice.

[RangerMauve]

Would the idea be to try to detect the presence of atom versions and render them out instead of the HTML version of the page + add the user styles?

This could be handy to add to the renderer extension by detecting the link or the contentType: https://github.com/AgregoreWeb/extension-agregore-renderer/blob/default/background.js#L35

From what I understand the browser engine should already be able to render XSLT, at least until Chrome removes support

[sjehuda]

I beg your pardon. I will explain myself properly.

Please ignore XSLT. XSLT is not to be involved with software that support "Ace". The use of XSLT is a recommendation, in order to make Ace documents usable over those archaic HTML/HTTP internet browsers.

"Ace" is still a work in progress. The "Ace" specification is merely
XML; that is, Atom Syndication Format. DOAP, OPML, Sitemap, and Urlset are optional components.

The whole "Ace" specification can realized merely with the Atom format.

The idea of "Ace" is to utilize Atom Syndication Format as a portable
document format, which can be seamlessly browsed over several different protocols, such as Gemini, Gopher, HTTP, SSH, Gnutella, eD2k, IPFS, etc.

This is an example document.

It might has discrepancies, due to manually editing it.

Notice the eD2k and Magnet links in elements "atom:link" with rel="navigation".

Suppose that this document is hosted over FTP or XMPP, and that the software which parses this document supports every protocol, be it centralized or P2P.

<?xml version='1.0' encoding='utf-8'?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <icon>ed2k://|file|icon.svg|FILE_SIZE|MD4_HASH/</icon>
  <title type="text">Welcome to Rivista Voyager</title>
  <subtitle type="text">Publish news feeds and consume contents via XMPP</subtitle>
  <link href="/feed" rel="alternate" title="Atom Subscription" type="application/atom+xml"/>
  <link href="/opml" rel="alternate" title="OPML Collection" type="application/opml+xml"/>
  <generator uri="https://git.xmpp-it.net/sch/Rivista" version="2025.09.07">Rivista Voyager</generator>
  <updated>2025-09-08T00:47:22.401848+00:00</updated>
  <entry>
    <title>Welcome to Rivista Voyager</title>
    <link href="ed2k://|file|rivista.tar.gz|FILE_SIZE|MD4_HASH/" rel="enclosure" title="Rivista Voyager Source Code" type="application/gzip"/>
    <link href="ed2k://|file|2024-11-05-xmpp-as-the-internet.atom|FILE_SIZE|MD4_HASH/" rel="navigation" title="previous" type="application/atom+xml"/>
    <link href="magnet:?xt=urn:btih:SHA1_HASH/2025-05-05-the-garden-metaphor-of-worldwide-finance" rel="navigation" title="proceed" type="application/atom+xml"/>
    <content type="text/html">&lt;p&gt;Rivista Voyager is a syndicated publishing system for XMPP.&lt;/p&gt;
&lt;p&gt;It is an open-source, minimalist journaling platform which is designed for
writers who want a simple and distraction-free writing experience.&lt;/p&gt;
&lt;p&gt;It is built to support the XMPP protocol, allowing people to publish content
which can be shared and discovered across different platforms, such as Blasta,
Libervia, and Movim, over the decentralized network.&lt;/p&gt;
&lt;p&gt;In addition to being cost-effective and having low maintenance overhead, Rivista
focuses on providing a clean interface that emphasizes writing and reading
without the clutter often associated with more complex content management
systems.&lt;/p&gt;
&lt;p&gt;&lt;a class="reference external" href="https://schapps.woodpeckersnest.eu/rivista/"&gt;https://schapps.woodpeckersnest.eu/rivista/&lt;/a&gt;&lt;/p&gt;
</content>
    <summary type="text/plain">Publish news feeds and consume contents via XMPP</summary>
    <published>2024-09-24 00:00:00 UTC</published>
    <updated>2024-09-24 00:00:00 UTC</updated>
    <category term="rivista"/>
    <id>index</id>
  </entry>
</feed>

Do you now realize the concept that I attempt to convey?

[RangerMauve]

I think I understand, yes. If you'd like to try to add this format I'd be happy to review PRs to the renderer extension. This is how we get gemini and markdown document support for example.

I'm guessing the mime type would be application/xml and the code would need to find the xmlns declaration in the <feed> element.

There's the DOMParser API that could be used to traverse the document as you turn it into HTML.

An example document published to either HTTP or a supported p2p protocol should also be added somewhere on the explore page

[sjehuda]

I think that mime type should be application/atom+xml.

I did write code to parse XML documents which includes Atom, OPML, RDF, RSS, and SMF, using module "DOMParser".

https://greasyfork.org/en/scripts/465932-newspaper-syndication-feed-reader

See function "renderXML". It will be shorter than a hundred lines.

I would appreciate if you would have a brief observation at that code, before I begin to work on it, as you might have relevant insights.

[sjehuda]

I have changed the title, just as a reference for future tests of navigating Atom files from one protocol to another.

[RangerMauve]

Just looked it over, ty for the link.

Your approach sounds good, my only desire would be to reuse the template code for the page structure and to keep the doc as flat as possible so that the user controllable stylesheet can apply for the look and feel. e.g. no css classes unless there's absolutely no way to represent the structure with semantic elements. I'm open to adding some more styles to the stylesheet if there are semantic HTML elements we're missing.

Ideally I'd like the details of specific file formats to wash away and to present the content as uniformly as we can across document types and transfer protocols.

[sjehuda]

Thank you, for your time.

Your approach sounds good, my only desire would be to reuse the template code (template.js).

Fine.

Ideally I'd like the details of specific file formats to wash away and to present the content as uniformly as we can across document types and transfer protocols.

Of course. I agree.

There is also other software-specific code, such as the Table of Contents, which I do not intend to include, unless you would want to include it, either now or in future.

A "Table of Contents", "tags" and other sorts of features are one of the interesting things that can be achieved with Atom Syndication Format.

[sjehuda]

Mauve. How are you?

Extension

I have enhanced my extension.

https://greasyfork.org/en/scripts/465932-newspaper-syndication-feed-reader

Now, it supports Atomsub (Atom Over XMPP), Atom Activity Streams, Sitemap, DOAP, FOAF, and RSS 0.90.

I have also segregated the tasks, of data downloading, data extraction, and data rendering.

So, you might want to explore it.

The actual code is roughly 3000 lines, due to CSS code, added features, and multiple formats (each format is handled by a function).

Ignore the first 5000 lines, because these are only XHTML documents that are attached to this software.

Activity Stream

Apparently, as I have explored the Friendica platform, there is "Atom Activity Stream", which is only an Atom Syndication Format entry, with optional extra elements (i.e. modules of external XML namespaces).

So, I intend to set all the single-entry Atom document pages into Atom Activity Streams, unlike the example which I have posted.

#302 (comment)

Navigation

Interestingly, navigation was already realized by RFC 5005: Feed Paging and Archiving.

I have added support for that also.

Conclusion

So, I ask to support Atom Activity Streams and Atom Syndication Format.