XML FolioXML Folio / Notebook
A feed for machines. A page for readers.
Create a human-facing reading copy without changing the feed’s address or pretending to reproduce its stylesheet.
Start with a small, representative export
Open the workbench and choose “Atom · relative links.” The example contains two entries, a feed-level base URL, an entry-level relative base, Unicode text and a deliberately rejected link. These details make it more useful than a large happy-path file. Download the authored Atom fixture if you want to compare the source directly.
The first entry’s identifier is urn:example:entry:17. Its title remains “Étude 東京 0017.” Its alternate link is resolved against the inherited base and then the entry’s 2026/ base. The result is https://example.test/releases/2026/entry-17.html?mode=read&lang=en. The entry’s self link is not used as the human article link.
Keep text as text
The first summary contains the literal text <b>Two fixes</b> & one note. The reading copy displays those characters; it does not insert a bold element. RSS descriptions can also contain encoded markup or CDATA. This profile deliberately gives you a text reading copy, with no imported images, event handlers, scripts or styles.
If rich media is essential, this output is not an equivalent replacement. Review a renderer that understands the required feed extensions and content types. A safe, readable snapshot is useful only when its selected fields serve the actual reader.
Review the link that did not become a link
The second example uses a javascript: alternate URL. It remains visible as text and increases the rejected-link count. XML Folio accepts only HTTP(S) destinations without embedded credentials. A missing link is counted separately from a rejected one. Neither causes an entry to be silently removed.
RSS uses the item’s link, not its GUID as an automatic navigation fallback. Atom uses an alternate link, preferring an HTML-compatible candidate when more than one is present. Other links and unused fields appear in the field audit. Relative feed links need a source URL or a usable xml:base; the workbench does not fetch that URL.
Download a reviewable publication
Build the reading copy, then inspect the Field audit tab. Download the HTML, field manifest and publication notes. The HTML is self-contained: its content and styling do not require XML Folio or a hosted stylesheet. The manifest includes record identifiers, titles, raw and resolved links, hashes, omissions and mapping warnings. Treat it as content-bearing material, not anonymous telemetry.
Feed exports with a source URL include an alternate-feed discovery link in the HTML head. Keep a visible original XML link as well. The snapshot’s noindex default avoids automatically publishing a duplicate, indexable version of every feed item; decide its indexing role deliberately.
A snapshot needs a freshness decision
For a release archive, a dated static copy may be enough. For an active feed, regenerate after updates or enable the optional manual refresh button. Refresh only works on the source’s own origin, has explicit request limits and does not save new content back into the downloaded file. The publication guide explains that distinction.
Format references: RSS 2.0 specification and Atom specification, RFC 4287. XML Folio implements the documented reading profile, not a complete feed validator.