The XML reference

The words the specifications assume you already know — CDATA, QName, PCDATA, canonicalization — each defined in a sentence, demonstrated against a real parser, and linked to where it bites.

For example CDATA section

Root element, namespace URIs, media type, and the constraints that cause real failures — sitemaps, RSS, Atom, SOAP, XSD, WSDL and more.

For example XML sitemap

Per-element rules for sitemaps, feeds, SOAP, and XSD: what may sit inside what, how many are allowed, and executable broken and fixed documents.

For example <lastmod>

Look up a namespace URI, its conventional prefix, the specification behind it, and the formats that declare it.

For example sitemaps.org 0.9

Every entry carries examples that were run on libxml2 and had their output compared, including the namespace trap that makes the obvious expression return nothing.

For example normalize-space()

What validates and what does not, with every literal put through the schema validator — because most schema failures are lexical rather than semantic.

For example xs:dateTime

Real XML you can fetch right now, from standards bodies to transit APIs. Every entry was fetched before it was listed and is re-checked daily, so link rot reports itself instead of accumulating.

For example Earthquakes, past hour

XML or JSON, RSS or Atom, XSD or RELAX NG — the differences that matter, demonstrated by running both sides, and an honest account of where the other one wins.

For example XML vs JSON

Keyed by the message your parser actually printed, and cross-referenced to how libxml2, Xerces, expat and .NET each word the same defect.

For example Premature end of data

The failures that only appear once an XSD is applied — a document can be perfectly well-formed and still fail every one of these.

For example Element is not expected

Validate, explore and convert XML in the browser. They parse in memory, store nothing, and fetch nothing — no sign-in.

For example XML validator