Version Packages

[github-actions]

This PR was opened by the Changesets release GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated.

Releases

@hyperbook/fs@0.29.1

Patch Changes

• 9cee0b6 Thanks @mikebarkmin! - dev: A build failure no longer takes the dev server down with it.

  hyperbook dev used to exit when the initial build threw, so the author fixed
  the file against a dead port and started over. The server now starts anyway,
  prints the failure, and serves it: routes the broken build never wrote answer
  with an error page, pages still standing from an earlier build get the overlay
  as soon as they connect, and the save that fixes the problem rebuilds and
  reloads everything. A first build that failed also no longer pins the session
  to full rebuilds — incremental mode takes over once a build succeeds.

  frontmatter: A page whose frontmatter is not valid YAML now names the file
  and the line. The parser's own complaint ("incomplete explicit mapping pair")
  arrived with no indication of which page it came from; it now comes with
  path:line:column, the offending line with a caret under it, and — for the
  mistake that causes most of these — the quoted form that fixes it:

  error book/tokenisierung.md:2:22
  incomplete explicit mapping pair; a key node is missed; or followed by a non-tabulated empty line
  2 | title: Text als Daten: Tokenisierung
                           ^
  
  A value containing ": " has to be quoted: title: "Text als Daten: Tokenisierung"
  

hyperbook@0.107.8

Patch Changes

• 9cee0b6 Thanks @mikebarkmin! - dev: A build failure no longer takes the dev server down with it.

  hyperbook dev used to exit when the initial build threw, so the author fixed
  the file against a dead port and started over. The server now starts anyway,
  prints the failure, and serves it: routes the broken build never wrote answer
  with an error page, pages still standing from an earlier build get the overlay
  as soon as they connect, and the save that fixes the problem rebuilds and
  reloads everything. A first build that failed also no longer pins the session
  to full rebuilds — incremental mode takes over once a build succeeds.

  frontmatter: A page whose frontmatter is not valid YAML now names the file
  and the line. The parser's own complaint ("incomplete explicit mapping pair")
  arrived with no indication of which page it came from; it now comes with
  path:line:column, the offending line with a caret under it, and — for the
  mistake that causes most of these — the quoted form that fixes it:

  error book/tokenisierung.md:2:22
  incomplete explicit mapping pair; a key node is missed; or followed by a non-tabulated empty line
  2 | title: Text als Daten: Tokenisierung
                           ^
  
  A value containing ": " has to be quoted: title: "Text als Daten: Tokenisierung"
  

• 706325b Thanks @mikebarkmin! - mermaid / pyide: Diagram labels and test code with non-ASCII characters are
  no longer mangled.

  A mermaid node labelled Überwachtes Lernen rendered as Ãberwachtes Lernen,
  and the same happened to every umlaut, accent, ß and emoji in a diagram. The
  text was stored correctly — the directives base64-encode it with Buffer.from(),
  which writes UTF-8 bytes — but the clients decoded it with a bare atob(), which
  hands those bytes back one character at a time. Ü (C3 9C) arrived as Ã
  followed by an invisible control character, and mermaid drew exactly that. Both
  clients now decode the bytes as UTF-8 before using them.

  pyide carried the identical bug in its data-tests attribute, where it hit
  the assertion messages a learner reads when a test fails: Die Größe muss größer als 0 sein arrived as Die GröÃe muss gröÃer als 0 sein.

  Authors who worked around the mermaid case by writing HTML entities
  (volljährig) can keep them — mermaid still resolves those — or switch
  back to plain characters.

  mermaid: A page with several diagrams no longer stops rendering at the first
  one that was already processed. The loop in loadMermaid used return where it
  meant continue, so one processed diagram skipped every diagram after it.

@hyperbook/markdown@0.79.5

Patch Changes

• 706325b Thanks @mikebarkmin! - mermaid / pyide: Diagram labels and test code with non-ASCII characters are
  no longer mangled.

  A mermaid node labelled Überwachtes Lernen rendered as Ãberwachtes Lernen,
  and the same happened to every umlaut, accent, ß and emoji in a diagram. The
  text was stored correctly — the directives base64-encode it with Buffer.from(),
  which writes UTF-8 bytes — but the clients decoded it with a bare atob(), which
  hands those bytes back one character at a time. Ü (C3 9C) arrived as Ã
  followed by an invisible control character, and mermaid drew exactly that. Both
  clients now decode the bytes as UTF-8 before using them.

  pyide carried the identical bug in its data-tests attribute, where it hit
  the assertion messages a learner reads when a test fails: Die Größe muss größer als 0 sein arrived as Die GröÃe muss gröÃer als 0 sein.

  Authors who worked around the mermaid case by writing HTML entities
  (volljährig) can keep them — mermaid still resolves those — or switch
  back to plain characters.

  mermaid: A page with several diagrams no longer stops rendering at the first
  one that was already processed. The loop in loadMermaid used return where it
  meant continue, so one processed diagram skipped every diagram after it.

hyperbook-studio@0.55.5

Patch Changes

• Updated dependencies [9cee0b6, 706325b]:
  • @hyperbook/fs@0.29.1
  • @hyperbook/markdown@0.79.5

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
hyperbook-4fwa	Ready	Preview	Sep 17, 2026 7:56pm UTC