Pelican 4.0.1

• Refactor pelican.server logging
• Fix bug in which all static files were processed as "draft"
• Bug fixes for Invoke/Makefile automation, Importer, and other miscellanea

If upgrading from 3.7.x or earlier, please note that slug-related settings in 4.0+ use {slug} and/or {lang} rather than %s. If %s-style settings are encountered, Pelican will emit a warning and fall back to the default setting. Some user-submitted themes might try to format setting values but fail upon site build with a TypeError. In such cases, the theme needs to be updated. For example, instead of TAG_FEED_ATOM|format(tag.slug), use TAG_FEED_ATOM.format(slug=tag.slug)

Pelican 4.0

• Replace develop_server.sh script with pelican --listen
• Improved copy/link behavior for large static files (e.g., videos)
• New {static} syntax to link to static content; content linked to by
  {static} and {attach} is automatically copied over even if not in
  STATIC_PATHS
• Pages can now have draft status
• Show current settings via new --print-settings flag
• New signals: feed_generated and page_generated_write_page
• Replace Fabric with Invoke and fabfile.py template with tasks.py
• New ARTICLE_TRANSLATION_ID and PAGE_TRANSLATION_ID settings to
  specify metadata attributes used to identify/disable translations
• HTML reader now parses multiple occurrences of metadata tags as a list
• New Blogger XML backup importer
• Wordpress importer now updates file links to point to local copies if the
  files were downloaded with --wp-attach.
• Many bug fixes, tweaks, and other enhancements

Pelican 3.7.1

• Fix locale issues in Quickstart script
• Specify encoding for README and CHANGELOG in setup.py

Pelican 3.7.0

• Atom feeds output <content> in addition to <summary>
• Atom feeds use <published> for the original publication date and
  <updated> for modifications
• Simplify Atom feed ID generation and support URL fragments
• Produce category feeds with category-specific titles
• RSS feeds now default to summary instead of full content —
  set RSS_FEED_SUMMARY_ONLY = False to revert to previous behavior
• Replace MD_EXTENSIONS with MARKDOWN setting
• Replace JINJA_EXTENSIONS with more-robust JINJA_ENVIRONMENT setting
• Improve summary truncation logic to handle special characters and tags that
  span multiple lines, using HTML parser instead of regular expressions
• Include summary when looking for intra-site link substitutions
• Link to authors and index via {author}name and {index} syntax
• Override widget names via LINKS_WIDGET_NAME and SOCIAL_WIDGET_NAME
• Add INDEX_SAVE_AS option to override default index.html value
• Remove PAGES context variable for themes in favor of pages
• SLUG_SUBSTITUTIONS now accepts 3-tuple elements, allowing URL slugs to
  contain non-alphanumeric characters
• Tag and category slugs can be controlled with greater precision using the
  TAG_SUBSTITUTIONS and CATEGORY_SUBSTITUTIONS settings
• Author slugs can be controlled with greater precision using the
  AUTHOR_SUBSTITUTIONS setting
• DEFAULT_DATE can be defined as a string
• Use mtime instead of ctime when DEFAULT_DATE = 'fs'
• Add --fatal=errors|warnings option for use with continuous integration
• When using generator-level caching, ensure previously-cached files are
  processed instead of just new files
• Add Python and Pelican version information to debug output
• Improve compatibility with Python 3.5
• Comply with and enforce PEP8 guidelines
• Replace tables in settings documentation with data:: directives