v1.4, 2016-02-24
    * Drop support for Django before 1.8.
    * Add a web interface for importing Akoma Ntoso files. #435
    * Show speaker's speeches most recent first. #482
    * Hebrew and Slovak translations.
    * Improved testing using tox.
    * Use regular arrows in back/forward buttons. #509
    * Akoma Ntoso import
        * Add improved skip/replace/merge import options. #456 #434
        * On directory import, import files in sorted order.
        * Raise exception on empty 'showAs' attributes. #442
        * Warn if docDate missing attribute or doesn't parse. #492
        * Print out statistics in command line import. #460
        * Print traceback for import exceptions. #475
        * Make SSL certificate verification optional. #476
    * Bugfixes
        * Add translation of metadata. #497
        * Fix issue with importing empty docDate/source_url. #484
        * Compare section and speech keys with same type. #472
        * Fix redisplaying of speech date in Django 1.7+.
        * Give sections with no headings a title of 'Untitled'.
        * Fix flake8 part of pre-commit hook, and colour code output.
        * Fix display of speaker name when used in code.

Thanks to James McKinney, Duncan Parkes, Jaroslav Semančík and Matthew
Somerville for their work on this release.

v1.3.2, 2014-12-13
    * Improved initial running of example project, providing login links
      and buttons for adding or importing data. #222 #411 #419

v1.3.1, 2014-12-12
    * Fix error in speaker colour display in Python 3.
    * Akoma Ntoso import
        * Better generic error handling, and of missing speaker TLCPersons

Thanks to Jaroslav Semančík and johnfelipe for reporting these issues.

v1.3, 2014-12-05
    * Delete/ merge speaker front end interface. #407
    * PopIt and Popolo JSON import management command and front end. #231 #310
    * OpenGraph data output. #364 #416
    * Dots in debate subsection timeline, fix section divider display. #372
    * Spanish translation, bundled .mo files. #356
    * Nicer section and speaker forms. #350
    * Update the slug if you edit speaker or section. #433
    * Show more editing buttons on home/ speakers pages. #417
    * Improve speaker sorting, including ability to provide sort name.
    * Use /section instead of /sections for editing URL endpoints.
    * Akoma Ntoso import
        * Options to clobber or skip existing data on import. #387
        * No error if no files found on --dir import. #409
        * Support for source_url. #354
    * Development
        * Quieten unneeded warning. #406
        * flake8 everything, and add flake8 to pre-commit hook. #432
        * Speed up Travis.

Thanks to Duncan Parkes, James McKinney, and Matthew Somerville for their work
on this release.

v1.2.2, 2014-10-01
    * Fix a bug in South migrations introduced in v1.2 (80e3f0b4) where
      migrating could give an error when it tried to rename the content type.
    * Upgrade django-subdomain-instances to get its migrations.
    * Don't show keyboard instructions when no next/previous section.
    * Fix typo in link in README.

Thanks to Andrew Langhorn, Duncan Parkes, Matthew Somerville, and Martin Wright
for their work on this release.

v1.2.1, 2014-09-26
    * Don't error if speaker image caching fails
    * Add new-style migrations for Django 1.7
    * Allow loading of remote Akoma Ntoso files
    * Make it easier to override the ImportCommand behaviour that is
      filesystem-specific
    * Remove requirement to run collectstatic before tests

Thanks to James McKinney and Matthew Somerville for their work on this release.

v1.2, 2014-09-19
    * Cache speaker image from URL, using face recognition for thumbnail
    * Improve display of subsections on section pages
    * Unicode slug support
    * Paginate /speeches
    * Link truncated speech text to speech
    * Tweaks to speech templates, including showing speech titles
    * Add a date picker to speech/ section forms
    * Unhide unimportant form sections which contain errors
    * More translatable strings, including form labels
    * Reduce database queries in get_ancestors and search indexing
    * Model additions:
        * Speech 'type'
        * Section start date and time
        * Section/Speech num, heading (renamed from title) and subheading
    * Akoma Ntoso import
        * Allow anonymous speeches
        * Add support for more preface elements
    * Development
        * Easier local development settings override
        * No requirement to have auth middleware installed

Thanks to James McKinney, Duncan Parkes, Matthew Somerville, Martin Wright, and
Zarino Zappia for their work on this release.

v1.1.1, 2014-07-04
    * Switch to PyPI-hosted fork of django-bleach to remove extra step running
      under Python 3.

v1.1, 2014-07-04
    * Display improvements, including to the default front page, the speech
      detail page, nicer empty content messages and various tweaks
    * Editing improvements, including a much nicer speech editing form, with
      inline creation of new speakers and sections
    * django-popolo is now used to store people data
    * Python 3 compatibility
    * Removal of unneeded dependencies
    * Included git pre-commit hook for CSS compilation
    * Realtime search processer enabled in the example project
    * Akoma Ntoso import
        * Support coverPage and docTitle
        * Log rather than die on error
        * Namespace handling fix

Thanks to Richard Lewis, James McKinney, Duncan Parkes, Matthew Somerville, and
Zarino Zappia for their work on this release.

v1.0, 2014-04-16 -- Initial release.
