Squeeze Alexa Versions Save

• Handle old and disconnected players properly (#28, #93)
• Handle now playing for tracks / streams with no artist (#94)
• Friendly error responses if transport is broken (#95)
• Update requests dependency for security (#96)
• Improvements to setup scripts (#100)
• Big improvements to SSL / howto guides (#104).
• Improvements to translations handling for Windows and MacOS mainly.

• Stop transport better (#52) and unsubscribe in MQTT on disconnect
• Improve long ValueError Alexa reponse when transport goes wrong.
• Now Playing: Support multiple artists if LMS is configured for these (#90)
• Now Playing: Support composer for classical (#13)
• Update MQTT docs e.g. systemd and Raspberry Pi

• Fix / rewrite caching layer. In ad-hoc tests, this averages ~600% faster (#88) This might also fix #52 in practice
• Drop v0 intents support for en_GB, v1 is now added.
• Last bits of Python 2 support code gone
• Use Amazon.GENRE instead of custom ones. Performance is sometimes not as good (for my genres at least), but it's standardised now at least and should work across locales - thanks to @NadirJ (#86)
• Documentation updates

• New deployment script (now entirely Python to avoid too many system dependencies) can create / permission / update lambda automatically (need AWS CLI login) (#82 and others)
• Split others scripts into build and release for clarity (#76)
• SSL: only use TLS v1.1 or v1.2 - deprecate everything else
• Setup / SSL: Add Nginx / Dockerised Nginx tunnel examples
• Local test: don't assume users have favourites (#75)
• MQTT: detect bad cert setups
• mqtt-squeeze: add deploy.py command for creating a .tgz, and update docs for this.
• Lots of documentation improvements including new screenshots thanks to @NadirJ (#78)

• MQTT transport is now supported for networks / firewalls where ports cannot be opened (#50 / #60)
• Settings modularised completely to support this, and other things (e.g. testability)
• More testing coverage (~94% now)
• Version 1 Interaction support, thanks especially to @NadirJ (#61)
• French support, thanks to @chatainsim (#66 / #70)
• FIX: volumes larger than 10 or 100% (#65)
• FIX: confusion between CA and cert path when these are not the same
• Better diagnosis of SSL read timeouts.
• Various documentation improvements.

• Fail fast if no players are found (#56)
• Check that the default player is found before using (#59)
• Test a few German translations to make sure they're always working

• Switch to Python3 (3.6 upwards, but 3.5 might work). Python 2 is no longer supported - but you can still use the older releases (e.g. 1.2). This brings new language features (type hinting especially) but allows to cut out some complexity. Docs are updated too
• Refactor: generalise transport (SSL)
• Build tooling improvements
  • We use Circle CI now, not Travis
  • Use pipenv (losing requirements.txt in favour of Pipfile)
  • ditch Tox
  • Use setup.cfg for config
  • Remove references to Python 2
  • Rewritten scripting for packaging / release process (taken from MQTT branch)
• Close SSL connections better (hopefully - though hard to replicate - see #52).
• Update translation tests

• The skill is now internationalised (#44), with en_GB as default, and German (de_DE) (thanks @DFranzen) (#45). For simplicity, just choose your preferred locale in the settings. en_US is supported by default now too.
• Better timeout handling (#48)
• Errors with the skill are now (generally) reported back by voice. It's as English as possible, and due the earlier diagnostics work usually contains useful hints about what the problem might be e.g. firewall issues / certificate mismatches, etc. This should help when users are getting set up, but you can turn this off in the settings if it's annoying :robot:
• There's now a contributing guide.
• A release script / process to, err, create these releases..
• Minor HOWTO improvements

Various smaller fixes:

• Extensive documentation updates around server setup / networking / AWS
• Better handling of exact genre matches (#2)
• Favourites supported (on server-side only)
• Added scripted manual zip file creation to avoid any 3rd-party tools.
• Support for Python 3 (3.6 in particular)
• Absolute volume setting (#12 / #35) (thanks @DFranzen )

First squeeze-alexa release.

Supports

• Custom or native Amazon audio utterance usage
• Basic querying of current playing track
• Playing by genre
• Playing LMS playlists
• Intelligent fuzzy matching for all of these
• Python 2.7, with PY3 code-compatibility in advance of full support
• Advanced diagnostics for helping setup / connection problems.

Test coverage > 90%