Device location settings, bug fixes

This release enables our new and enhanced location configuration services at https://home.mycroft.ai. Users can set default locations for their account and for individual devices. This provides rich location information for Skills running on the Mycroft device, including governmental entity names (city, state, providence, country), GPS coordinates and timezone information.

General enhancements

• Location integration (#487)
• BUGFIX: Authorization token was not updated after expiration (#452)
• BUGFIX: Speech To Text config initialization (#484)
• BUGFIX: Shutdown all skills on process termination (#479)

Developer tools

• BUGFIX: dev_setup.sh ignored WORKON_HOME when VIRTUALENV_ROOT was unset (#372)

Fixes for Picroft and general bug fixes

General enhancements

• Changed network timeout to improve disconnected behavior (#461). This is most important when first bringing up a unit which has no network connection configured yet.
• Updating 'requests' package version in requirements.txt (#464). This fixes sporadic SSL3 errors that were appearing in the logs with long STT requests on some platforms.
• Pairing now says "zero" instead of "oh" (#401)
• NEW: Skills will load from one-level-deep subfolder. This makes it simpler to use git to install a group of skills if the author keeps several in one repo. For example, it will now find "/opt/mycroft/skills/bobs_skills/calendar/init.py" as well as "/opt/mycroft/skills/bobs_calendar/init.py"

Picroft support

• Stop splitting speech at period when on Picroft (#466). The lockup that can happen with analog audio output (an issue with Raspbian/Pi hardware) happen much less frequently.

Developer tools

• Fixed unit tests, added "cleanup" to prevent blocking (#453)
• Added skill_container to packaged distribution (#450)
• Bringing Google TTS support up to speed (#454)
• Update build_host_setup_arch.sh for Arch Linux (#459)

Connection to new back end, Mark 1 unit specific enhancements

v0.8.0

New API and web service integration

• Connenction to the new account management back-end. We have moved to home.mycroft.ai from cerberus.mycroft.ai. Older code is incompatible with the new API so we will continue to maintain Cerberus for some time. Not sure where to pair? Let Mycroft tell you the correct website.
• Many more configuration options, including support for multiple social media login types as well as stand-alone accounts.
• Change TTS engine, wake-word, units, time and date format, and more settings from the web interface.
• We lost support for loading location and timezone configuration from the web but this is coming again soon. These settings can still be found in the mycroft.conf file; Take a look here.

Core enhancements

• Major configuration format refactor. The flat mycroft.ini file has now been replaced by a json implementation in mycroft.conf. This should allow for simpler usage and will be much more powerful.
• The identity.json has been renamed to identity2.json, to avoid conflicts with the Cerberus account management system.
• Configuration loading changes: home.mycroft.ai will periodically be polled poll for changes and no longer requires a complete restart for settings changes to take effect.
• Wolfram Alpha and Open Weather Map now have a isolated API layer, we need not package those libs.
• The listening process has been refined, removing unnecessary threads for better performance.
• The message bus websocket port and path have changed to avoid conflicts. Mycroft now listens on port 8181 with the route '/core'

Mycroft Mark 1 unit enhancements

• Our unit specific enclosure client (for updating Arduino sketches) and packaging method have been modified for platform detection either by configuration file or from serial port communication. This will allow for better compatibility for other systems, such as standalone Raspberry Pi's.

Enhanced wifi-setup for Mycroft devices

This release is focused on the process of getting a new Mycroft device connected to the internet easily. It now walks you through the setup process "out of the box" and will also help you setup up if the connection is lost at a later time -- i.e. if the unit gets moved to another location.

NOTE: These changes apply to the ARM build, so they will also impact a generic Raspberry Pi with mycroft-core installed on it. The walkthru makes reference to the button on the top for now, but the triggering interface and terminology will move into the Enclosure in the future so custom hardware can customize it as appropriate.

WiFi setup client disabled

We discovered an issue using the new WiFi setup client from release 0.7.17 with certain Android phones, so we are holding off on releasing that piece. This release disables that functionality until we get this resolved.

WiFi Setup Client, Push to talk, Voice Speed, Tweaks and Bugfixes

Wireless setup client

@aatchison and @jdorleans added a new client to set up wireless networks on the Holmes unit.
PR: #362

Speech Speed Setting

@forslund added configuration for the Mimic TTS engine to change speech speed.
#334

@ParkerMc made some changes to the ArchLinux build script.
PR: #359

Push to talk and bug fixes

@penrods added push to talk as well as some usability tweaks.
PR: #365

• Pressing the button when it isn't listening starts it listening
• Pressing the button when listening will stop the listen
• Added a mycroft.util.signal() mechanism for out-of-thread communication
• Pressing the button now creates an "buttonPress" signal from the Enclosure
• The viseme playback and aplay check for the 'buttonPress' signal to abort
• Removed "Sorry I didn't catch that", irritating during false activations

Enclosure API for the visualization of Text to Speech

Enclosure API for the visualization of Text to Speech

• Added support to the Enclosure API for the visualization of Text to Speech (using "visemes")
• Mycroft's default Mimic text to speech support now sends visemes to the Enclosure. On a Mycroft unit, this produces lip syncing on screen when it speaks

This integrates changes from the last enclosure code release:

• Illumination changes also impact the matrix dispaly
• Added visemes, which will be visible once changes are made to the TTS client
• Removed the weather icons, added a generic mechanism for sending encoded icons across the serial line

It also breaks display of text and sprites, see #358
This will be addressed in an upcoming release.

Enclosure menu updates

Refined the enclosure menu behavior:

• Menu now comes up after pressing the button for 3 seconds
• Reordered the menu
• Added Illum menu item, which controls the brightness of the eyes
• Several minor visual refinements to the menu

0.7.14: improvements to hardware and TTS

This release adds several different features to improve user experience:

--Hardware units now show a spinning eye animation when shut down, rebooted, or turned on, and stop displaying it when power is fully on or off
--When volume is changed using the unit encoder dial, the eyes now display the current volume (0-11) for a few seconds, then reset to normal
--By holding down the encoder button on a unit for five seconds, users can access menu mode, which allows power control (reboot and shutdown), factory reset, and the hardware test (replacing the automatic hardware test trigger), and will soon include wi-fi setup
--Volume levels can now be configured by the user to determine what each level (0-11) corresponds to
--Mycroft's speech is now broken up into smaller individual chunks, making it easier to understand

Wolfram improvements and hardware test

With this new release, the Wolfram Alpha Skill now gives much more concise responses. Rather than reading all 10 definitions, Mycroft will only read the first. Along with that, when Wolfram cannot find anything for a particular query, rather than speaking the result from the Did you mean? query, he simply mentions he could not find any results but did find something for an alternative query.

This release also brings in a basic hardware test. When activated by holding the button for 5 seconds, Mycroft will now run through tests to verify that the display, knob, microphone, and speaker are all working properly.