v0.5.0 / 2018-01-13 =================== - Fix: recognition option in settings - Fix: no_voice flag in neurotransmitter neuron no longer lost - Fix: retrieve parameters when user order contains non matching words - Fix: Update Voicerss TTS - Fix: Usage of double brackets with json sentence - Fix: Remove acapela TTS - Feature: Kalliope can be started muted - Feature: add geolocation signals - Feature: add Watson TTS - Feature: Hook. WARNING: This is a breaking change. Settings must be updated - Feature: add normal, strict and ordered-strict order v0.4.6 / 2017-10-03 =================== - add core neuron: neurotimer - add core neuron: mqtt - add core signal: mqtt - new feature: community signal now supported - dict can be used in global variables - bug fix: python 3 execution with snowboy lib - new feature: kalliope memory - add espeak tts to core - add stt options. manual or dynamic threshold - Fix: neurotransmitter bracket in answer v0.4.5 / 2017-07-23 =================== - add keyword_entries attribute to CMU Sphinx - add simplerate attribute to Pico2wav TTS - API: convert mp3 file to wav automatically - add sensitivity attribute to snowboy trigger - add grammar attribute to CMU Sphinx - add no_voice flag to api - add possibility to send parameters when using api with run synapse by name v0.4.4 / 2017-05-20 =================== - Fix: Uppercase in order/parameters/global variables are now handled correctly - Fix: usage of integer in neuron parameters - Fix: encoding with special character - Refactor main controller. Use a LIFO to allow full usage of kalliope via API (even with neurotransmitter) - Add a systemd script to start kalliope automatically - docker testing - python 3 support 3.4, 3.5, 3.6 - Increase testing code coverage - Fix: Raspberry performance. CPU usage from 120% to 15% - Input value refactoring. "args" parameter replaced by jinja templating - Review TTS overriding config in neuron declaration - Fix: accapela TTS - LED and mute button support for Raspberry Pi - Player modularity v0.4.3 / 2017-03-11 =================== - Update Documentation - Improve API + manage incoming audio - Fix bug maximum recursion - Disable REST API by default - Add command line run-order to start a synapse - Run default synapse if STT engine do not return an order - Manage global variables - Remove password from log - cli command to delete a neuron - Fix neurotransmitter - Fix Acapela TTS - Remove Voxygen TTS (not working anymore) v0.4.2 / 2017-01-18 =================== - fix community neuron installation v0.4.1 / 2017-01-15 =================== - add CORS support for REST API - fix installation of community STT and TTS - add feature: sound/sentence when Kalliope is ready - fix Raspbian install - add offline STT (CMUSphinx) - orders are not anymore case sensitive - add STT utils class to split the audio catching from the processing - fix CTRL-C to kill Kalliope v0.4.0 / 2016-12-28 =================== - Add resources directory (neuron, stt, tts, (trigger)) - Install community modules from git url - Fix API (strict slashes + included brains) - starter kits (FR - EN) - Add support Ubuntu 14.04 - Fix neurotransmitter (multiple synapses call) - Neurotransmitter improvements (pass arguments to called synapse) - Split Core Neurons and Community Neurons - update some pip dependencies - Add Russian snowboy model for каллиопа v0.3.0 / 2016-12-7 ================= - add unit tests for core & neurons - add CI (Travis) - refactor Event manager - support installation with setup.py - support pip installation - fix ansible_playbook neuron - add rss_reader neuron - review settings and brain file loading - add default neuron used if Kalliope does not match the spelt order v0.2 / 2016-11-21 ================= - add neuron URI to call web services - fix wikipedia neuron. renamed wikipedia_searcher - update neuron neurotransmitter: can now call another synapse directly without asking a question v0.1 / 2016-11-02 ================= - Initial Release