• Add module system using sraix and services

• Async (stops reading when after new response)
• Mute function
• Do not disturb in selected time settings
• Language depends on user profile (primary), OS settings (secondary) or geolocation settings
• Voice depends on Language (primary), Robie gender (secondary) or user selection.
• Change Robbie gender when no gender voices available.
• Change Robbie voice language when no voices for actual language are installed.

Default OS settings

• Linux
• MacOS
• Windows

Inbuild solution as second option

• Linux
• MacOS
• Windows

• Module classes to robbie/src/programy/utils/modules
• Helper classes (shared methods for multiple modules) to robbie/src/programy/utils/helpers
• Module name as "sound.py" or something
• Public getters and setters for access from main application (bool is_muted)
• Public "Mute" method for Mute/Unmute with possibility to set parameter like Mute(true)
• Public "Say" method for voice output (Apple's "say" command only actually when running on mac)
• Public "PlaySound" method for playing sound effect from robbie/sounds
• In consoly.py run all sounds and output all voices using this method
• Add startup sound support

• Add bot integration command
• Remove bot integration command
• Asking another bots
• Getting response from another bots

Multiplatform desktop GUI application with input, voice recognization support and nice design.

• Remove Pannous service
• Remove Pandora service
• Remove mentions from categories too

• Ask for term meaning
• Only first paragraph is returned
• "Tell me more" category
• "Source: Wikipedia" after definition

Ask for term meaning

Robbie - "What you want talking about? Nikola Tesla, Organizations...?"

User - "About organizations"

Robbie - "What you want talking about? Tesla Inc., Tesla (Czechoslovak company) or Tesla Science Center at Wardenclyffe?"

User - "About Tesla Inc."

Robbie - "Tesla, Inc. (formerly named Tesla Motors) is an American automaker and energy storage company co-founded by Elon Musk, Martin Eberhard, Marc Tarpenning, JB Straubel and Ian Wright, and is based in Palo Alto, California. The company specializes in electric cars, their powertrain components, and also produces battery charging equipment."

User profiles with users switching ("Hello, here is Thomas. Can you do some notes for me?", "Hi Thomas, what do you want to note?")

• "Source: Wolfram Alpha" at end of answer

• Sound OK after successful command or understanding of sentence
• Sound Failure after error, problem or not understanding of sentence (default text)

• Add requests service

From e-mail

Hello jiaaro,
I like your pydub sound library for Python. It works like a charm. But I have one issue. I have console application and I want let it clean as possible. Only user inputs and application outputs. But everytime when sound plays, something like this shown:
avplay version 11.4, Copyright (c) 2003-2014 the Libav developers
built on Feb 11 2017 01:10:02 with Apple LLVM version 8.0.0 (clang-800.0.42.1)
Input #0, wav, from '/var/folders/fx/y_8sjl8d48g51x2c9g_55yyw0000gn/T/tmpvwr2374p.wav':
Duration: 00:00:00.11, bitrate: 1539 kb/s
Stream #0.0: Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
1486777806.72 A-V: 0.000 s:0.0 aq= 0KB vq= 0KB sq= 0B f=0/0
Is there something for playing of sound in silent mode. Another play method or parameter? Thank you for answer.
Tomáš Tříska

I would suggest creating a copy of the play function in your own codebase and setting stdout to sys.devnull (or adding a flag to reduce the verbosity of ffplay)

• Best answer is returned
• Only first paragraph is returned
• "Tell me more" category
• "Another definition" category
• "Source: Answers.com" after definition

• Music stored to Robbie/Libraries/Music

Queues

• "speaking" queue is using 2-dimensional array with texts and priority. Priorities are Disabled, Low, Normal and High. Low are used for not much important info like Robbies random conversations, normal for standard results and high for very important informations like "I have no power source and battery is on 10%" for example. Disabled is used for special cases, when user can tell "Disable battery notifications" for example. First item in queue is actual output which stays until text is written and spoken. After that, first item is removed and first one with highest priority is processed using "speak" method in "speaking" thread. Exclude disabled, which are not removed completely, but waiting to potential change back to Low, Normal or High. Priority can be changed using modules except first one in queue which is running actually.

Threads

• "speaking" thread is special thread for speech of text. It takes texts from "speaking" queue, displays them, use text to speech on them and take another one. When no items in queue, it waits.
• "listening" thread is special thread for listening of inputs from user. When command for stopping of speak is used, text with stoping of talking is added to "speaking" queue and actual text to speed method in "speaking" thread is stopped.
• Other threads have own names and are used for modules purposes.

Safely wrapped geolocation function

• AdminLTE integration
• Console

• OS detection

• USB
• SD card reader
• Restart
• Autoload
• Intel Edison programming in Python 3 support
• Arduino modules support

• Set timer function
• OS support primary (for example MacOS calendar)
• Inbuild timer method with alert sound when no native support

• Fix Unicode in YAML
  or
• Change default "o" in "R🤖bbie ➜" to "🤖" in code
  and
• Change ">>>" to "➜" in code

Protocols

• TCP/IP
• COM
• JSON
• GET
• POST
• Bluetooth
• LAN
• IR
• Remote file

Features

• Add sensor/device
• Remove sensor/device
• Turn on sensor/device
• Turn of sensor/device
• Send data to sensor/device using protocol
• Get data from sensor/device using protocol

Bootstrap template
Robbie info (Age, Name, sensors, devices, dictionary size etc...)
Protected using Password, PIN or security questions

• "What time is it" category after #20

Starting sound when lauching using console after #13

• Robbie waits for his name keyword
• He listens and write text to input after few seconds of silence and message is send

• Lauching and terminating of web console using console.py as new thread instead of starting sh file