Just finished a small video series that shows how to integrate Mycroft.AI, Rhasspy, and Node-RED. It uses Mycroft Precise as the wakeword-engine and makes all skills from mycroft available in rhasspy/Node-RED. It touches on several issues and tries to give solutions and work arounds. You will also see me struggle with some small quirks in Rhasspy 2.5.
Enjoy and share your ideas and feedback here, on github, or YouTube.
Thanks for doing this, @ulno! I’m hoping some of my recent changes to rhasspy-voltron will help eliminate the installation quirks for virtual environment users.
We still have a ways to go before 2.5 is considered “stable”, but I think we’re just about to the “release candidate” stage. Your time spent testing and running issues to ground is much appreciated
Thanks for checking things out - quirks are always to be expected and just show the life in projects, so looking forward to making my way around more.
I was wondering if I should also give voice2json a go but I am a little confused about the future of it - I wonder if it could be not run as some kind of subproject of rhasspy?
If you look at the requirements.txt file in voice2json, you could say it’s arguably already a sub-project of Rhasspy
As for its future, voice2json is here to stay. I use it at work all the time, and often incubate new features for Rhasspy there. After working on something as complex as Rhasspy, voice2json can be a breath of fresh air with its limited command set. It’s definitely not for those unfamiliar with the command line…
This was sooo great!!! Especially the second video helped me out finally installing Rhaspy without having to use Docker.
Rhaspy is now working awesomely and I am implenting and upgrading my old snips skills (I made a translation, timer, volumecontrol and a mp3 player skill in the snips days)… I am so happy that I can finally continue developing now
