Watson-Intu Self Docker container build for x86 (amd64) Requires sound card, mic, speaker (or headphones), and camera (webcam preferred)
Dev version (leaves build artifacts in place, retains utilities like vim) # 5.9GB image
docker build --squash --build-arg "mode=dev" -f Dockerfile.x86 -t openhorizon/amd64-x86-intu-self:edge .
(omit "--squash" if you're not using docker experimental features)
Prod version (cleans up build artifacts, removes vim and unncessary deps, cleans apt-cache) # 1.9GB image
docker build --squash --build-arg "mode=prod" -f Dockerfile.x86 -t openhorizon/amd64-x86-intu-self .
docker pull openhorizon/amd64-x86-intu-self:edge # approx 1.9GB
docker pull openhorizon/amd64-x86-intu-self-dev:edge # approx 5.9GB
-
Clone this repo, and cd to the ./config directory. Use bootstrap-example.json as a starting point
-
Set up an IBM account with credentials for Watson services, according to this guide
-
Copy your Watson services credential pwds/URLs into bootstrap-example.json, save as bootstrap.json.
-
On your x86 system, run
aplay -l
andarecord -l
, identify the integer number of your sound card. Edit the alsa.conf file with these values. (see below for an example, if your sound card number is '2'.) Save the file.defaults.ctl.card 2 # approx line 79
defaults.pcm.card 2
-
Run Self in a container, linking your current config files:
docker run -it --rm --privileged -p 9443:9443 -v $PWD:/config openhorizon/amd64-x86-intu-self:edge /bin/bash -c "ln -s -f /config/bootstrap.json bin/linux/etc/shared/; ln -s -f /config/alsa.conf /usr/share/alsa/alsa.conf; ./bin/linux/run_self.sh"
-
When Self starts, it will verbally greet you. You can browse to
http://localhost:9443/www/dashboard#/
to see the Self dashboard. (substitute your TX2's IP address in place of "localhost" to visit the UI from another computer on LAN)