Giters

chrisl8 / Orac

Personal service to do stuff.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

This is a personal project to make my own sort of "home assistant" system. Named Orac after the intelligent computer in Blake's 7.
I did not build this as a general purpose tool, but instead as a sort of playground for me to just write code to do things that I want to do with it.

I do not expect this code to work generically for anyone else, but it is here if in case it can help you as examples.

Personal Development Notes

TODO: 2. Notify if car doors are not locked. 3. Notify if lights are on or doors open after a certain time of day. 3. Enable finding items by searching obsidian documents. 4. If it loses track of me sometimes, use devices (like car) to override current isHome status when received, if they are far away.

Alert me when my watch is fully charged, so I can put it back on.

Check if:

  • Watch is not fully or almost fully charged
  • and I haven't charged it today
  • and I appear to be in my office
    • Not known to be away from home and
    • Motion sensor recent
    • Lights on in office
  • then alert me to charge my watch.

Some events happen on HA. It has its own "automations" but I find them clunky, but could I have HA essentially CALL this API instead of polling a LOT? Or should I just poll A LOT?

If the wall switch in the office is turned off:

  1. Turn off all office lights.
  2. Turn the wall switch back on. ;)

When watch is 100% charged, alert me to put it back on. battery_status: 'Charged', battery: 100,

Things HA is doing by itself already: Turn on office lights when Office motion happens.

Pi Setup

Raspberry Pi Install and Configuration

  • Use the Raspberry Pi Imager to install Raspberry Pi OS (64-bit)
    • If you want to, you can use the Config icon (gear) to customize some things:
      • Set host name to Orac
      • Enable SSH
        • Add your Key if you want to
      • Set up my username/password
    • and boot it up, no need to edit anything on the config before first boot
  • Initial boot
    • If you set things up above, it should just boot right into xWindows with no password prompt
  • Configuration
    • SSH In and
      • Install my dotfiles setup, reboot, updateAllTheThings.sh, and reboot
      • Run sudo raspi-config (Note: There is a GUI version of this, but it seems to be missing several options.)

Set up Orac on this Devices

Use syncOrac.sh to sync code to Pi

cd ~/Orac/
./setupAndUpdate.sh

Reboot and everything should work.

Auto Login

If the auto-login to GUI quits on you, run sudo raspi-config and turn the Auto Login OFF and ON again and reboot, and it will fix it. System Options->Boot / Auto Login

DO NOT DO STUFF BELOW HERE It is just for reference

Old Setup Info

This is old junk from when I was trying to use all of the hardware on the Mycroft Dev Kit, but it was not reliable.

  • Run sudo raspi-config (Note: There is a GUI version of this, but it seems to be missing several options.)
  • NOTE: You do not need to enable Legacy Camera Mode, and should not. It is only required for old Python scripts, such as Mycroft.
  • Install Dotfiles per Readme instructions and also run updateAllTheThings.sh and reboot.

Fix up boot config

Run diff ~/Orac/pi/boot/config.txt /boot/config.txt to see what you need to tweak,
then edit or replace /boot/config.txt and reboot.

Replace command: sudo cp /home/chrisl8/Orac/pi/boot/config.txt /boot/config.txt

Install audio drivers and libraries

See further down for full information. This is the "quick start"

cd
pip install smbus2
sudo apt install raspberrypi-kernel-headers espeak
gh repo clone OpenVoiceOS/VocalFusionDriver
cd VocalFusionDriver
sudo cp xvf3510.dtbo /boot/overlays
cd driver
make all
sudo mkdir /lib/modules/6.1.21-v8+/kernel/drivers/vocalfusion
sudo cp vocalfusion* /lib/modules/6.1.21-v8+/kernel/drivers/vocalfusion
sudo depmod 6.1.21-v8+ -a

Make Orac run on startup, but from GUI so Audio works

Note: If a user autostart file exists at /home/pi/.config/lxsession/LXDE-pi, then the System autostart file is totally ignored (for that user). Reference: https://forums.raspberrypi.com/viewtopic.php?t=294014

mkdir -p ${HOME}/.config/lxsession/LXDE-pi
cp /etc/xdg/lxsession/LXDE-pi/autostart ${HOME}/.config/lxsession/LXDE-pi/
vi ~/.config/lxsession/LXDE-pi/autostart

and add this at the bottom: @/home/chrisl8/Orac/startpm2.sh OR if you want the terminal to open and see it: @lxterminal -e bash /home/chrisl8/Orac/startpm2.sh

If it isn't already done, add: pm2 log to the bottom of startpm2.sh to prevent if from just shutting down immediately.

boot config References

https://www.raspberrypi.com/documentation/computers/config_txt.html https://github.com/raspberrypi/firmware/blob/master/boot/overlays/README

Hardware Information and Reference

SJ201 Information

SJ201 Board

The board on the top of the Mark II dev kit is called the SJ201.

Hardware Information https://github.com/MycroftAI/hardware-mycroft-mark-II/blob/master/mark-II-Rpi-devkit/KiCAD/SJ201-Raspberry%20Pi%204%20Daughterboard/README.md

The primary information about it is here: https://github.com/MycroftAI/hardware-mycroft-mark-II but there is a more thorough testing suite that should help guide how to connect to all the features here: https://github.com/MycroftAI/mark-ii-hardware-testing

Verify SJ201 board exists and find version

References: https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/init_hardware.sh https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/hardware_tests/auto_detect_sj201.py

View all i2c devices:

❯ sudo i2cdetect -a -y 1
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00: -- -- -- -- 04 -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 2f
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- -- -- -- -- -- -- -- -- -- -- -- -- 6d -- --
70: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

tiny_address = "04" xmos_address = "2c" ti_address = "2f"

"Old" SJ201 have "04" (tiny) The documentation says they have "all 3" but the "2c" (xmos) seems to go away sometimes, and the presence of the "tiny" is the key to knowing this is an "old" or V1 SJ201.

View all spi devices:

❯ ls -l /dev/spidev*
crw-rw---- 1 root spi 153, 0 Feb 28 19:41 /dev/spidev0.0
crw-rw---- 1 root spi 153, 1 Feb 28 19:41 /dev/spidev0.1

Fan

Reference: https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/hardware_tests/old_fan.py

Hardware speed range is appx 30-255. sudo i2cset -a -y 1 0x04 101 0 i # Stop sudo i2cset -a -y 1 0x04 101 255 i # Full speed

For some reason we always need -a to see 04 both with i2cdetect and i2cset

Touchscreen

Turning on SPI and/or I2C (not sure which) should have allowed touchscreen to work.

  • Test it in xWindows, and it should work.
    • Try opening Menus
    • Try inside web browser
      • Both should work

Touchscreen from Terminal

/dev/input/event0:      raspberrypi-ts
  • Note that if set up correctly, that is the ONLY device listed apart from possibly any keyboard/mouse you connected.
    • While without SPI or I2C turned on, you will see an array of devices for some reason, none of which really do anything.
  • Select it, and tap on the screen, and it should output all sorts of fun data about touch/release events!
    • It does multi-finger-tracking, etc.

I'm not sure yet if I'll try to like display an image and track touches on their own (as Mycroft did),
or try to use a web browser interface and use the browser touch input. For now, just know it works.

LED Ring on top

Reference: https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/hardware_tests/ledtest_old.py

Use i2cset to set all LEDs to white:

i2cset -a -y 1 0x04 0 250 250 250 i
i2cset -a -y 1 0x04 1 250 250 250 i
i2cset -a -y 1 0x04 2 250 250 250 i
i2cset -a -y 1 0x04 3 250 250 250 i
i2cset -a -y 1 0x04 4 250 250 250 i
i2cset -a -y 1 0x04 5 250 250 250 i
i2cset -a -y 1 0x04 6 250 250 250 i
i2cset -a -y 1 0x04 7 250 250 250 i
i2cset -a -y 1 0x04 8 250 250 250 i
i2cset -a -y 1 0x04 9 250 250 250 i
i2cset -a -y 1 0x04 10 250 250 250 i
i2cset -a -y 1 0x04 11 250 250 250 i

Use i2cset to set all LEDs to red:

i2cset -a -y 1 0x04 0 250 0 0 i
i2cset -a -y 1 0x04 1 250 0 0 i
i2cset -a -y 1 0x04 2 250 0 0 i
i2cset -a -y 1 0x04 3 250 0 0 i
i2cset -a -y 1 0x04 4 250 0 0 i
i2cset -a -y 1 0x04 5 250 0 0 i
i2cset -a -y 1 0x04 6 250 0 0 i
i2cset -a -y 1 0x04 7 250 0 0 i
i2cset -a -y 1 0x04 8 250 0 0 i
i2cset -a -y 1 0x04 9 250 0 0 i
i2cset -a -y 1 0x04 10 250 0 0 i
i2cset -a -y 1 0x04 11 250 0 0 i

Use i2cset to set all LEDs to green:

i2cset -a -y 1 0x04 0 0 250 0 i
i2cset -a -y 1 0x04 1 0 250 0 i
i2cset -a -y 1 0x04 2 0 250 0 i
i2cset -a -y 1 0x04 3 0 250 0 i
i2cset -a -y 1 0x04 4 0 250 0 i
i2cset -a -y 1 0x04 5 0 250 0 i
i2cset -a -y 1 0x04 6 0 250 0 i
i2cset -a -y 1 0x04 7 0 250 0 i
i2cset -a -y 1 0x04 8 0 250 0 i
i2cset -a -y 1 0x04 9 0 250 0 i
i2cset -a -y 1 0x04 10 0 250 0 i
i2cset -a -y 1 0x04 11 0 250 0 i

Use i2cset to set all LEDs to blue:

i2cset -a -y 1 0x04 0 0 250 250 i
i2cset -a -y 1 0x04 1 0 250 250 i
i2cset -a -y 1 0x04 2 0 250 250 i
i2cset -a -y 1 0x04 3 0 250 250 i
i2cset -a -y 1 0x04 4 0 250 250 i
i2cset -a -y 1 0x04 5 0 250 250 i
i2cset -a -y 1 0x04 6 0 250 250 i
i2cset -a -y 1 0x04 7 0 250 250 i
i2cset -a -y 1 0x04 8 0 250 250 i
i2cset -a -y 1 0x04 9 0 250 250 i
i2cset -a -y 1 0x04 10 0 250 250 i
i2cset -a -y 1 0x04 11 0 250 250 i

Use i2cset to turn all LEDs off:

i2cset -a -y 1 0x04 0 0 0 0 i
i2cset -a -y 1 0x04 1 0 0 0 i
i2cset -a -y 1 0x04 2 0 0 0 i
i2cset -a -y 1 0x04 3 0 0 0 i
i2cset -a -y 1 0x04 4 0 0 0 i
i2cset -a -y 1 0x04 5 0 0 0 i
i2cset -a -y 1 0x04 6 0 0 0 i
i2cset -a -y 1 0x04 7 0 0 0 i
i2cset -a -y 1 0x04 8 0 0 0 i
i2cset -a -y 1 0x04 9 0 0 0 i
i2cset -a -y 1 0x04 10 0 0 0 i
i2cset -a -y 1 0x04 11 0 0 0 i

Switch

Reference: https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/utils/async_button.py raspi-gpio get 25

You can flip the switch back and forth and alternate between: GPIO 25: level=0 fsel=0 func=INPUT pull=DOWN and GPIO 25: level=1 fsel=0 func=INPUT pull=DOWN

Buttons

References:

Camera

Reference: https://www.tomshardware.com/how-to/use-raspberry-pi-camera-with-bullseye

In X Windows run:
libcamera-hello

  • You can run this via SSH, but you won't see the output! It should open a screen and show you the output from the camera as a video, while streaming data about the images to the terminal.
  • If you want to have it stay open indefinitely, run libcamera-hello -t 0

To take a picture and save it run:
libcamera-jpeg -o test.jpg

Speaker

Things I've tried:

Note: The "bcm2835" is the Raspberry Pi onboard chip (Broadcom), so if you see that audio devices, it is NOT the SJ201 speakers.

sudo vi /boot/config.txt UN-comment the line

dtparam=i2s=on

COMMENT the line:

#dtparam=audio=on

In other words turn on i2s and turn off the default audio device and reboot.

Reference: https://forums.raspberrypi.com/viewtopic.php?t=350269 sudo apt install linux-headers

References: https://github.com/OpenVoiceOS/VocalFusionDriver/tree/main https://github.com/OpenVoiceOS/ovos-image-arch-recipe/blob/master/docker_overlay/overlays/03_sj201/setup_sj201.sh https://github.com/xmos/vocalfusion-rpi-setup/blob/master/setup.sh cd git clone https://github.com/OpenVoiceOS/VocalFusionDriver.git cd VocalFusionDriver sudo cp xvf3510.dtbo /boot/overlays cd driver make all sudo mkdir /lib/modules/6.1.21-v8+/kernel/drivers/vocalfusion sudo cp vocalfusion* /lib/modules/6.1.21-v8+/kernel/drivers/vocalfusion sudo depmod 6.1.21-v8+ -a modinfo -k 6.1.21-v8+ vocalfusion-soundcard

Should show:

filename:       /lib/modules/6.1.21-v8+/kernel/drivers/vocalfusion/vocalfusion-soundcard.ko
alias:          platform:vocalfusion-soundcard
license:        GPL v2
author:         OpenVoiceOS
description:    XMOS VocalFusion I2S Driver
srcversion:     BD3D235129B2BE11A13B7A6
alias:          of:N*T*Cvocalfusion-soundcardC*
alias:          of:N*T*Cvocalfusion-soundcard
depends:
name:           vocalfusion_soundcard
vermagic:       6.1.21-v8+ SMP preempt mod_unload modversions aarch64

sudo dtoverlay xvf3510

aplay -l

Should now include the xmos/vocalfusion device:
**** List of PLAYBACK Hardware Devices ****
card 0: vc4hdmi0 [vc4-hdmi-0], device 0: MAI PCM i2s-hifi-0 [MAI PCM i2s-hifi-0]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 1: vc4hdmi1 [vc4-hdmi-1], device 0: MAI PCM i2s-hifi-0 [MAI PCM i2s-hifi-0]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 2: sndxmosvocalfus [snd_xmos_vocalfusion_card], device 0: simple-card_codec_link snd-soc-dummy-dai-0 [simple-card_codec_link snd-soc-dummy-dai-0]
Subdevices: 0/1
Subdevice #0: subdevice #0

and this should play some White Noise python ./mark-ii-hardware-testing/utils/set_volume_tas5806.py 370;aplay /usr/share/sounds/alsa/Noise.wav

sudo vi /boot/config.txt ADD the line:

dtoverlay=xvf3510

And you won't have to run dtoverlay after rebooting.

Unfortunately, now the Touch Screen is not working. evtest will show no raspberrypi-ts entry.

You must also: sudo vi /boot/config.txt COMMENT the line:

dtoverlay=vc4-kms-v3d

In other words turn OFF all other audio devices (HDMI) and reboot.

aplay -l Should now show only one Device:

**** List of PLAYBACK Hardware Devices ****
card 0: sndxmosvocalfus [snd_xmos_vocalfusion_card], device 0: simple-card_codec_link snd-soc-dummy-dai-0 [simple-card_codec_link snd-soc-dummy-dai-0]
Subdevices: 1/1
Subdevice #0: subdevice #0

Microphone

So far I have NOT made the Microphone work, but maybe I don't need it.

Old Information

Grab the Mycroft test code in case you need it

cd
git clone git@github.com:MycroftAI/mark-ii-hardware-testing.git

Possible audio helps from Mycroft code:

NOTE: I'm not 100% sure whether the above is all I need,
as I also edited init_hardware.sh to change /home/pi/venv to /home/chrisl8/venv
and then ran:

cd ${HOME}/mark-ii-hardware-testing
init_hardware.sh

and rebooted.

You may have to run this on every boot?

cd ${HOME}/mark-ii-hardware-testing
init_hardware.sh

Test sound driver and speakers

Reference: https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/utils/init_tas5806.py https://github.com/MycroftAI/mark-ii-hardware-testing/blob/main/utils/set_volume_tas5806.py

i2cdump -y 1 0x2F W

sudo python ${HOME}/mark-ii-hardware-testing/utils/set_volume_tas5806.py 200

Test buttons

Test LEDs

Test microphone array

Mycroft Dev Kit on OpenVoice Operating System

Getting all the hardware to work on a bare installation of Raspberry Pi OS wasn't working, so I tried this OVOS image that runs on the Mycroft Dev Kit and everything works!

This seems to be the way it is released: https://community.mycroft.ai/t/openvoiceos-a-bare-minimal-production-type-of-os-based-on-buildroot/4708/414?page=21

I downloaded this image: OpenVoiceOS_rpi4_64-gui_dev-202301171328.img on https://drive.google.com/file/d/1PUtNXfZ5jMUlVAgyN-KXPdVdX6r51eBw/view?usp=share_link on March 3 and installed it.

I think that you can also get the images from the links here: https://openvoiceos.github.io/community-docs/getting_image/

Documentation for OpenVoiceOS is here: https://openvoiceos.com/

Install image using the Raspberry Pi Imager.

I plugged into Ethernet, so it didn't ask about Wi-Fi or network setup.

Select "No Backend"

Once it boots up, reboot right away because the first time it starts MyCroft twice, causing everything to be VERY SLOW.

Use the menu system to enable SSH. Default credentials 'mycroft/mycroft'

Rename it: sudo vim /usr/lib/systemd/system/hostname.service Comment out existing hostnamectl command and replace it with one to name it Orac:

#ExecStart=/bin/sh -c "hostnamectl set-hostname OpenVoiceOS-`sed 's/://g' /sys/class/net/eth0/address | tail -c 7`"
ExecStart=/bin/sh -c "hostnamectl set-hostname Orac"

and reboot of course.

See if it will run Orac:

  • Install node

cd mkdir bin cd bin wget https://nodejs.org/dist/v19.7.0/node-v19.7.0-linux-arm64.tar.xz tar xvf node-v19.7.0-linux-arm64.tar.xz rm node-v19.7.0-linux-arm64.tar.xz ln -s node-v19.7.0-linux-arm64 node

Add this to .bashrc: export PATH=${HOME}/bin:${HOME}/bin/node/bin:${PATH}

Exit and reconnect: node -v

  • Copy unison over so that we can sync it

scp dotfiles/pi64/unison* mycroft@orac:./bin/

I cannot find any way to add a new path to the pre-login connection, so link to this file from the default /usr/bin path

sudo ln -s /home/mycroft/bin/unison /usr/bin/

Make sure that the IP address in syncOrac.sh is correct, and run:

syncOrac.sh

  • Run Orac

cd Orac/node npm i node server.js

.config/systemd/user/orac.service

[Unit]
Description=Orac Server
Wants=network-online.target
After=network.target network-online.target

[Service]
Type=forking
PIDFile=/home/mycroft/.pm2/pm2.pid
ExecStart=/home/mycroft/Orac/startpm2.sh

[Install]
WantedBy=multi-user.target network-online.target

NOTE: We did NOT ask systemd to ever restart pm2. The assumption is that pm2 is smart enough to do that itself, but if not, we could ask it to.

systemctl --user daemon-reload # To allow systemd to see the service exists systemctl --user status orac systemctl --user start orac

And to make it start at boot: systemctl enable --user orac

To see logs: journalctl --user -f

TODO:

  • Change the wake word
  • Have it not say "I'm mycroft" but "Orac" on boot

Trying the "Manjaro" OpenVoice Operating System (OVOS) instead of the Buildroot based image

https://community.mycroft.ai/t/openvoiceos-a-bare-minimal-production-type-of-os-based-on-buildroot/4708/440?u=chrisl8

This is an Arch based image, but a FULL Linux install, rather than the "Buildroot" thingy, so it should be nicer for me to

Download latest Manjaro-based image from: https://downloads.openvoiceos.com/images/manjaro/stable/

I'm using "mark2" edition, assuming that means/includes the Dev Kit. This is the download link I used: https://downloads.openvoiceos.com/images/manjaro/stable/

  • Install image using the Raspberry Pi Imager.
    • Note that it requires at least a 32GB card. 16GB is too small and will appear to just fail to image the card.
  • Insert SD Card and boot up.
    • I plugged into Ethernet, so it didn't ask about Wi-Fi or network setup.
  • Select "No Backend"
  • Use Default Speech Engines

Swipe from top down to see menu for

  • Brightness

  • Volume

  • Home, Wifi, Rotation, Microphone, Settings, Restart, Reboot

  • Do an initial reboot of the system to ensure everything is working.

    • I found that my initial settings didn't stick after the first reboot, but they did after the second.

  • Use settings to customize system

    • Customize -> Pick the color you like
    • Use About to find the IP Address Note that SSH is already enabled by default, so there is no need to enable it in the config menu.

Connect via SSH

https://openvoiceos.github.io/ovos-technical-manual/manjaro/#core_configuration

You can initially SSH in as EITHER ovos or root and the password is ovos in both cases. ssh ovos@192.168.1.157 or ssh root@192.168.1.157

Display info during boot

edit /boot/cmdline.txt Remove the words quiet splash

Fix-ups

shadow.service won't start until these exist:

sudo mkdir /home/pulse
sudo mkdir /home/pulse-access

Mycroft likes this to exist, although it never gets used: sudo mkdir /var/log/mycroft

Force a fsck check:

shutdown -Fr now The -F forces a check disk.

Use Gparted to expand the filesystem.

If it is broken when you restart

This is what finally worked to put it back into a working state: Log in as root.

cd
git clone https://github.com/OpenVoiceOS/ovos-image-arch-recipe/
cd ovos-image-arch-recipe/docker_overlay/scripts
chmod +x run_scripts_mark2.sh
./run_scripts_mark2.sh
shutdown -r now

And it seems good again!

System

Use systemctl to see if everything is working or not.

Logs

To see what is going on run: journalctl --user -f

Set up my Public Key

sudo pacman -S vim
mkdir .ssh
vim .ssh/authorized_keys

Insert SSH Public key

chmod go-rwx -R .ssh

Test Public Key login and if it works then sudo vim /etc/ssh/sshd_config and set

PasswordAuthentication no
PermitEmptyPasswords no
PermitRootLogin no

and also set a strong password for yourself and root

Rename System

hostnamectl set-hostname orac Reboot

Install Node.js

I'm trying this as the user ovos for now.

cd
mkdir bin
cd bin
wget https://nodejs.org/dist/v19.9.0/node-v19.9.0-linux-arm64.tar.xz
tar xvf node-v19.9.0-linux-arm64.tar.xz
rm node-v19.9.0-linux-arm64.tar.xz
ln -s node-v19.9.0-linux-arm64 node

Add this to the very top of .bash_profile:
export PATH=${HOME}/bin:${HOME}/bin/node/bin:${PATH}

Exit and reconnect: node -v

Profile

Move any .bashrc files to .bash_profile, which is the correct filename.

Sync Orac to Device and Setup

From Remote scp dotfiles/pi64/unison* ovos@192.168.1.157:./bin/

On Orac sudo cp /home/ovos/bin/unison* /usr/local/sbin/

From Remote again syncOrac.sh

On Orac

cd Orac/node
npm i
node server.js

Make it run automatically, npm i -g pm2

as a Service

mkdir -p .config/systemd/user
vim .config/systemd/user/orac.service

[Unit]
Description=Orac Server
Wants=default.target
After=default.target

[Service]
Type=forking
PIDFile=/home/ovos/.pm2/pm2.pid
ExecStart=/home/ovos/Orac/startpm2.sh

[Install]
WantedBy=default.target

NOTE: We did NOT ask systemd to ever restart pm2. The assumption is that pm2 is smart enough to do that itself, but if not, we could ask it to.

systemctl --user daemon-reload # To allow systemd to see the service exists
systemctl --user status orac
systemctl --user start orac

And to make it start at boot: systemctl enable --user orac

To see logs: journalctl --user -f

Set your timezone timedatectl set-timezone America/Chicago

Reboot to test that it starts up.

Configuration

The primary config file seems to be /etc/mycroft/mycroft.conf See here for all the defaults and what you can change them to, since the file only has "differences" in it: https://github.com/OpenVoiceOS/ovos-config/blob/dev/ovos_config/mycroft.conf Edit it and:

  • Add this near the top: "system_unit": "imperial",
    • To get Fahrenheit instead of Celsius temp display and spoken.
    • Note that the example site above is WRONG, suggesting to use "english".

You can set the log level to "DEBUG" to see a lot more info about what files are written and read.

Mostly instead of mycroft- tools use ovos- tools.

Package updating for the Insane

I'm not suggesting this is a good idea, but you can force all the Python packages to update, and force pip to use the latest code from GitHub for OVOS by doing this:

  • Be root sudo su -

  • Update all OS packages: pacman -Syu

  • Update pip /usr/bin/python -m pip install --upgrade pip

  • Update setuptools, as it is often out of date and breaks things pip install --upgrade setuptools

  • Use the unpinned requirements file to force everything to install with latest versions possible pip install --upgrade -r /home/ovos/Orac/ovos-setup-stuff/requirements-unpinned.txt

Now reboot, because Ovos needs to start over.

Now you shouldn't see a lot of out of date packages when you run:

pip list --outdated

and the few that are probably are a result of some dependency of other packages in the list.

If you want to know WHY something is still old, run: pip show

It will show what is requiring it, and then you can look in that package's requirements.txt file to see what version it wants.

If something is listed as old, but just isn't in requirements-unpinned.txt, you can add it, and it should update if you run the pip ... --upgarde command again.

FYI: This file started when I had a corrupt system and had to do a force reinstall of all packages: pip freeze > requirements.txt pip install --upgrade --force-reinstall -r requirements.txt (Don't do this now.)

Noted old stuff and who done it: Flask-WTF: https://github.com/flask-extensions/Flask-SimpleLogin/blob/main/pyproject.toml more-itertools: https://github.com/OpenVoiceOS/ovos-ocp-files-plugin/blob/dev/requirements.txt ovos-lingua-franca: https://github.com/OpenVoiceOS/skill-ovos-homescreen/blob/dev/requirements.txt ovos-PHAL-plugin-alsa: Not sure, it seems the GitHub repo shows an older version than pipy? ovos-stt-plugin-chromium: Not sure, it seems the GitHub repo shows an older version than pipy? ovos-stt-plugin-server: Not sure, it seems the GitHub repo shows an older version than pipy? ovos-tts-plugin-mimic: Not sure, it seems the GitHub repo shows an older version than pipy? Pillow: https://github.com/OpenVoiceOS/ovos-PHAL-plugin-oauth/blob/dev/requirements.txt pyee: https://github.com/MycroftAI/mycroft-messagebus-client/blob/master/requirements.txt PyGObject: OS maintained. Also, not sure what it is used for. PyYAML: https://github.com/OpenVoiceOS/ovos-config/blob/dev/requirements/requirements.txt qrcode: https://github.com/OpenVoiceOS/ovos-PHAL-plugin-oauth/blob/dev/requirements.txt smbus: OS maintained it seems, and I don't think anything even uses it timezonefinder: https://github.com/NeonGeckoCom/neon-utils/blob/dev/requirements/requirements.txt watchdog: https://github.com/OpenVoiceOS/ovos-core/blob/dev/requirements/requirements.txt websockets: https://github.com/OpenVoiceOS/ovos-PHAL-plugin-homeassistant/blob/dev/requirements.txt

Run Dashboard

TL;DR: It doesn't work, even after you fix it up.

Update the Dashboard

cd /usr/local/share
rm -rf ovos-dashboard
git clone https://github.com/OpenVoiceOS/OVOS-Dashboard.git

export PYTHON_SYS_SITE_PACKAGES=/usr/lib/python3.10/site-packages
export PYTHON_USER_SITE_PACKAGES=/home/ovos/.local/lib/python3.10/site-packages
export MYCROFT_SKILLS_LOCATION=/home/ovos/.config/mycroft/skills

About

Personal service to do stuff.

Languages

Language:JavaScript 79.1%Language:Python 9.4%Language:C++ 7.9%Language:Shell 2.8%Language:CSS 0.4%Language:HTML 0.3%