OpenVoiceOS - A bare minimal (production type of) OS based on Buildroot

Release Spotlight: Backend UI

A benefit to hosting the back end for OVOS is better control over the systems within, no longer relying on another party to uphold promises of privacy or for server uptime. However a drawback comes in the form of it being command line focused, with little in terms of visual feedback. But thanks to the work of @JarbasAl, running a local back end now comes with the ability to monitor and manage it with a UI!

Check out the Github page here: GitHub - OpenVoiceOS/ovos-backend-manager: A simple web UI for personal backend, powered by PyWebIO

What does the UI do for me?

It was possible before to run a local back end for OVOS, but in order to set various settings you would need to do so via a command line. This is a barrier for entry for some people, passing the line of wanting better privacy but requiring too much investment to manage such systems. The user interface for the back end allows people to not only configure settings within the local back end for OVOS, but to also view metrics and information about the system.

OVOS is meant to allow developers to create voice activated systems that respect your privacy. But we also want to enable as many people as possible to partake in this mission. Creating easier to navigate systems plays an important role in allowing people to participate. Raising the bar for FOSS voice activated systems, lowering the barrier for entry.

2 Likes

Announcing: OVOS Community Docs

No project is complete without documentation directing the users how to use said project, and OVOS now has it’s very own Community documentation github page! Check it out here: GitHub - OpenVoiceOS/community-docs: community maintained documentation

What can I find in the docs?

Right now? Well as I write this, it includes sections for understanding what OVOS is, why it was made, what some terms mean, some developer FAQ, some compatibility reading, and some additional reading. But these are community docs, meaning that if someone would like to make additions or changes, they can easily submit a pull request and participate in growing the knowledge base!

What can I contribute to the docs?

Anything and everything you find relevant! The information presented is meant to answer questions users and developers have about the system, whether they’re juts getting started or want to dive deep into the systems and components that makes OVOS tick. Users who wish to make a change can do so via a pull request. We encourage those who want to get involved to feel free to suggest changes and improvements.

How can I get started?

  1. Read the docs
    • Simple as that. Read them and see where that leaves you in terms of understanding.
  2. Contribute where you see fit
    • Create an issue if you have a question, or create a pull request(pr) if you have something to contribute.**
  3. Enjoy.

Call to action

Want to get involved? Check out our website! Find links to our Matrix chat and other information about our work and find ways to get involved!

1 Like

Announcing the Release OVOS 0.0.5


As these dot releases do we find ourselves with a handful of helpful fixes, as well as a plethora of bug fixes. These major changes includes a kick…uh…butt new listening mode!

update to Log location
#210 - ovos-core will now automatically write logs to $XDG_STATE_HOME/mycroft/xxx.log, until now the launcher scripts were responsible for writing and rotating logs, this often caused permission issues or images to suddenly run out of space, we now ensure logs directory is writable and enforce a limit of 50MB for logs, you can see all logging config options here

i’ll be ready in a minute
#211 - mycroft-core had a hardcoded dialog “please wait a moment as i finish booting up”, this was not actually related to the boot process and misleading, ovos-core now plays a error sound instead, other components may reuse the same sound to indicate failures if appropriate. the sound file can be changed under mycroft.conf just like the start_listening and acknowledge sounds

ch-ch-ch-ch-changes
#204 - the helper method ask_yesno was very limited, it has been updated and will now handle things like double negations and changing your mind mid sentence, the logic for this has now been moved to ovos-lingua-franca OpenVoiceOS/ovos-lingua-franca#30

end to the backend
#189 + #201 - backend has now been abstracted away, by default ovos-core uses no backend but multiple backends are supported, check out our new backend docs Backend Services - Documentation

now you listen here
#150 - new experimental listening modes have been added, you can now completely disable wake word or not require it for follow up interactions, this is not enabled by default and should be considered only a preview


Full list of changes can be found on our Github, or just below these words:

Fixed Bugs
Implemented enhancements

Come hangout

Want to get involved? Check out our website at openvoiceos.com. Find links to our Matrix chat and other information about our work!
4 Likes

GUI and Skill Updates For OpenVoiceOS

Some interesting projects to test and run on your OVOS powered images / devices

Home Assistant PHAL Plugin:

Introducing Home Assistant PHAL plugin it provides GUI interfaces and API for Home Assistant Instants. The plugin is in early development, so there are some features that are not yet implemented. It currently supports the following entities:

  • Media Player: Start / Pause / Change Tracks / View Status
  • Light: Turn On / Off / View Status
  • Vacuum: Turn On / Off / View Status
  • Binary Sensor: Status Updates
  • Sensor: Status Updates
  • More device support and features will be added, need community help to get more actual operation data for different type of devices

Demo:

Available to Install For Testing From:

Routine’s Skill:

A new routines skill is available for testing:

  • Supports create schedules to run utterance one or several action on specific days at specific times
  • GUI and Voice interface to easily add and remove routines
  • A dashboard to view and manage current routines

Demo:

routines-skill-demo

Available To Install For Testing From:

Food Wizard Skill:

Introducing Food Wizard, a recipes skill, it was a skill around for a while as a demo for GUI API usage, but now has been converted into a fully functional recipes skill.

  • Adds voice and ui supports for ingredients and cooking instructions (cooking instructions for supported recipe providers)

Demo:

Available To Install For Testing From:

3 Likes

Took some time over the holidays to work on the documentation of the OpenVoiceOS - Buildroot image.

Not completely finished as of yet, however here it goes;
https://openvoiceos.github.io/community-docs/buildroot/

Added documentation for most of the different features of OpenVoiceOS including the Spotify, Airplay, KDE Connect and other small things that are included by default.

2 Likes

Can this also be loaded to the Mark 2? I assume using the RPI4 version?

yes, mk2 is one of the target platforms, we have been supporting ovos in the mk2 devkits since the beginning

This looks very interesting. I’m looking for a way to make my Mark 2 more usable and Spotify would be the icing on the cake. Don’t mind at all if it’s a work in progress.

Could you tell me what I need to do to get it started please?

I downloaded a buildroot pi4 image and used the Raspberry Pi Imager to write it to a usb drive. I put it into my Mark 2 and plugged in the power. I saw an OVOS logo and a loading progress bar, that completed. Then another screen came with two revolving circles and a logo in the middle but it seems to be stuck there …

Did I miss something simple? Thanks.

Can you give this image a go?
https://drive.google.com/file/d/1GVC0HnxaudVRsTPT4_9KCQam3B7AVP_b/view?usp=share_link

Hi @j1nx ,

What is different about this image vs. the latest release in the beta channel? I am going to give you guys a go tomorrow when I have some time and wanted to see if anything fun or important was added.

Hi thanks for helping.

It gets stuck at the same screen but I noticed by accident that I can swipe down from the top of the screen and bring up a menu and brightness and volume sliders. The brightness control works, none of the menu works except the one that rotates the screen.

@dunkum Just a newer release with more fixes. This one specifically has the support of the sj-201 hat included.

I always include a timestamp at the end of the file so you know when it was build. Development is still going at a rapid pace so…

@postman Which device are you using? Just out of curiousity could you try booting with a LAN cable connected? WiFi has been a trouble from the start. Anyhow, will look into it.

1 Like

Nice! I appreciate the labor you are putting in. I’ll give you some feedback tomorrow or Sunday once I load it and take it for a spin.

@postman Jup, again wifi crap. Will look into it. It boots with lan cable connected before powering on.

Will report back asap

1 Like

Yep, with a LAN cable it goes on to the setup … and I could get into the settings and connect to WIFI successfully.

Seemed ok but doesn’t respond to any wake command or to the big button on the top and there’s a small circle in the top right of the screen that is spinning.

But I can ssh into it so will start looking around…

[update] I can’t figure out what’s wrong. It appears to be running OK, the home screen is working and so are the menus but the fan is running at full speed and it doesn’t respond to any sounds, wake words or otherwise. I rebooted into dinkum to check it wasn’t the hardware and everything works normally.

@j1nx Any advice on what to try to get OVOS working on my Mark 2? Thanks.

There is also a Manjaro based OVOS image for the mark2 hw and dev kit. It does not have all airplay, spotify stuff yet like buildroot, but everything else should be working. If anyone wants to give it a go: https://downloads.openvoiceos.com/images/ovos-dev-edition-mark2-161222-031558.img.xz

Nice to know there are options but Spotify and Airplay are the must have pieces for me if I am going to use this thing. Otherwise, I’ll have to go back to being spied on by Amazon or Apple as a trade for the device doing what I actually want.

@j1nx I’m using a Mark II … I think there’s only one model?

Ran into some connectivity check troubles which I currently am looking into.

Anyhow, I only have the dev-kit and not the real Mark2 retail device so interested in getting to the bottom if your issues as the support work for it was done kind of “blind” without the actuall device.

Considering that a lot of Mark2 buyers would like to try out our image wiuld be nice to get this sorted for all the others.

my mk2 should arrive next monday, so i will be able to test all that moving forward

2 Likes