mike_wooskey

Can someone please paste the configuration yaml for an M5Stack Atom Echo that is currently working as a voice assistant satellite for Home Assistant using open wake word? (remember to redact your private data)

I installed a bunch of these Echos following these instructions back when they were first introduced. They have tiny little speakers which are essentially useless, but I routed the output sound to a media player in the same room. At that time the only option was to use open-wake-word - Nabu Casa even provided instructions and virtual-gpu resources to train your own wakeword, which I did.

The Echos worked great with my own custom wakeword for a long time!

Then sometime within the past year they stopped working. Their "wake word" entity is unavailable. I know Home Assistant came out with micro-wake-word, but I wasn't aware they ended (or broke) the open-wake-word functionality. I'd love to get my Echos working again - I'd prefer to use open-wake-word, but if the only way they're going to work now is with a pre-set wakeword, that's better than no voice assist at all!

Thanks for any help you can provide.

I used to do very basic "script-kiddie" coding about 30 years ago, but haven't really done any coding since then. Until about 6 months ago when I started being able to use LLMs to vibe-code. I speak my ideas and LLMs do a remarkable job of creating them.

I understand enough logic and basics of common programming language structures (e.g., functions, variables, loops) to be able to do a little creating on my own and some editing of what LLMs create, but LLMs are good at correcting themselves if I have a little patience and ask them to iterate for a while (usually with a guided goal).

It's absolutely remarkable that this technology exists in my lifetime. I feel like we're way closer to Star Trek's "Computer, do this thing" than I ever thought I'd actually see.

Sometimes I let the LLM choose the stack to use and sometimes I tell it what I want. But how far away are we from lanugages not mattering at all? How long until I say "make a program that does X and host it here" and LLMs use whatever stack is appropriate and create and host the result, and the language and dependencies don't matter to me?

And one step further: how long until LLMs (or probably more general AI technology) create languages on their own and humans don't even participate in that low level of programming any more?

I want to install smart blinds on 4 windows, from small (2'x3') to kind of large (4'x5').

I need them blinds to be local-only and integrate into Home Assistant.

I'm looking at SmartWings, but blinds with a z-wave motor for one of my large windows would cost Just under $400. That seems a bit expensive to me (especially considerin that I got my current non-motorized, non-smart blinds for those windows for like $50 each - and they're good blinds!). Is that just what smart blinds are going to cost? Or did I land on an expense option?

Do you have a preferred make or model of local-only smart blinds?

Is anyone familiar with gum?

I have a simple:

while true; do
    CHOICE=$(gum choose "one" "two" "three")
    case "$CHOICE" in
        "one")
            <do something>
        "two")
            <do something>
        "three")
            <do something>
    esac
done

If the user hits ESC, gum displays "nothing selected" and exits.

Is there a way to disable the "nothing selected"? I tried redirecting 2>/dev/null but it suppresses the entire menu. I also tried adding $'\x1b' to the case, but it did nothing.

I also tried adding:

    if [ -z "$CHOICE" ]; then
        exit 0
    fi

directly after CHOICE=$(gum choose...).

No dice.

Cross-posted from "finance management software suggestions" by @mike_wooskey@lemmy.thewooskeys.com in !foss@beehaw.org

I used to use Quicken long ago but have been using Moneydance for many years to manage my personal finances. But Moneydance is propietary software and I prefer FOSS now. What's your preferred finance management software?

Thing I need it to do:

• manage different accounts (common types are banks, credit cards, loans, assets)
• have basic reporting (e.g., categorized expenses per time period)
• preferably uses tags
• export or copy reports or data - could be excel or csv or something that can be pasted or imported into spreadsheets
• self-hostable with a web app

I'm not a fan of software that's budget-focused. I don't mind it having some budgeting functionality, but I don't want opinionated software to force me to manage my money a certain way. I just want felxible software to help me manage my software how I want to manage it.

I'm wary of source-available/freemium/dual-licensed/open-core licenses. It can't hurt to suggest such apps if you like them and I'll take a look, but I think it's not likely that I'll buy into that philosophy.

So what software do you use and like?

I used to use Quicken long ago but have been using Moneydance for many years to manage my personal finances. But Moneydance is propietary software and I prefer FOSS now. What's your preferred finance management software?

Thing I need it to do:

• manage different accounts (common types are banks, credit cards, loans, assets)
• have basic reporting (e.g., categorized expenses per time period)
• preferably uses tags
• export or copy reports or data - could be excel or csv or something that can be pasted or imported into spreadsheets
• self-hostable with a web app

I'm not a fan of software that's budget-focused. I don't mind it having some budgeting functionality, but I don't want opinionated software to force me to manage my money a certain way. I just want felxible software to help me manage my software how I want to manage it.

I'm wary of source-available/freemium/dual-licensed/open-core licenses. It can't hurt to suggest such apps if you like them and I'll take a look, but I think it's not likely that I'll buy into that philosophy.

So what software do you use and like?

I requested to schedule an annual wellness exam with my doctor, they said "She's booking out into 2026, would you still like to schedule?"

They're scheduling out a YEAR in advance!

I realize this is likely a regional issue to some degree, but I have a hunch that in the US (the country I'm in) it's very difficult to get a appointments for non-urgent issues. It's also incredibly difficult in my area to be able to find a doctor that's accepting new patients!

Why can't medical facilities afford to expand to support the demand? This isn't meant to open a discussion about healthcare - it's just a rant. Grrr!

After a recent update to ESPHome, my ATOM Echos no longer show a wake word in Home Assistant and the wake word field isn't even editable:

And indeed they aren't responding to me speaking the wake word.

Has anyone encountered this? Does anyone know of a fix or has a suggestion?

I've got this command in a bash script:

TEST=$(curl -o /dev/null -s -k -w "%{http_code}" -u "${USERNAME}:${PASSWORD}" "${URL}/dashboard/")
echo "${TEST}" #debug

When the script runs, the output is "000".

When I run the same curl command from the shell, the output is "200" (which is correct, since the URL is valid).

I verified that the USERNAME, PASSWORD, and URL vars are being passed to the subshell.

I'd appreciate it if you could point out what I'm doing wrong here. :)

UPDATE: This has been solved.

Hi.

This works fine in the template editor:

{% if states('sensor.apollo_mtr_1_cca750_presence_target_count') | int > 0 %}
  Detected
{% elif is_state('binary_sensor.apollo_msr_2_c7bfe8_radar_zone_1_occupancy', 'on') %}
  Detected
{% elif is_state('binary_sensor.apollo_msr_2_c7bfe8_radar_zone_2_occupancy', 'on') %}
  Detected
{% elif is_state('binary_sensor.apollo_msr_2_c7bfe8_radar_zone_3_occupancy', 'on') %}
  Detected
{% else %}
  Not Detected
{% endif %}

But when I try to create a binary_sensory from it in configuration.yaml, I get "'binary_sensor' is undefined":

template:
  - binary_sensor:
    - name: "Lounge Presence"
      state: >-
        {% if states('sensor.apollo_mtr_1_cca750_presence_target_count') | int > 0 %}
          Detected
        {% elif is_state('binary_sensor.apollo_msr_2_c7bfe8_radar_zone_1_occupancy', 'on') %}
          Detected
        {% elif is_state('binary_sensor.apollo_msr_2_c7bfe8_radar_zone_2_occupancy', 'on') %}
          Detected
        {% elif is_state('binary_sensor.apollo_msr_2_c7bfe8_radar_zone_3_occupancy', 'on') %}
          Detected
        {% else %}
          Not Detected
        {% endif %}
      device_class: presence

I'm probably using incorrect syntax or something. Can anyone help me with this?

I host a website that uses mTLS for authentication. I created a client cert and installed it in Firefox on Linux, and when I visit the site for the first time, Firefox asks me to choose my cert and then I'm able to visit the site (and every subsequent visit to the site is successful without having to select the cert each time). This is all good.

But when I install that client cert into GrapheneOS (settings -> encryption & credentials -> install a certificate -> vpn & app user certificate), no browser app seems to recognize that it exists at all. Visiting the website from Vanadium, Fennec, or Mull browsers all return "ERR_BAD_SSL_CLIENT_AUTH_CERT" errors.

Does anyone have experience successfully using an mTLS cert in GrapheneOS?

[SOLVED] Thanks for the solution, @Evkob@lemmy.ca