@KristjanESPERANTO

Thanks! Maybe I take a look into the authenticating stuff …

@KristjanESPERANTO

AFAIS the wrong license is coming from the gitHubData.json file.

So looking into https://github.com/MagicMirrorOrg/MagicMirror-3rd-Party-Modules/blob/main/scripts/updateGitHubApiData.js I stumbled over the line
let maxQueryCount = 58; which is used in the shouldFetch function.

Maybe I’m wrong but this looks like we only fetch 58 repos from github and then switch to historical data?

@sdetweil

thanks for the hint, had to check mmos …

@p1lspeda said in MMM-PublicTransportHafas - Error when loading - DeparturFetcher.mjs:

I’ve upgraded to the latest MM Version

not really, 2.31.0-develop <> 2.31.0

I did a test with default calendar and custom.css:

.calendar div.module-content::after {
  content: url("/css/k13.svg");
}

or

.calendar div.module-content {
  content: url("/css/k13.svg");
}

which works (svg is visible).

So maybe something with your css concerning the child-stuff is wrong (my css knowledge is to bad to help here)

@com1cedric

the /icons subfolder will not work, you have to use one of these

We’re happy to announce that KristjanESPERANTO’s 3rd-party modules repository has found a new home at MagicMirrorOrg.

Thank you, Kristjan, for donating your repository to the public and for all the work you’ve put into it.

We have (hopefully) updated all links:

• Github repository: https://github.com/MagicMirrorOrg/MagicMirror-3rd-Party-Modules
• 3rd-party-modules website: https://modules.magicmirror.builders/
• 3rd-party-modules json file: https://modules.magicmirror.builders/data/modules.json

In the medium term, we plan to replace the currently used wiki with a different onboarding process for 3rd-party modules.

@smegbadger said in I cannot disable screen blanking with MagicMirrorOS:

@karsten13 thank you for all the effort you are putting into this issue.

Well, it is a pleasure to have such a good tester as you

I agree, let’s go back to the version that will kill and restart the container if the display doesn’t respond to wlr-randr (and then waits for a MMM-Universal-Pir cycle to sleep it again).

for now I did exactly this, but this issue is still working in my brain …

So to revert all the changes done before:

cd /opt/mm
git switch master
git pull

Change the images back to latest in .env:

MM_IMAGE="karsten13/magicmirror:latest"
LABWC_IMAGE="karsten13/labwc:latest"

Get new images and restart:

cd /opt/mm/run
docker compose pull
docker compose up -d

Last point: You should add the --transform 90 to the onCommand of the MMM-Universal-Pir config:

                {
                        module: "MMM-Universal-Pir",
                        position: "top_right",
                        config: {
                                deactivateDelay: 30000,
                                onCommand: "wlr-randr --output HDMI-A-1 --transform 90 --on",
                        }
                }

@smegbadger said in I cannot disable screen blanking with MagicMirrorOS:

Please check I did the pull etc correctly

yes

this seems to become a never ending story …

With the last changes I wrote the display state (on or off) to a temp file so the labwc container knows the last state even after a restart. Problem now is the mm container, here the logs after restart:

> magicmirror@2.32.0-develop start:wayland
> WAYLAND_DISPLAY="${WAYLAND_DISPLAY:=wayland-1}" ./node_modules/.bin/electron js/electron.js --enable-features=UseOzonePlatform --ozone-platform=wayland

[2025-04-04 21:19:30.050] [LOG]   Starting MagicMirror: v2.32.0-develop
[2025-04-04 21:19:30.084] [LOG]   Loading config ...
[2025-04-04 21:19:30.087] [LOG]   config template file not exists, no envsubst
[2025-04-04 21:19:30.608] [INFO]  Checking config file /opt/magic_mirror/config/config.js ...
[2025-04-04 21:19:30.674] [INFO]  Your configuration file doesn't contain syntax errors :)
[2025-04-04 21:19:30.675] [INFO]  Checking modules structure configuration ...
[2025-04-04 21:19:30.778] [INFO]  Your modules structure configuration doesn't contain errors :)
[2025-04-04 21:19:30.780] [LOG]   Loading module helpers ...
[2025-04-04 21:19:30.786] [LOG]   Initializing new module helper ...
[2025-04-04 21:19:30.788] [LOG]   Module helper loaded: MMM-Universal-Pir
[2025-04-04 21:19:30.788] [LOG]   No helper found for module: alert.
[2025-04-04 21:19:30.793] [LOG]   Initializing new module helper ...
[2025-04-04 21:19:30.793] [LOG]   Module helper loaded: updatenotification
[2025-04-04 21:19:30.794] [LOG]   No helper found for module: clock.
[2025-04-04 21:19:30.924] [LOG]   Initializing new module helper ...
[2025-04-04 21:19:30.925] [LOG]   Module helper loaded: calendar
[2025-04-04 21:19:30.926] [LOG]   No helper found for module: compliments.
[2025-04-04 21:19:30.927] [LOG]   No helper found for module: weather.
[2025-04-04 21:19:31.034] [LOG]   Initializing new module helper ...
[2025-04-04 21:19:31.034] [LOG]   Module helper loaded: newsfeed
[2025-04-04 21:19:31.035] [LOG]   All module helpers loaded.
[2025-04-04 21:19:31.043] [LOG]   Starting server on port 8080 ...
[2025-04-04 21:19:31.046] [WARN]  You're using a full whitelist configuration to allow for all IPs
[2025-04-04 21:19:31.681] [LOG]   Server started ...
[2025-04-04 21:19:31.683] [LOG]   Connecting socket for: MMM-Universal-Pir
[2025-04-04 21:19:31.684] [LOG]   Connecting socket for: updatenotification
[2025-04-04 21:19:31.685] [LOG]   Starting module helper: updatenotification
[2025-04-04 21:19:31.686] [LOG]   Connecting socket for: calendar
[2025-04-04 21:19:31.687] [LOG]   Starting node helper for: calendar
[2025-04-04 21:19:31.688] [LOG]   Connecting socket for: newsfeed
[2025-04-04 21:19:31.689] [LOG]   Starting node helper for: newsfeed
[2025-04-04 21:19:31.690] [LOG]   Sockets connected & modules started ...

The container is running but electron not (the line Launching application. is missing), I think it is waiting for an active display …

The Pir-Sensor which would active the screen again is running inside mm (which is not really running).

So I have no idea at the moment, maybe we have to live with the situation before, where after restart we need a sensor interval before the screen is blanked again.

@smegbadger

I think I have now a working solution for this.

For testing

use LABWC_IMAGE="karsten13/labwc:develop" in .env.

In /opt/mm do

# check if worktree is clean
git status
# switch to develop branch
git switch develop
# pull
git pull

Go to /opt/mm/run and execute docker compose pull and docker compose up -d --force-recreate.

With these changes the screen should still be blank after a restart of the labwc-container if the screen was blank before. Let me know if it works.

@smegbadger

had to write it down:

labwc	after start			wlr-randr --output HDMI-A-1 --transform 90
mm	starts and mm is displayed
mm	offCommand			wlr-randr --output HDMI-A-1 --off
mm	still up but not displayed
labwc	fails after a while, killed
labwc	after start			wlr-randr --output HDMI-A-1 --transform 90
mm	still up and displayed again
mm	offCommand			wlr-randr --output HDMI-A-1 --off
...

so another idea would be to execute the command defined in RANDR_PARAMS only if the display is not switched off. Will check if I can get this information …

@ajoreis said in Black screen on the radxa ROCK 2F, using docker:

Segmentation fault labwc

that tells you that labwc crashes, further investigation without having the used hardware is impossible.

@smegbadger said in I cannot disable screen blanking with MagicMirrorOS:

the labwc container is then killed and then automatically restarts - will the screen be turned on then until the next PIR activity and subsequent timeout of MMM-Universal-Pir?

yes, I tested this by killing and restarting the labwc-container from outside.

Not nice but I have no idea how to solve this.

From the sample config the normal feed url is https://rss.nytimes.com/services/xml/rss/nyt/HomePage.xml (they maybe have more).

My first thought was to use https://username:password@rss.nytimes.com/services/xml/rss/nyt/HomePage.xml but the problem here is another one. The feed is free so you need no authentication for the feed but some articles of the feed that should be displayed with ARTICLE_MORE_DETAILS are protected. Because the article urls are extracted from the feed I have no idea how to implement this at the moment.

@sdetweil

and I learned more about pi imager, the stuff it writes seems to depend on the OS, with my examples above I used fedora, now looking at a raspian pi os lite I don’t see the files.

There now is a userconf.txt with only user/password …

@sdetweil

files should be in same directory as config.txt and cmdline.txt. Maybe the OS deletes them after initial reading? So you may have to look for this files before you insert the sd-card into the pi …

example of user-data:

[mm@pi4-argon efi]$ cat user-data
#cloud-config
hostname: pi4-argon
manage_etc_hosts: true
packages:
- avahi-daemon
apt:
  conf: |
    Acquire {
      Check-Date "false";
    };

users:
- name: mm
  groups: users,adm,dialout,audio,netdev,video,plugdev,cdrom,games,input,gpio,spi,i2c,render,sudo
  shell: /bin/bash
  lock_passwd: false
  passwd: $5$wEw...
  ssh_authorized_keys:
    - ssh-rsa AAAA...
  sudo: ALL=(ALL) NOPASSWD:ALL


timezone: Europe/Berlin
keyboard:
  model: pc105
  layout: "de"

example of network-.con:

[mm@pi4-argon efi]$ cat network-.con
version: 2
wifis:
  renderer: networkd
  wlan0:
    dhcp4: true
    optional: true
    access-points:
      "k13":
        password: "cbd..."

@sdetweil

another interesting fact, if I open the settings in the imager to flash another
the password fields (user and wifi) are filled with junk text, maybe encrypted… but the values I entered are not visible

thats normal

which version of pi imager are you using? Latest which is v1.9.0?

And you are using the right user + passwd (I often forget to set the correct user)?

The pi imager writes a file user-data into the root dir of the sd-card which has yaml content, you can check the content (e.g. user) and also try to override passwd, maybe it works unencrypted, otherwise you can try to encrypt with htpasswd.

wifi has own file network-.con (yaml too).

@rkorell said in Standard Weather Module showing unwanted header:

appendLocationToHeader

read the docs, there is no property with above name …

@smegbadger

I build another new labwc image. You can revert the changes done to MM_IMAGE and the on-/offCommand in your setup.

The new karsten13/labwc:develop now runs labwc with pid=1 (I thought that was already the case) and additionally kills labwc if the wlr-randr command fails with failed to connect to display.

Now the container will restart if labwc is terminated inside the container. This does not address the root issue but should be a good workaround (hopefully).

So please do again a docker compose pull && docker compose up -d to get the latest changes …

for testing I added wlopm to the current mm-develop docker image.

So for another test use this lines in the .env file

MM_IMAGE="karsten13/magicmirror:develop"
LABWC_IMAGE="karsten13/labwc:develop"

Then change the on/off commands in the config.js

                        module: "MMM-Universal-Pir",
                        position: "top_right",
                        config: {
                                onCommand: "wlopm --on HDMI-A-1",
                                offCommand: "wlopm --off HDMI-A-1",
                        }

and then do a docker compose pull && docker compose up -d