None of those fix the issue.
Try the HACS card āmodern circular gaugeā. It has all these options and a ton more. Iām loving it.
My Two Bits : I only upgrade when the 0.3 or 0.4 revision is out . HA has become too complex . My opinion , start cutting down on integrations etc. Perhaps , integrations with less than 1000 installations can be given a goodbye !! Concentrate on the next the few months in cutting flab . Simplify !!
Works for me with exception of white background of selected item, when doing manual selection by keyboard arrows, the selection moves and is white. I f using mouse selection what is under cursor changes color properly.
I thing the issue here is that there are so many cross-references between variables (e.g. you have real definition of colour, then some elements are references to this variable and yet another element references to this one), so it hard to figure out at which stage colour is really defined. I use theme that was defined ~3 years ago and since then barely update for cases like this one. So perhaps in your case there is something in between, that overrides these variables⦠Just thinking loudlyā¦
Modern Circular Gauge is really good, but Ring-Tile Card is even better.
There are many alternatives to the core Gauge which do not even have a scale.
Does this mean previously locks like U200 or SwitchBot are natively supported? I still do not understand why they need an app to setup when any other device can be setup via matter natively.
Not sure, because i tried a lot of variables during the beta, but i think at the end it helped to make my theme dark.
from
darkorange_rk_pc:
primary-color: "#ff9800"
...
to
darkorange_rk_pc:
modes:
dark:
primary-color: "#ff9800"
...
After this, my 'input_*'s all where dark again.
Hello,
Since last Core update 2026.4.1 , the alarm panel stopped to work ( digit 1 displayed as undefined1)
The error log is
Logger: homeassistant.components.websocket_api.http.connection
Source: components/websocket_api/commands.py:321
integration: Home Assistant WebSocket API (documentation, issues)
First occurred: April 4, 2026 at 11:05:07 PM (2 occurrences)
Last logged: April 4, 2026 at 11:05:33 PM
[140612771289568] Invalid alarm code provided
Is it a known issue ?
Thank you
Thanks so much for this. I spent hours yesterday wading through āinspectā and with chatgpt trying to fix this via the theme file or using card-mod - it had dozens of attempts, none of which did anything.
According to the integrations analytics site, there are 775 integrations with <1000 installs. Thatās a lot of potentially pissed off people.
Love that IR support got added!
However, as a previous Seeed Studio XIAO IR Mate owner, Iām really missing the āLearnā functionality their previous firmware had. Made it super easy to pick up new remotes.
Iāve worked around this by finding my deviceās IR button definitions online (for example: https://lirc.sourceforge.net/remotes/logitech/Z-5500D) and modifying the ESPHome definition and reflashing.
My new IR Proxy ESPHome yaml for this device, including the Logitech speaker buttons, now looks like this:
substitutions:
name: xiao-ir-mate-aa1714
friendly_name: XIAO Smart IR Mate aa1714
packages:
Seeed.Xiao IR Proxy: github://esphome/infrared-proxies/xiao-ir-mate/xiao-ir-mate.yaml
esphome:
name: ${name}
name_add_mac_suffix: false
friendly_name: ${friendly_name}
api:
encryption:
key: abc123
wifi:
ssid: !secret wifi_ssid
password: !secret wifi_password
ap:
ssid: "ESPHome-Setup"
password: "esphome"
button:
- platform: template
name: "Speaker Power"
on_press:
- remote_transmitter.transmit_nec:
transmitter_id: ir_tx
address: 0x10EF
command: 0x08F7
- platform: template
name: "Speaker Volume Up"
on_press:
- remote_transmitter.transmit_nec:
transmitter_id: ir_tx
address: 0x10EF
command: 0x58A7
- platform: template
name: "Speaker Volume Down"
on_press:
- remote_transmitter.transmit_nec:
transmitter_id: ir_tx
address: 0x10EF
command: 0x708F
- platform: template
name: "Speaker Mute"
on_press:
- remote_transmitter.transmit_nec:
transmitter_id: ir_tx
address: 0x10EF
command: 0x6897
Normally, I would. However, my whole house went berserk at 3:30 in the morning. So, no, I didnāt have a chance to grab the logs before immediately reverting to a snapshot to get my house back to normal before waking everyone else up. ![]()
I suppose my hope is that at least one other person out there encountered such an issue (doesnāt feel alone haha) and did grab some more information. If not, Iāll await 4.2 and hope for the best. Iāll also install 4.2 a little earlier in my day so that I can actually keep an eye on it this time. Iām just glad I decided all those years ago to run HAOS in a virtual machine so that I can readily revert to snapshots when things like this happen. I genuinely feel sorry for anyone who runs HA bare metal on something like a Pi and runs into this type of catastrophic crash. ![]()
Iāll agree with the first half. I think Iām gonna start waiting until the final release of the month to upgrade. .0 and .1 have been particularly problematic for me over the last seven or eight months now. HA has gotten plenty large enough for things to start slipping through the cracks between final beta and .0. The only downside to this approach is what Iām about to mention nextā¦
As far as the second half is concerned, I think you are halfway right. I feel that HA has been a little too quick to bring new integrations in-house. Integrations with low adoption rates being integrated directly into HA as a native integration starts to become problematic because there isnāt enough testing done by the community before those changes make it into the next HA build. These days, when I see an integration move from HACS to HA-native, I always wonder about the future of that integration and how updates are going to be moving forward. Some of these integrations see updates weekly in an attempt to keep up with API changes at the 3rd party they are āhackingā access to.
The best example I have in recent times is Music Assistant. When the integration was not built into HA, it worked great because they could push updates out as quickly as they were putting updates out for the server add-on. But now, I often find that the MA integration will be slightly broken for a week or two after an MA add-on update because of a mismatch between the integration and the add-on. When the Integration was OoB, it got more frequent updates alongside the add-on.
Only the most stable and least frequently changing integrations should be brought into the fold. Anything beyond that starts creating problems for those of us using them. I suspect at least half, possibly two-thirds, of those using HA are well aware of and use HACS plenty. There is little reward to rush bringing any integration in-house, and plenty of risk to doing so.
I would recommend adding them to IRDB if they are not already added. If you only have the raw timing data, you can get the protocol data by using a tool like IrScrutinizer
Auto height for cards
This change has broken all my Grid-Layout dashboards.
How can I go back to the previous behavior in YAML-code?
Iām thankful for any help.
I agree.
The existing cards should get new options that will allow it to be backward compatible.
The bare minimum should be:
- Options to override the heading/title/name so all the cards can have the same text size (and color).
- Options to set all the card text to the same size and color
- Icons should be template-able
- Options to set card border and background
I agree - we dont want everything to break. But for the love of God - can we get cards that can be forced to look like they were designed by more or less the same people?
After many years of constant forced changes and updates, I am done using custom cards.
nice one, i currently use GitHub - guiohm79/custom-gauge-card: Dashboard homeassistant Ā· GitHub but ring tile looks to have some interesting options
I get the following warning in the Samsung Smart TV-integration, it presumably also impacts other integrations for network connected devices
āImplicit Wake-On-LAN is deprecated and will be removed in a future version of Home Assistant. Please create an explicit Wake-On-LAN automation to turn on your TVā
Does this mean you cannot turn on the TV from the media player power button anymore? What is the reason for this removal? If I understand it correctly, it seems like it will diminish the functionality quite a bit?
