Do you use ha_xiaomi_home integration? There was a change in v0.3.0
that changed the name of the entity ids.
I have several of the round ones and all still work for my Automationâs which are just triggers to turn on and off devicesâŚCan only guess different to your method.
Example Only (if it helps)
mode: single
triggers:
- device_id: 7a377acb6ed63ba687ffe0df47fe347d
domain: zha
type: remote_button_long_release
subtype: button
trigger: device
conditions: []
actions:
- type: turn_off
device_id: 9ee18c5c008085fbe651aa44695cba21
entity_id: light.lifx_a19_02
domain: light
So, this basically took me renaming every single entity in Home Assistant (changes within my ESPHome config didnât fix it, including adding a friendly name since you canât create friendly names for your entities in ESPHOME, you can only create one for your device, which is useless to cleanup this particular issue). I left all of the diagnostic entities alone as they werenât worth bothering to change.
All in all, it took about an hour and a half to correct once I got started, but more important than that, this wasnât in the breaking changes, so, it wasnât time I could properly plan for or something I could check for, but in the PR it very clearly indicates it wasnât going to be a change thatâd make some people happy, but it was worth it since it was a one-time thing.
I donât really understand why this isnât something that only applies at onboarding of the device ⌠in my opinion there is zero reason for HA to ever bulk change already existing devicesâ names without the owners permission, âkeeping it in lineâ with everything else or not once the device is created in HA and been there for years, itâs âin lineâ with the rest of MY devices and that should really be all that matters. I honestly donât think creating a device with an explicitly defined name and expecting it to be honored is all that weird either, but recognize that if ESPHOME becomes more commercially successful, that it would need to be more in line with the rest of the integrations.
Itâs also really weird that the devs will add whatâs essentially âif you have an automation to remove the persistent notification that you have new devices, it wonât work anymore because itâs not thereâ as a breaking change, but I had hundreds of entities that were all screwed up and it didnât even get an honorable mention in the Backward-incompatible changes and the only mention of ESPHOME at all is that itâs now reached platinum status, which ⌠cool, I guess, itâs great that this behavior is up to their highest level of standards.
Someone might say âhow is it a breaking change? It didnât change the entity ID or anything like thatâ â Well, I have announcements and other automation based around the names that I explicitly gave these devices.
Now, with all that said, is this really a big deal that this happened? No, not really, a one-time fix on my end in exchange for the ESPHOME integration getting better and now acting in accordance with the rules that other devices have to follow is objectively GOOD, but, without a breaking change notification itâs been significantly worse for me than it needed to be (and again, I donât see why this is done at all after onboarding a device is complete). I actually watch the release party videos and read the change logs and there was absolutely nothing in this release that would have indicated to me that this was happening, ESPECIALLY to already existing devices (at least not that I noticed/heard).
In other forums, Iâve had people say âIâm a single parent, I donât have time to fix HA every time the devs intentionally break something, so mine never gets updatedâ and Iâve argued that if you read the changelog and breaking changes youâll be fine, so, Iâm kind of feeling like I have egg on my face right now.
Good morning since I updated Home Assistant to release 2025.5.0 if the modem restarts or I restart it manually the Raspberry that has the SD Home Assistant loses the connection. To restart it I have to turn the Raspberry off and on again. On the same network there are 3 other Raspberry that reconnect normally. To test I also reversed the SDs between the different Raspberry, but when the modem restarts the Raspberry that has the SD Home Assistant always loses the connection to the modem. Does anyone have the same problem?
Congratulations on making home assistant more difficult to use, just for an integration rating. And failing to mention the change in the release notes.
You have completely underestimated the âlittle transition painâ.
Personally I face the choice of changing all my carefully chosen 422 ESP device entity names in the device configs or renaming them all in the HA UI.
I will probably have to use the UI as renaming them in esphome will cause the entity ids to change necessitating even more remedial work.
Iâve seen one person with 1000s of entities to rename.
Itâs just one of those rash decisionsâŚthey tend to happen lately more and more often⌠last one was changing backup to useless with mandatory encryption⌠but in that case at least they admitted the fault amd changed it.
If I had gone in and renamed all my diagnostic entities, Iâd easily be in the thousands. As it is, Iâm definitely in the hundreds. It took me an hour and a half of dedicated time to only rename the ones that are going to interface with other people besides me and/or be seen in the dashboards!
Personally, I donât see a way at all to make changes in the ESPHOME config and have them be honored in HA with the names I want them to have. Just because 4 entities are on the same ESP device doesnât mean I want them all pre-pended with the same name! Thatâs why I named them properly in my config in the first place!
If friendly_name were an option at the entity level (and that would actually be honored), that would work, but, itâs not, the friendly_name only applies at the device level (and a blank name doesnât get honored either, haha), which makes it useless for my purposes.
Again - I donât see why this is even a thing at all once the device is added. During onboarding if it does this and doesnât honor my name as-is, thatâs still not great, but at least then itâs only during onboarding that I have to deal with the need to rename.
ETA: In the Github issue that someone created, someone advised that if Iâd set the name in HA this wouldnât be happening right now, which âŚyeah, but, I made a custom device in ESPHOME, I named it the way I wanted there. Kudos to the people that make their devices more generic and then customize in HA, but, I make them the way I want them to be and in the past HA has respected that.
Please allow us to set a local mount as backup storage.
For users that virtualize, like proxmox and a HASS OS VM, we can now use vitiofs for âDirectory Mappingâ. So, instead of having to mount an NFS/SMB share for backup storage, I can instead bind mount the host storage into the VM.
I cannot save my backups to that location locally, I have to instead expose it via NFS and then set it up as an NFS share to store my HASS backup data there. If we could set a local directory for backups, I wont need all that abstraction and NFS to save my HASS backup data.
This might be helpful.
I havenât used the UI version that Coenenp built, I believe Iâve been using the edits he made to the CLI version as I just recently updated both entity_ids and friendly names in batch via a spreadsheet which made find and replace pretty efficient. Much better than the HA UI.
The 2025.5.1 update, which includes the frontend change to 20250509.0, seems to have broken the way my Simple Thermostat card displays the target temperature. On the Companion app, it appears in a very small font and cannot be adjusted.
@elchininet Nope, just using it âbare bonesâ in ZHA.
@myhass140 Iâm using AppDaemon. I set up all my automations in AppDaemon way back when the core automation process was extremely lackluster (e.g., you couldnât mix âandâ and âorâ statements), and Iâll probably never find the time to move even just the ones that can now be done in the âstandardâ automation. Thanks for the share, though.
Could you explain a little bit on this? I have 456 entities from ESPHome, and I donât understand the change nor the scope of what youâre saying.
I didnât update yet, but right now I donât have several hours to spend renaming things. So skipping for now, but it would be nice to understand the issue.
I am surprised that changes with so much impact dont get announced beforehand to give developers time to prepareâŚ
Iâm worse. I found out after a 6 he coding binge. Iâve changed hundreds of entities so the thought of backing things out to make a change before upgrading to put them back is terrifying
If you didnât update is great. Just follow THIS guide from petro and youâll be fine.
it also has steps if you updated and donât want to downgrade.
Thank you, I was completely lost on this. Who I should speak with to add this in bold in the breaking changes section of the release notes blog post? Itâs for a friendâŚ
Hey @Protoncek,
Yeah we did some checks against the cloud integration, because there is the cloud backup location. This was a bug that is fixed in 2025.5.1
This is an amazing QOL release. Scrolling badges, automation UI elements, backup settings, and especially the iOS dropdown list bug! Thanks to everyone who assisted in this release. Every month, Home Assistant just keeps getting better and better and I really appreciate everyone who helps make it happen.
Hi, for me the 2025.5 release breaks the energy dashboard unfortunately. From the looks of it my âbattery consumed ernergyâ is now actually âbattery consumed energyâ minus âexported energyâ (to the grid). A negative value is not displayed though. The net effect is that most of the battery consumption is gone from the stats, except at night time (when I donât export energy). My system is a seperate battery system (only supplies the house) and a seperate ongrid solar system. Anyone else having this problem?