New badges allow more customization for showing info, much less customization for styling and considered as small by some users. Myself never used both old and new ones, only posted tutorials about styling the old ones. More worried about a more general issue about interaction between devs and users.
This applies for both sides of the medal, has someone counted all reactions here in the forum as „research“?
I love the new badges a lot, so count me on the positive side
I used the old badges. I now use the new ones. They’re working out OK. In some ways better, in others not.
The problem I saw was the way they were “sprung” on users. It was a pretty dramatic change, and at first there were limited customization options. For example, no way to disable showing an icon. This made it impossible to consider them a direct replacement for the old style. That was, frankly, a bad design decision. If I display a temperature sensor, which shows a number, the “degrees” symbol and either C or F, I don’t need a little picture of a thermometer to be able to tell that’s a temperature. And, at first, there was no clear way to retain the old style badges while figuring out how to work the new style one into our design.
Those issues were eventually resolved. But the idea that a change like this could be forced on users who wanted to update, in an incomplete state and with little documentation, was the real problem.
Especially when there are hundreds of FRs on this forum, full of changes that users actually want.
Same for me. No update of switches in 2024.10.1
Thanks a mil.
How did you figure it out or found the solution? I couldn’t figure it out from my logs.
From one side, as it was already mentioned, people more report their negative observations than expressing their love in a technical thread.
From other side, majority of users use default settings and do not use complex customizations: partly because they do not know about such possible customizations, partly because they do not have advanced skills. So, majority of users have nothing to report. Contrary, users with advanced skills may experience lack of smth and report their observations.
It is true that there’s a bias when it comes to online feedback, but in this case many people have been quite specific with the challenges they now face – it’s not all just ranting.
Dumb luck mostly.
I saw the file not found error and assumed that uv
wasn’t installed to start. From the venv tried uv
, that worked, so thought maybe there was something with the way it was called. Modified util/package.py
to log the arguments it was called with, then ran uv
with the same arguments from the command line, that worked. I then thought maybe it was an issue with how Popen was calling uv
, so I pulled that out and ran it from the python shell in the venv and that worked. At that point I went back to util/package.py
and added logging for the path environment variable and then compared that to the one in the venv python shell and noticed they didn’t match. From there it was just finding the systemd docs and adding the proper path.
For some reason this update is causing a memory leak. Needed to downgrade in order to stop it from crashing.
No not the upgrade itself, it is probably one of your integrations that is causing this. I guess as @Nerdix (below) it’s your Alexa Media integration.
Alexa Media Player? If you scroll up this thread you find a lot of people having trouble with it…
I’m a big fan of Home Assistant and very appreciative of all the folks who work on it to make it so great and to become even better over time.
But the superlatives for every item in every new release are silly. Raving about how amazing every little tiny change is makes it difficult to identify which ones are actually significant.
But the purple prose doesn’t reduce my appreciation of Home Assistant and its team. I just mentally remove all the adjectives when reading the release notes to make my own evaluations on what changes are important to my setup.
As I said, new and old badges both have positive and negative sides. Also, the new badge looks very consistent with Tile and Heading card, allows to create nice views.
But the way what used to force users to migrate to new badges is wrong. And no alternative to keep using old badges was provided (that “custom-hui-badge-thing is not a 100% replacement).
So, saying about a “fantastic reaction” reminds me a story with a new HA logo:
NB: We made a research before replacing a logo.
Me: Where this research was done?
NB: In this thread.
Me: But this thread was created after you changed a logo.
Silence.
Can we please avoid this conversation again? It was already had by the exact same people in the last release thread. We do not need it here again.
Great detective work.
My first reaction too was to check if uv
is installed. I know the build process switched to it, because it’s so much faster. Then I thought that I should install HA Core now with uv
instead of pip
, but couldn’t find anything on that in the changelog or installation docs. At that point, I was a bit out of time, because I was hoping to do just a quick upgrade. Luckily, I found your post.
Btw, in my /home/homeassistant/.local/
I only have a lib
and share
.
And exact same people tried to shut them down.
This is not about badges (which I do not hate and consider them as one more useful feature), this is about “devs and users” communication.
Ildar, stop. Please.
Petro, I just want users be happy. Really. That is why I spend some amount of my time for supporting them.
So, please: less marketing, more communications with users.
I have an issue with my Zigbee buttons not working with 2024.10.1 and I had to revert back to 2024.09.3. using “ha core update --version 2024.9.3”
The issue is that all my button’s action sensor becomes unavailable. Is anyone else having this issue?
Zigbee2MQTT version 1.40.0-dev commit: 0d46354
Logs.
Got to say, mine are working normally. Just tested.
You don’t automate via the action sensor anyway.