Just tried it out yesterday, the LLMs feature is amazing. It understands my intents far better than Alexa or Google and doesât pretend to know nothing about the device because of a misplaced command or âsâ. It also handles more complicated questions pretty well, e.g. whatâs the difference between outside and inside temp. Having the ability to ask non smart home related questions is just a nice side effect in my opinion.
For me, Home Assistant is the main smart home controller that assists me with all smart related things in my home.
Having hard to maintain expensive hardware is not a must at all. All the features like voice assist are optional and if you still want to use them, you can subscribe to HA cloud. Iâm personally running HA on a RPi 4 for years and it is still powerful enough to power the entire system without problems.
question regarding the new conditional section visibility:
wasnt the whole idea of those sections, that per device layout, one would keep the same order and placement of the sections in a view? As we dragged and dropped them in the view design.
with these conditional sections, that concept is no longer happening, as the view completely rearranges with the visibility of a view section, the same way it did before (sections were introduced)? Filling/replacing the now empty spot of that cards conditional visibility?
Not saying it shouldnât, its just that that behavior seems counter intuitive to the concept of Sections.
I have a person group. If only person entities are involved, the status is still home or not_home.
But I had also a generic sensor which was based on an input_boolean. I use this to manually set the status to home if nobody is at home. E.g. if a baby sitter is here.
So this was a template sensor, if boolean is on then home, if off then not_home. This broke now as it was setting the group.home to the on/off states.
I now switched to a fake person with a fake device tracker by using the device_tracker.see service.
Donât have paid subscription but getting the same error when I tried it out (Iâm currently having limited access to gpt-4o in the free version).
Google AI works.
Thanks, yes, I noticed the developer section, but since the change is also called out in the general blog section, there must be a user-related side to this.
Do you have an alarm control panel of sorts? Could you go to the entity settings and tell us if there is a default code setting?
As for myself, I have two alarm control panels defined in YAML using the manual alarm integration, which cannot be managed via the UI.
I was able to add a dummy code for all my automations, but still the tile card failed.
just disabling the requirement fixes the issue. I think this one should somehow be in repair of at least back ward incompatible changes because it took me a lot of time before I found your post
Speaking about professionalism.
Is there any SLA, protocol or whatever regarding âWithin what period a broken integration should be fixedâ?
Traccar server integration - was broken in 2024.5 and still not fixed.
Well, this is a free SW - but I guess there should be a normal & predictable order of fixing bugs.
No, you need to read up more about FOSS (Free AND Open Source Software). While you can probably predict how long it usually takes to fix and issue based on history, there is no garantees that an specific issue in FLOSS (Free Libre AND Open Source Software) will ever be fixed. Also understand that Home Assistent has thousands of integrations where most of those are developed by volenteers in their spare time as a hobby, and often primarly for their own private use and personal interest, and many times they just been generous and wanted share the project with the public as you might want to do with things you like/love. i can not speak for the developers for the the Traccar integration so this is more a general explaination, however I do not believe that any developer is employed by a company to specifically maintain the Traccar integration(?) or other integrations for third-party cloud services?
I am not a developer of this project but know from my experince of FOSS/FLOSS project that if you are not capable of fixing an issue yourself then what YOU can do is do your best to report the issue/bug to the proper channel so that it there can be tracked publicly (which is not here) + provide enough descriptive information there in that tracked issue so that others can read about it and maybe help confirm that it is an issue and perhaps try to find out how it can be replicated. In general, when reporting potential bugs related to an integration to the issues trackers, it is normally a good idea to try tl provide detailed and specific description + information in addition to the information requested by the standard issue template(s):
So no, there is no SLA (Service Level Agreement) you get included for free in FOSS/FLOSS software, and those free software that offer some kind SLA normally do it via a compnay that had full-time support employees with a helpdesk and developers that is in some way paid to do exactly that (with revenues coming via either paid-support or other means such as for example ads-revenues). Some very popular open source software (e.g., OSS distributed by Red Hat or SUSE) offer plenty of paid-for support where you can pay for various SLA contracts, and then they have employeed dedicated support teams and devlopers in place. Otherwise, users can find help best-effort through user forums and mailing lists or other channels. That is why most companies only use software that offer paid-for support that meets their SLA needs for their critical or time-sensative service.
Here is an article to get you started on reading up on what you get and do not get for free with open source software, and more importantly what to expact and how to best proceed to try to get help for free:
One thing to know about the Aladdin Connect Integration is that if you, like me, have gotten rid of the app and were waiting for the update, you must use it to update and generate the new API connection to their cloud services. Then, you can easily add your garage door/doors to HA.
After playing around with the Google Generative AI for a bit I decided to add the integration a second time so I could have two separate personalities. Nothing seemed to stop me doing this however there is no way to select which version of the Integration is used when in the settings for a Voice Assistant device.
Is this something that will come? Or should I have not been able to create two copies of the Integration in the first place?
Yeah but you can only pay Nabu Casa for their Home Assistent Cloud service (i.e. remote access + Google Assistent and Alexa cloud API integration), not for development of all the hundreds of integration componets that come with Home Assistent Core. Nabu Casa have no SLA on the maintainence and upkeep of all of those, most of those are only developed by community members as their personal hobby.
It was just meant as a general answer. I did also link to the proper way to report issues, which again is not here.
Still, the code owner of the traccar integration listed as being individual, not company, and there is no mention of service level agreements for it or other integrations, (regardless if the individual that develops an integration also happen to work for Nabu Casa or not).
As you could notice in my post, it was never mentioned that âNC is obliged to fix some particular bugâ. So no need to write long posts about it.
OK, hope this is your speculation.
Do not think this is needed.
Again - what for?
And do you think that MY issues for THIS integration were not registered properly?
I asked a simple question - and you wrote a long post.
OK, may be.
Then I will just hope that these bugs will be considered sooner or later.
There are some issues. I myself created 2 of them.
BTW, links you posted are not related to Traccar server integration. Probably you did not understand what integration I am talking about. So, your remarks are still âin generalâ and just very very partially related to the issue.
I have a usecase for a conditional background.
Iâve got a main dashboard, used by the phones, the pc and the tablet stuck to the wall in my living room. Behind the tablet, there is a very beautiful wallpaper. Iâd like to use an image of that wallpaper as a background for the dashboard, but only when opening the dashboard on the tablet, not on any other device (the tablet has a specific user, which I use to show a conditional battery card in the same dashboard).
you might be able to do all that using card-mod (themes), or some other custom utility.
the new UI background selector has no option for conditional options like that. It merely is a selector, and acts as a frontend to the already available background option in Yaml.
Which is not conditionable either.
