I don’t think you read through properly or are you telling us that has_value works on the states value and not passing in the entity? Because from what I gather from the docs, it only works on the entity you pass in, not the states value.
It accepts entity ids only, no need to grab the state before hand
He want’s to check against the trigger.from_state, has_value I can only assume will give the current state aka trigger.to_state
You can use trigger.entity_id. But he can filter out unknown/unavailable in the trigger without templates
Sure he can do has_value(trigger.entity_id) and test against it’s current state but that’s not what he’s after… He is after the state it was before it’s current state. The triggers also check against the current state or you could manually add before states but we can’t assume his automation or logic.
Assuming he has a trigger for a state change which is anything but doesn’t want the automation to happen if it was previously unknown or unavailable.
Mine was not. I also do not have a weather section in my configuration yaml. I had a darksky platform sensor.
1 Like
State triggers can use not_from: with a list of states that he doesn’t want to trigger from if the from state is in that list.
not_from:
- unknown
- unavailabile
2 Likes
We have more work to do to reduce the size of MariaDB in future versions. I suggest waiting to migrate until June or later, as the MariaDB size should shrink again, making migration easier.
We may eventually offer a built-in migration, but it’s likely months or years away if we can work out the performance challenges.
You can create a new home-assistant_v2.db file with default schema by starting up HA without MariaDB, and you’ll get a fresh home-assistant_v2.db in the correct format (assuming one does not already exist).
You could then shut down, clean up any data from the sqlite database, and copy over your data from MariaDB.
This process is a significant manual pain; there isn’t a great way to do it now.
Keeping 2x the space available for table rebuilds, auto repacks, schema changes, etc., is generalized SQL database advice and not specific to Home Assistant.
Adding it to the release notes when we migrate would give the impression that you only need it during migration.
If that’s something you feel would be valuable and help others. It might be an excellent contribution to add it to the recorder docs home-assistant.io/source/_integrations/recorder.markdown at 12e29a81f6a75e909e2cc1e5e06644bd85c5e74c · home-assistant/home-assistant.io · GitHub
2 Likes
He might choose that route if that’s his goal, hopefully he see’s your post
Guys I have 2 homes, so I have added second server in companion settings, but with regards to this declaration “They will all show up in the app” I can’t find my second server unless I go to settings-companion app and activate server from there. Is this normal or second server shall showup somewhere else for easy of access, just like Xiaomi and other smart home apps do, there is drop down at the top and you can chose other homes.
Also if for instance I buy Nabucasa account, why other added homes of mine do not show there, as I have entered my account credentials in both home assistant installations? (Im still using trial).
Im using static IP services, generaly speaking I do not need nabucasa, I can go with duckdns, but two homes with static IP already makes more than nabucasa claims for subscription, so it makes sense to buy subscription but only if it will support more than one homeassistant instance. Of cause would make sense for security reasons not to use any cloud service, but I can survive if that is financially beneficial.
3 finger swipe to the left-right changes between servers when you have multiple registered
Bit of a false positive here that might benefit from a ‘regex comparison’ rather than a ‘string contains’ check.
i.e. /\b(entity_id)\b/
That check is a bit naïve to avoid the overhead and complexity of parsing the SQL.
That wouldn’t match states.entity_id but I think we could make it a bit safer by changing it to match entity_id instead
In minecraft server the aiodns was upgraded to 3.0 in 2021, so there is something wrong in yiur system.
Oh great! Thanks…I’ll hold tight for now and explore migrations in the future. As someone that has no desire to be a deep DB admin, I’d like to move back towards the preferred DB.
I’d be happy to make a suggestion based on what I have learned from this experience with the 2023.4 release, however I only know about this from the video of the release notes and from this post, which was prompted because of the 2023.4 migration update…
Maybe I’m the only one here, but I could imagine there are others that do not know that they should always keep 2x their database available in free space for the future. And it was repeated a few times in the video, so clearly it is an important point. It still would have been good to include this in the release notes and clarify that it’s always the recommended advice, not just during the migration. It clearly isn’t common knowledge to everyone, but it should be considering there have been others with larger databases that hit the same issue that I did.
With that said, I really do appreciate your responses, insights, and willingness to help make this part of HA better for everyone.
1 Like
The amount of users that keeps adding identical issues to several custom cards that needed (and received ) updates for this release or keep adding ‘yes for me too it is still broken ‘ is staggering.
Seems there is a new wave of users not used to updating custom resources, and who won’t search how to update their own config, but simply throw in the ‘ it does not work ‘.
Not sure why I post that here, but having tried to help more than a few, I guess I’ve lost inspiration to search through completely irrelevant and huge card configs, or none at all ;—)
Maybe we should have a small pinned post here in the community with a tiny guide how to ‘break cache’ on those custom resources.
So we can point all of these posters to it and be done with it.
End rant.
2 Likes
First time I’ve ever had safe mode after upgrading to this release.
I have no idea whats causing the issue as when I try to look at the log I get this error…
failed to get logs, r is undefined
Can’t check the log to see what the problem is. I’ve checked the breaking changes but I’m lost.
Particularly with the ubiquitousness of HACS and the easy updates it offers.
2 Likes
I must say in their defence that if they use IOS it often happens that a js resource is sticky like hell. I reported an issue on a card a month ago and I had upgraded the component but it still failed. The dev told me to check my version and it was shown correctly in HACS but with further help I found that Safari was still holding on to the version. Clearing the cache is all or nothing and that makes your next month on the net miserable.
The fix when that happens is to enter the direct URL to the resource. And then reload it. That finally makes it update the cache. It happens rarely but it happens often enough to be a pain in the…
2 Likes
Well. I reduced my number of days from 730 to 90 to 30. The reasons were basically ignorance but also reflection. I took initially 2 years to compare year on year. But you only need macro info for that not micro. So I thought let’s take a quarter. But even that when I reflected I never look back more then a month. So I finally decided 30 days. No rocket science.
3 Likes