Hi !
Please open an issue at GitHub for such problems providing more information like logs (see also https://www.home-assistant.io/integrations/knx/#logs-for-the-knx-integration).
Its hard to find and keep track of such things when their only mention is the 239th comment on a very broad topic thread.
I am just getting started with HA and have no sensors like that, but I can definitely see a problem with masking lost sensor readings.
Next time flag the post, only a few of us are on here way too much.
My deconz sensors and lights are all messed up now. Lost the light levels of my Aqara motion sensors and quite a few devices showing in Phoscon not showing up on HA. Anyone else noticing this?
Probably just need to refresh your web browser.
3 posts were merged into an existing topic: Journal issues with supervised install after updating to 2022.11
Read a few posts above
https://community.home-assistant.io/t/2022-11-a-heck-of-a-release/483295/252?u=sparkydave
Nevermind, after updating Mushroom theme all good.
A post was merged into an existing topic: Journal issues with supervised install after updating to 2022.11
A post was merged into an existing topic: Journal issues with supervised install after updating to 2022.11
This is misleading. You use heavily customised dashboards and views with third-party cards. Who knows whatās affecting your performance.
after updating to 2022.11.1 last night, my cameras only pop up on press the same size as in my dash lol.
I already had it like this on 2022.11.0. Hope to have camera restored to previous behavior when it opened in full size.
BTW are there somewhere release notes regarding version 11.1 ?
EDIT
Ok, I see now
I have tested it in vanilla HA. Of course no difference. Where should it come from in average? Letās say you have a sensor which is not changing. Then with history you have to read and display two datapoints and a line and 1-2 hover points with display text per 24 hours. For LTE min max and mean, you have read and draw 24*3 datapoints with at least 72 hover points. Where should this extreme example be faster?
If performance would be a reason, a display of a history graph in history work center for 20 days and 100 entities should not be possible at all. But it is. Since āforeverā and displayed in microseconds.
Of course, this is extreme, but same the other way round. In average, there will not be any difference. This is unfortunately only a given reason to give any text.
haha, never short on hyperbolic statements you are arenāt you
Of course, you didnāt realize or cared to ask if this was experienced on a test setup or not. Which, of course, it was and is. Added to that, even in my production instance, thereās an extensive Test dashboard.
So yes, I quite extensively tested this new feature in the last beta cycle. Did you?
Yet, since you mention the heavily customized dashboards (which practically everyone uses I suppose, considering the thousands of posts in the community/reddit/discord/Facebook using the many available custom cards), even on those, the now legacy state more-info panel graphs were instantaneous. (with a few mentioned exceptions).
So, if the move to statistics graphs is about backend efficiency (which I would applaud vehemently) or frontend speed, it yet is not visible nor āfeltā in handling the Dashboards.
If it has to do with a move away from āstatesā centered backend, I could understand that too.
The dev/states page is the worst page in the system, and has been like that for a very long time. So things could be improved upon states handling, agreed on that. Not sure if this is related though.
Other than that, currently, personally, I do not experience any advantage in replacing the history state graphs with statistics graphs. And thatās an understatement.
never really noticed that tbh, but, if I click that popup (on desktop), it expands to to full view
No, HHCC plant sensors didnāt make it into this release, the PR is not merged yet into xiaomi-ble.
2 posts were merged into an existing topic: Journal issues with supervised install after updating to 2022.11