If you are talking about the portal, it has had that since the initial “version” (it uses the host/browser to determine it), if you are talking about the theme setting in HA that’s not off topic and already noted by @SeanM?
I mean some switch, like in your HACS dokumentation, as i always prefer dark mode.
TBH those are just stupid and not needed.
Set your preference in your OS / browser and sites that have it should just use that.
Not for me, Ubuntu 18.04 with Compiz, and i love it.
Stats I would like to see… hmm how about…
-
Total/Average/Max of devices broken down by integration
-
Total/Average/Max of devices broken down by integration
-
Total/Average/Max of Integrations
-
Total/Average/Max of Automations and/or Scripts
-
Avg/Max Age of Install (by both time and/or number of times updated)
-
Avg/Max Uptime
It would be neat if you could filter multiple integrations: tasmota, esphome
google trend style (with a graph or without)
It’s going to be interesting to see these numbers change as non-beta users start sharing!
Kinda pointless now but it would eventually be interesting to be able to see all the stats over a time period to see things like how quickly a feature or integration is adopted when it comes out or if an integration use is in the decline etc etc.
Right now the maps shows all the installed instances per country.
I would like to see installed instances per capita, i.e how big percentage of the countries population is using HA.
I don’t understand this grammar. In general, this data is meant to be aggregate, and not allow to track you.
Yes, but it apparently knows at least which country the reporting users are in since the stats are broken down by country.
I think they are asking to take the reported stats for each country and dividing by the country’s population to get the percent of the country that is using HA or it’s components.
However, I think they are going to be disappointed in those results since say for the US that has a population of ~>330 million and only 1695 installations I’m not sure that knowing that .00000000514% of the population is using HA is very useful.
Plus, it would be non-representative to those who would have opted out of analytics, or haven’t updated. Also, the core repository has 41.8k stars, and the reddit has 118k members.
True but you’re going to have to factor that in to any decision based on the need to obtain those analytics in the first place. The whole idea of analytics is useless if you are concerned that it likely isn’t really a good representation of the users who didn’t report. If that’s the case then what’s the point?
I’ve never starred the repo and I am not a reddit member. so there’s that too.
Would love to see info on which addons are in use and how many users are using them. Also if possible, would be great if each addon included the url from its config so people could click through and discover addons others are using they weren’t aware of. Other then the core and community addon repos its pretty difficult to discover addons in other repos but there are a lot out there which are quite useful.
Also could be used to make a case for bringing an addon to the core/community repos if one is particularly popular.
This would only be for supervised and whatever else.
Right. So roughly 80% of the userbase would find it useful and 20% would ignore it
EDIT: Actually I could even argue the Container users might at least check it out. They don’t use or care about addons but if there’s an addon there’s probably a standalone docker image. Might be useful for discovering self-hosted software they weren’t aware of.
While I agree with the sentiment, I think that making add-ons/containers more visible is going to lead to more people (especially those running HAOS on a rPi3b) to bloat their instances which in turn leads to more posts about “My home assistant is [crashing|running slow|won’t load]”.
It’s tiring trying to explain to non-technical users the limitations of a SBC like the rPi and why they can’t run 15+ add-ons or containers on a SBC with a tiny processor and 1GB of ram. A lot of them get all bent out of shape about it and start blaming the platform.
Fair point. Although I would argue that at least for HAOS/Supervised people the damage is already done in that regard. Between the core and community addons repo there’s a pretty solid selection, more then enough to get people with underpowered hardware in trouble. I wonder how many issues like that the plex addon alone has caused
LMAO! I had forgotten all about the plex addon, but looking at the thread, seems people are still using and abusing it.
Most add-ons don’t bog a system down too much, it’s when people start doing video type stuff that it becomes an issue, like Plex as you mention, and another one is MotionEye…‘why can’t I record 6 HD CCTV streams on my RPi3…’.
I do like the idea of seeing the stats on add-ons being used.