It’s not going to improve if people continue to make up their own names or refer to it by its traits instead of its name.
I think you know I was not a fan of how the product’s name was redefined. Nevertheless, it’s a done deal and in the interests of minimizing confusion, it benefits the community to adopt the new nomenclature. Otherwise, we only have ourselves to blame for the confusion because we avoid using the official names consistently.
Totally agree with this!
Ummm… what are they again?
All variations available with BLT as an Add-On.
It’s only 10am here and I now I’m starvin’! 
In that case, I suggest the Triple-Docker or Supervised with BLT. In comparison, HassOS on rye is very small and lean.
As long as no APIs change with firmware updates or if you’re not using external polling data, yes you could stay on a version until the heat death.
Just an idea, but would it be an option to merge the supervisor/hassio with https://dietpi.com ? It is also a minimal linux distro that runs on a plethora of platforms and embedded devices and also uses docker containers for “apps”. And both addressing a quite similar target audience it IMO would be a win-win. So simply making HomeAssistant an app there. I know that there already is a HA core app for dietpie, but well, it misses the nice add-on integration of hassio.
Mostly likely the infrequent experiencing of any issues is due of all of the work that’s been continuously poured into making it as stable as possible.
While I hope that it will still be supported, I think their original message here is that the amount of work to continue to do so is unsustainable.
I’d like to know if an alternative solution to the problem would be for them to use e.g. Kubernetes as opposed to an in-house made supervisor (as mentioned in a more recent post in this thread). This is outside of my area of expertise though.
On the other hand, to be fair, I didn’t ask “what Home Assistant are you running?”. I asked “which install method did you use?”. In that way the poll questions do make sense with the exception that I should have split the “Core” question out into individual Docker and venv options. So I guess there should have been four options instead of three.
And as of right now we have a pretty clear winner and I don’t think it’s what the devs think it is.
But again in regard to the question that you seem to want to ask instead ( “what Home Assistant are you running?”) I can’t even figure out how to ask the question in a way that makes sense without asking for many more additional details for clarification.
I don’t think that what anyone does or doesn’t do is going to help clarify things with the current naming paradigm that’s being used.
If you start looking at how users are organically evolving how they talk about the names you can start to get an idea on how it should be called.
The problem with the poll is that it only reaches those who have come to this thread. It’s not necessarily representative of everyone who uses home assistant.
I don’t mean to downplay the significance of this install method being potentially deprecated though.
Also there may be many ha users have never even been on the forum. Also 120 votes (as ir stands as i write this) is statistically insignificant.
Also I say also too often.
Right, probably haven’t run into any issues due to all of the work that’s continuously being put into making it as stable as possible.
But it appears that they’re saying that this amount of required work is not sustainable.
Of course since I use this method as well for similar reasons & it’d be nice if it can continue somehow, but we’ll have to wait and see.
See the Kubernetes comment in this thread. Is there a good alternative to the in-house supervisor that already exists?
But it’s representative to the users who are here every day, read the subject of about every topic and try to help where they can.
The most of ‘everyone’ come to the forum, asks about something that ‘doesn’t work’, get help and will never read this topic.
EDIT: These people are the ‘first level support’ of HA. A company should never ignore their support team, they are a part of it.
Then “Disk-image” would’ve been appropriate because that’s the documentation’s recommended installation method (for Home Assistant). All other installation methods require the user to pre-install some distro of linux whereas that disk-image does not. The average user probably doesn’t know what’s inside of the image (i.e. HassOS, docker, several docker containers) other than it makes their computer run “Home Assistant”.
That’s a fair point.
I had already acknowledged the error in bundling the “core in docker” and “core in a venv” together.
But I don’t think that’s a show-stopping error. The point still is being made. The supervised install is still a favorite over both of the other two choices combined by a two to one margin. I am even surprised by those results so far.
I’ve already considered that as well and had been considering starting a dedicated thread to have that poll included outside the context of this thread.
It would obviously split out the two install versions as noted above.
I think I’d need some kind of flow chart and explanation of the choices tho, since the user might not have any idea how they installed HA by just listing the “official” names since those make no real sense by most opinions.
At least it’s an attempt to gather more information than has been attempted by any official means. At least that I’ve seen.
So are you saying that trying find out any information at all is a waste of time because the poll isn’t comprehensive?
Do you have any other ideas on how we can get even some idea where this stands?
I’d assume the devs are working at trying to find out officially, if you’ve noticed this log output:
2020-05-15 10:13:55 INFO (MainThread) [homeassistant.components.updater] Submitted analytics to Home Assistant servers. Information submitted includes {}
Maybe so.
I don’t have the updater integration running since it never seemed to reliably work for me.
But that’s good to hear if so.
I think this would be beneficial, but it needs to be stickied to the top so everyone who visits the forums sees it, even when signed in.
Perhaps one of the mods who shall remain un-tagged could help out with this? The poll already created in this thread gives an interesting take on what a larger vote count could capture.
I would make one small change to this, you may not agree
Good idea as the poll is buried in the middle of this one now…
I suspected 40% for Supervised installed - but that was just raw rank speculation by me…