Just my personal opinion, but I don’t care much about voice. It’s more of a gimmick than a feature to me. And even though I am not a native English speaker, I don’t care much about localization of voice commands either. I set my interface language to English on most devices if I have the chance. I appreciate that this is an important feature to others but I’d much rather see time and effort spent on other things.
1 Like
Local voice in HA will not be running perfect when 2023 runs out.
Local voice is a long term investment.
When HA started out the first year was most likely not with a perfect running release and features were missing, but over time, years that is, it have become a great product.
Local voice is a feature that you can not really get with other modules, so it makes sense to invest in it.
If resources are an issue, then abandoning the media features would make more sense, since there have for a long time been several alternatives and also several local open source solutions.
Likewise the internal scripts/automation feature could be scrapped and instead NodeRed could be used, since it is local, open source and more capable.
There are probably other areas too. The essential thimg here is that there are other stuff that could be scrapped, but there are probably users for them, even though I am not one of them.
Local voice is something I look forward to. Google and Alexa is data mining a lot and Siri require expensive and a fenced off ecosystem based on Apple products.
1 Like
Oh no, thanks 
6 Likes
Your kidding right?
April Fools Day is several months away but you can try to use that one again at that time.
While we’re at it why don’t we just gut everything that makes HA what it is so that you can have voice recognition.
Really?
4 Likes
I am not trying to gut everything.
I want the voice assistant, because I do not trust the cloud solutions.
And I accept that other features have no value for me, but they might have for others.
The post was meant to have others look critically at other parts of HA with the same criterias they use to disapprove the Nabu Casa’s decission to bet on voice recognition.
You suggested to remove script and automation functionality from a home AUTOMATION project as “bloat”.
If that doesn’t totally gut HA then I have no idea what the concept of “gutting” would even mean.
8 Likes
I searched for “Voice assistant” in Feature Requests and the most recent WTH. Here are the results (feel free to use different search terms):
https://community.home-assistant.io/search?q=Voice%20assistant%20%23feature-requests
https://community.home-assistant.io/search?q=Voice%20assistant%20%23what-the-heck
What the two have in common is that the members of this community forum haven’t been requesting a native voice assistant. That’s not to say its addition wouldn’t be appreciated only that, historically, there hasn’t been a demand for it (in contrast to many requests for other things).
The top FRs focus on the organization, calculation, and presentation of data. Given that the focus of 2023 will be on something else may explain why its announcement has been received with some indifference.
9 Likes
I agree. Looking critically at HA as a media player, it’s hard for me to justify. Just like voice, I think it’s great that interested developers volunteer to work on it, but it’s hardly a core function of the product. On the other hand, automations are a core capability. I chose HA for many reasons, including its automation capabilities. I don’t see how abandoning HA automations in favor of NodeRed makes any sense.
Stepping back to look critically at the big picture, Voice makes even less sense. Wouldn’t it be better to do this as a separate, open-source project? HA developers can choose to join that project, and make sure it integrates well with HA. A lot of non-HA developers would see this as a worthy challenge and join in, too. We would get more developers, a more versatile product, and something HA users can choose to use. Or not.
And the HA team can focus on keeping the core parts of HA running, and maybe even attack more of those FRs and WTHs which their users actually asked for.
4 Likes
Rhasspy is/was a standalone project, but voice assistants is maybe a bit too big to have as a hobby project and the number of developers that can actually contribute to such a project is really limited.
Mike Hansen started Rhasspy but was then snatched up by Mycroft and the project went on pause, since hardly no one else have the knowledge and also the passion to put energy into an open source project.
Neavy Casa got a chance to hire Mike Hansen and jumped on it. If they had not, then Mike would be working somewhere else now and Rhasspy would probably be dead/on pause still.
And Nabu Casa would proba ly not have a chance to hire a developer with that knowledge in the years to come.
I still opt for a decent poll amongst the community to get a real picture for the interest in VR.
2 Likes
It would not say anything useful, because what really matter is the interest among the core developers.
The community is mostly just coming along for the ride. Of course some developers find joy in making something others want to use, but the core developers is the ones choosing the direction for this project and it is their project and their dream that counts.
If they
This touches probably on the basis of leading a community and taking it to the right direction. I believe we have more the 500k members and still growing. It requires a better governance structure then just what pops up with the developers. The WTH month and pull requests are a nice attempt to get input on what is living in the community but very granular with limited impact.
Have an anual well prepared referendum with a number of directions where resources will be put into. Well prepared means that nabu casa leaders address the main topics and ask the community what is important to them. Limited choices maybe but the ones which are moving the needles and the big bets for the future of HA.
In this case VR could be one of them. Same for automation, AI and machine learning.
At the end there is a members base which rely on the decision making of the devs. And vision should outperform well meant hobbyisme.
2 Likes
Not sure where you take those numbers from. On github there is a total of ~3000 contributors to the HA code. Most likely many of them with just a one-off contribution. Not sure how many are real active ones. Even the top 100 of them are not continuously contributing.
Coming back to the 500k+ users (which probably is what you are referring to). Vast majority has never even opened an issue on github, most can’t even read the docs and ask simplest questions on fb. I don’t think that directions for this project should be democracy based this would be unmanageable.
Regarding the WTH, the highest number of votes got a request that is minor in my opinion and potentially not that easy to implement, with existing simple workarounds. I would not agree that just because it got highest number of votes it should be a number one priority. It won’t add significant value to the project and will consume devs’ time.
And a side note. As much as I’m totally non-Apple user I really appreciate the famous quote from Steve Jobs:
“Some people say, “Give the customers what they want.” But that’s not my approach. Our job is to figure out what they’re going to want before they do. I think Henry Ford once said, “If I’d asked customers what they wanted, they would have told me, ‘A faster horse!’” People don’t know what they want until you show it to them. That’s why I never rely on market research. Our task is to read things that are not yet on the page.”
1 Like
Henry Ford didn’t invent the car; it already existed. He applied assembly-line techniques and vertical integration to develop the mass production of cars so they can be more affordable for a broader market (the burgeoning middle class). In other words, he provided what was really being asked for at the time namely “an affordable car”, not “a faster horse”.
However, the “faster horse” anecdote makes for a more pithy sound-bite that’s a better fit for Jobs’ management style (and his infamous “reality distortion field”).
As for “500K”, the closest we can get to a measure of installed systems is from Home Assistant’s analytics. It represents the number of installed systems based on those that have opted-in to report it. The actual number of installed systems is higher (because not everyone has chosen to share data). The number of users this represents may not be exactly one to one (because a single user may administer multiple systems). Nevertheless, a conservative estimate remains in the ‘hundreds of thousands’.
1 Like
With respect to the numbers of users /members read this:
But let the number not be the issue. It’s not the point.
I think as stated earlier a clear vision and direction is more important then hobby horses of developers ( and I am not trying to offend them as a lot of great work has been accomplished). And I also think that direction and vision can be build with the help of the community!! I am convinced of that.
You are giving the big number of users and votes on WTH as an argument for choosing something else as priority for development rather than the devs’ choice.
My point is that only a small fraction of users have even made an effort to vote on WTH ideas not to mention other contributions.
And Jobs didn’t say that Henry Ford had invented the car. He has made it available to the masses (relatively). Same way as nether Jobs nor Apple has invented a mobile phone. But he’s made it wanted by the people even though they have not asked for it in that form.
Same here. I don’t think that HA is going to invent anything in the VR area but there is a chance it will make it work in the home automation context
I am one of the people who heard that 2023 will focus on voice commands and thought “oh no, another focus that has very little benefit for me”.
And I think there is a reason why few people really care about voice commands, and it is not just because they are not local.
I would have loved something more practical and useful to all, like finally making the dashboard (including all the cards) work properly.
I think that the UI is the most important way of interacring with HA and it is the least reliable/stable/working.
Nevertheless, I wish all the devs the best of luck with getting voice commands up and running.
I will keep my fingers crossed for all of you 
P.S.: Does anybody know how the voice control is supposed to be triggered on iPhones and Android devices? If I am not mistaken, the companion apps will not be able to be triggered like Siri or Bixby (e.g. the dedicated button) with locked screen. As soon as you have to unlock and open the app, voice becomes less attractive than Android tiles (which work with locked screen).
1 Like
And I didn’t say that he did. Re-read what I wrote.
You supplied a Jobsian anecdote claiming the market doesn’t necessarily know what’s best for it (whereas some ‘visionary’ does). Unfortunately, it wasn’t the best choice of anecdote because Jobs’ comment was based on a distortion of Ford’s achievement. Ford’s market knew cars existed and wanted an affordable one, not a ‘faster version of a horse’; Ford gave his market what it wanted.
Home Assistant users also know what they want and it’s documented in Feature Requests (and WTH). A local voice assistant isn’t one of the high-demand features. My point is that it may explain why the community has been relatively indifferent to the announcement of dedicating 2023 to the development of this unsolicited feature.
2 Likes
I wonder why google with all their infinite resources hasn’t already done this.
I’m guessing this is pretty high on uses feature requests for Home Assistant although I don’t recall seeing it in month of WTH.
This guess is based on what exactly?
I have nothing against a voice assistant. I use Alexa right now but not extensively (an occasional kitchen timer and very rarely operate a switch in HA).
But I use it way more extensively thru the Alexa Media Player custom integration for HA triggered announcements. IMHO, that’s one of the most useful integrations in HA (aside from automations and scripts of course) either built-in or custom.
I want my HA itself to run automations and with very minimal interaction with it.
Sure, spend resources on voice assistants if it gets your juices going but please don’t mis-allocate resources away from maintaining (and actually fixing) things that are at the base of HA.
Unfortunately it seems that the things that the HA devs seem to be focusing on more and more are the “shiny and flashy new things” to bring in more and more (non-techy) users instead of spending time on the more mundane (but arguably WAY more important) job of fixing the things that are broken and have been broken for literally years.
4 Likes