Hubitat/RM Future

Well, OK.
It looks like we have very different way of thinking and vision on how HA should be implemented.
Nothing is wrong with this, everyone's vision is very different for sure.

For all current hubs this is 100% true.
But could be not a requirement if things are implemented a bit different.

Here is one quick example on what I meant by "simple and intuitive HA".
Jasco Z-Wave Motion Dimmer is only heir away from being really Smart Dimmer/MS.
It is missing built-in web server with somewhat minimal built-in Application
say, similar to "Motion Lighting" application in this case.
Yes, I am EE but with some programming skills.
I myself experimented with "lwip" (stand alone light way ethernet implementation
which does not require any OS) and run very successfully few examples for
controlling LEDs and/or Relays. These examples worked right away without any problems.
But since I am not experienced SW designer I had a difficulties how to setup
real SW development environment and to apply more complex modifications
to these examples.
LWIP can run over any Packet Communication Protocol including WiFi, Bluetooth,
Zigbee, ZWave, Wired Ethernet and whatever else as soon as it is "packet communication".
Memory requirement and CPU speed is minimal. Even 8-bit PIC potentially can run it.
Since function is limited to only one very simple application and does not
require any OS it could be very simple and quick implementation for the experienced
SW engineer.
Bottom line - this is really doable because I already tried this.

I guess, the Shelly Devices already running built-in web server.
So, these devices are even closer for implementing built-in applications.

Also the Devices should have an ability to talk to each other directly
bypassing Central Hub. Insteon kind of took this approach.
The Insteon Network did not require Central Hub but Insteon Devices did not have
any built-in applications either. So, for any real HA Insteon required Central Hub.

Another benefits for all this distributed computing power will be
significant reliability improvements.
Central HUB (no doubts here) is a Single Point of Failure.
HUB is down and nothing works.
That is why my wife requested that every light and appliance absolutely
must have local control. This means no hidden inline dimmers/switches with
nice looking remotes mimicking real switches on the walls. Plus every switch paddle
and/or remote button must have only one assigned function.
So, my HA done exactly this way but it does not mean I did not hide some functions
behind multi-function paddles and buttons.
But I am not against Central Hub.
Central Hub could/should be used for implementing very complex automation
and for managing all devices from one place.
My vision for Central Hub is something like Local Cloud Computing.

I guess, the above should be better explanation on my vision for HA implementation.
But again, all the above is nothing more than my "loud thinking",

OK, what's stopping you from pursuing your vision? We have pursued ours, you should too.

6 Likes

I am good as EE (at least all my bosses think so) and can design good HW.
I have good understanding in SW and can do some programming.
For instance, writing test routines and drivers for the HW I designed.
So, I can do a lot of different engineering job.
But I have no idea how to run business and at this time have no any
desire to start and run one.
However I am OK to share all my ideas with anyone who might be interested.

Elementary schools should be teaching programming-style logic based on pseudocode/English (eg IF homework done THEN play ELSE do homework) and then a real programming language in high school. I believe the vast majority of people are capable of dealing with it if taught properly and encouraged. Just like with dyslexia there will be a minority who cannot, and like with dyslexia we will learn to be understanding of the problem.

2 Likes

Just wanted to say 'thanks' to those who have contributed to this thread with the informative/inciteful replies. Interesting reading.

/ lack of input =p

I'm not so sure i would agree. There is definitely a learning curve starting out, once that looks far steeper than it really is. I am probably still be considered beginner, or maybe slightly more advanced. I also started with Smart Things, and switched when they dropped support for Windows phone and started the transition to the new app. I have been on Hubitat for about 3 or 4 years now. I can definitely say that Hubitat has lead me to try and do far more than I was ever inclined to do with ST. There I was pretty much of the mind set of set it and forget it. With Hubitat feel more of an urge to tinker, figure out new ways i can use automation , or improve what I already have.

Well you say you too started with SmartThings. Can you be sure you would not have found Hubitat's complexity and power a bit daunting had you not? So many different menus it takes a long time just to remember where you saw what! I am like that with Blender, dread opening it. If there had been a Blender Beginners Mode with far fewer options visible I might have been happily using the advanced features by now instead of giving up on meshing.

I started with wi-fi smart plugs then graduated to an Alexa/Zigbee hub when I wanted to add smart lighting into our home.
I also wanted to move away from wi-fi smart devices as much as possible.

While the 4th generation Echo with built-in Zigbee hub did work for most of my devices, Alexa routines are not reliable.
Mine would just stop working at random.

That's when I looked around for better options and to get an idea from the various communities about the products and their issues.

That's what pushed me towards Hubitat.

I only have one RM 5 routine at the moment and it works/doesn't work.

The doesn't work part of that statement comes from me assuming something worked the way I expected it to vs how it actually works.
As a long time software developer, I didn't read the documentation :rofl:

As a long time hardware developer usually I am not reading docs either.
Furthermore, if I cannot do simple tasks intuitively I am immediately ruling
out this SW/HW products. But I am OK to read docs for whatever features
which are not obvious.
As a EE FPGA designer I am using VHDL/Verilog (these are Hardware
Description languages). Both projects initially came from DOD requirements
to eliminate extra layers of documentation. I.e. the VHDL/Verilog design
must be self explanatory. And it is self explanatory if it is done right.

Returning back to Hubitat.
Visualized RM Rule is very intuitive and self explanatory (at least to my eyes).
However GUI for creating rules is not intuitive at all.
When I started to create Rules in RM I new exactly what I was looking for.
But it took me a while to realize all GUI is a dynamic pages and you are not seeing
the entire picture right away.
One improovent to the RM GUI will be the ability to see all available options but
whatever cannot be selected should be grayed instead of being invisible.
I am not sure if this is possible to implement on current Hubitat platform.
And if this is possible what efort it will/may take.

1 Like

I hesitate to say this for the retribution that may come...but many will invest the time to start climbing the learning curve of a lot of things as long as they know it will lead to building on itself, and resulting in implementations that won't require, or tempt, endless reworking as the platform/tools evolve every 6 months.

It has been said many times that "you can always keep using what you built" but nobody likes being left behind as significant improvements are made. There is enough mentioned in this forum of the likes of *being stranded" in an old version of RM (because folk don't have the time & energy to port to the latest version) that I bet there are a lot of folk who look long and hard for a HE supported path to do as much as possible without going the roll-your-own route.

Call that beginner, novice, or perhaps... just dis-interest to get knee deep in this.

While it is totally satisfying to configure home/facility monitoring and automation, not everyone wants to make a hobby of maintaining solutions to keep them in sync with the evolving platform. Yet they don't like being left behind either.

Yes, there is a point where it becomes necessary to roll up your sleeves....back in the day I can remember major re-writes of systems. But those usually happened after YEARS of evolution of the platform they were resident on; and they were for systems that actually made money and paid for themselves.

3 Likes

Every userā€™s needs will be different. I have setup some HA at both my father and my motherā€™s homes.

At my fatherā€™s place, I setup a few devices and related rules, and only need to touch it again if he wants to add a device. This occurs maybe once a year. The rules are written in whatever engin was available then and never get touched again. New rules are typically written in the new engins, but thatā€™s really not a necessity - just my personal preference.

At my motherā€™s place, the bits of automation have been setup and never touched again. She is happy and all is good.

At my place, it is a totally different story. I enjoy tinkering with my setup to add this and that improvement, updating to the latest version of hardware at times, keeping old stuff at other. It depends on how frequently I need to use it, and on whether or not my current/future needs are met.

That said, I donā€™t feel the need to rewrite rules that work in the latest rule engines. Iā€™ll use the new engines for new rules, but those that work well stay in the technology they were written in. If I have time that I feel like spending on my HA and I want to update a few things, I do. The beauty of Hubitat is that I donā€™t need to. I can keep the same old platform if I want to, generally donā€™t need to update code or rules. Works great for my purposes!

3 Likes

This is my personal approach as well. I don't generally have time to go rework things that work just fine as is. Most rules I create these days are for development and testing purposes. Every once in a while I will revisit an old rule and redo it. I'd guess that on average the automations in use in my house are 3 years old.

Some people have a compulsion about updating old automations to new rules, and then complain about the "need" to do that. I don't find that a very compelling criticism of the platform and how we evolve it.

8 Likes

And my golden rule is : Don't touch and fix whatever is not broken".
Works very well for me and saves a lot of time.

1 Like

I'm a bit of a categorizer and have found myself using the different rules apps to separate rules that are for different types of purposes. :smiley:

I am very well organized/categorized.
Unfortunately Hubitat as is today does not offer any kind of folder structure for rules.
But for years I am/was using very structural naming convention for designing electronics.
To begin with - every name is prefixed.
So, I am using the same approach for naming everything in Hubitat (devices, variables,
rules, etc.).
Here is a short example for my rule names:

I wouldn't be happy categorising by room other than plugs and bulbs etc. For example the pills reminder I'd want to categorise under health or reminders. I have a lot of intruder detection rules that are not really associated with rooms anyway. So I prefaced them with s_ but kept forgetting what it meant haha.

Yes, I 100% agree with you.
But currently there is no way to create nice top level hierarchy.

That would be good cos some of my rules are just supporting rules to the main one

I think so, With smart things I did 99% of what I did solely through the app. While I had very minimal exposure to the IDE, When I did get into it, I was typically very lost and confused. In that sense I think Hubitat is somewhat easier to "ease into". While there is certainly a learning curve, I think Hubitat has a way of drawing you in and the curve felt a bit more natural and intuitive.

I can't argue with this, though again in a fairly short time I seemed to just "know" where to go. Again I think Hubitat tends to be more intuitive, especially once you learn and understand some basic terms and ideas like driver (vs "device handlers" in ST),

Now, with all of that having been said the one area when I still have issues in Hubitat is in Rule Manager. Sure I can get most of the basic rules I want to do done without too much effort, but if they start to get mildly complex I start getting lost in the weeds and find myself doing a lot of trial and error. It is that same trial and error that usually leads me to some other new adventure! Rule Manager has been steadily improving since I first came to the platform,

1 Like

I have never set out to rewrite rules when a new version came out unless it was because I had a rule that wasn't working the way I wanted to begin with, As a result over the years I have had a more natural progression to update most if not all of my rules at one point or another. The last major rewrite was last summer when we moved and some devices were left in the old place, some were moved to different rooms or replaced all together. In any case the old rules didn't work in the new house and needed to be re-done. Some were redone out side of Rule Manager in simple lighting etc, because they just didn't necessitate the the complexities of rule manager. In the near future there will be another revision as I am working on generally cleaning up and streamlining my system. I have a set up with multiple hubs, one for devices, one for rules, apps, etc. During a recent transition wound up having some rules on the devices hub. Now most of those are paused and the other hub is controlling the automation. Additionally in the next year or so I expect to expand my system, which itself will drive new rules as well as some rewrites.

Yep, that was me for a long time. If it wasn't built in or official I didn't have any interest in it. As things progress , I'm finding I'm looking into more options (both in terms of devices as well as apps, etc.). It's hard form me to wrap my mind around how one could just have a set it and forget it mind set with this, I've tried that, but it seems like I just keep thinking of new ways I could make it work, but that's probably just me.