I am new to Hubitat, but a long-time Smartthings user. I have a dozen or so smartapps that need to be ported over from Smartthings.
- Is it better to try and port these over or start from scratch with Rule Machine rules?
- From a performance perspective, would it better to have multiple Rule Machine rules vs. multiple smart apps?
If rule machine can work you'd probably be better off using it, if the app makes API calls to a cloud service you probably don't have much choice but to try to port the app if it isn't already.
It depends on the smartapps themselves. There is a general consensus here that Groovy code (if written well) will outperform Rule Machine apps because RM has a lot of overhead that comes with it. I myself have a ton of small, custom apps that I have written to do things that I just don't need RM rules for and they perform really well.
On the other hand, the RM rules that I have also perform really well. So, it's kind of a toss-up depending on the apps in question.
The smartapps don't have any API calls to a cloud service. Many of them are blingy things such as custom faders and logic that reacts to external stimulus (someone turning a switch off manually) vs. actions performed by the automation. Groovy is a lot more expressive than RM rules.
I also have cast-web-api running on a Raspberry Pi. I need to find the client-side equivalent for this setup in the Hubitat world. I have seen some threads on this topic; I will explore this once I have my basic setup done.
Thanks for this. My current thinking is to do what you have done - use the current smartapps and later convert the real simple ones to RM rules.
Welcome to the Hubitat Community!
You’re going to find that many of the simple things on ST that required either a smart app or webCoRE, can be easily accomplished with Rule Machine 4 on HE with no noticeable effect on performance.
Custom apps on the other had can bring the hub to its knees if there are uncaught errors, incompatible code, or mistakes in the original code that the ST cloud instance would not have issue with because of the computing power behind it.
If these are not critical apps, I would suggest you make your early experience as good as possible by leaving them out. Ask the community for help with RM4 if you need it. Better to learn the system first, and customize later. There are also many built in app's that can get you up and running quickly, without a significant time investment.
I run Cast-Web API. There’s already a client here.
Thank you for your response.
With regard to the cast-web-api client, that thread points to a github link that does not exist anymore?
That's not available any more?
Boooo! Well then, it was ported once, it can be ported again.
I was having some perfomance issues and hub slow downs. I had a lot of RM rules. I started the process of writing small custom apps to take the place of the rules. Which is what I had over in ST.
When I got all my rules into my own apps. I disabled RM. I wanted to see what the difference would be in performance.
I have NOT experienced any hub slow downs since then. Performance is as good if not better.
So in my case the apps are better then RM. But that may not be the case for everyone.
I guess the complexity of them might make a difference. Most of my rules are very simple, and some work in conjunction with each other. I have 99 rules at this point and don’t have slowdowns I’ve ever been able to attribute to them. Custom code and network (IP and Zigbee) is where I have found issues in the past.
Just built my 100th Rule Machine rule!
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.