Two weeks in

[zarthan]

Some comments and observations

I think I made a pretty smooth transition from HomeAssistant to HE. I had all my devices moved to HE fairly quickly. After a quick test with a couple of devices, I made the leap in a couple of days. HA has a very large list of supported stuff and I will need to find some way to work with some of those things. They aren't critical and I won't be disappointed if I don't have some of those things controlled. Automations and such took a few more days to get to a decent minimum of functionality. More to do but I am pleased with what I have achieved so far.

I ran HomeAssistant as a virtual machine on ESXi. I have used ESXi for 15 years or so and it just seemed criminal to not use it. My HA instance had lots of CPU and RAM but one thing that was instantly clear was how much Zwave performance improved on this tiny HE device. I have some Aeon Minimotes that control some Hue lights. With HomeAssistant, a single button push would toggle on a light but a second push might take a few seconds to register and toggle the light. With HE I can push the remote button as fast as I want to the point where the light is flashing. I thought the improvement moving from HA on a Pi 3B to ESXi was huge but the move to HE was just wow. So I am pleased as can be with the overall performance of this little device. Hats off to the HE bunch.

If I can move back to Rules and such, I would like to comment on something I find difficult. I feel that Rule creation is like I am on a journey with no map. I only see what's around the corner after I have turned the corner. Once I turn the corner, I can't look back to see where I was. For a new user, this can be both frustrating and it tends to make learning a lot harder. I don't know whether I might have an easier time had I had previous exposure to Smarthings or was a programmer but I find it hard to visualize the rules. I have spent 25 years at the Linux command line. I can eventually craft a decent bash script so I am not totally inept.

I would love to see a rule fully exploded to reveal the structure and options. Screenshots don't really reveal much for me. It's somewhat like looking at random snapshots of someone's vacation. Without the narrative, they are just random images. An exploded, structured view of a rule would allow someone to post the rule they are having trouble with, and have someone comment like:

Select Actions for False
    Delay by 2 minutes  <--This should be 1 minute
    Off: sleep_mode_switch

It would also allow someone to post an easily understandable recipe for some novel way they worked around some issue.

Like I said, had I come from ST I might not care, but I am new. I think the long term success of this platform will be the ease of transferring understanding from one generation of a user to the next. At some point, users with the most experience will drift away. Their systems will be working smoothly and they won't need to learn new things and will stop participating.
Thanks

[erktrek]

Thank you for sharing your experiences! I came over from Smart Things last September. I was using WebCoRE at the time and I will say the switch to RM has been a bit challenging - and I am a programmer. I prefer to see the logic like you do. RM definitely has it's own style to it. I will say though that you can accomplish most anything you need to. The only mild complaint is you can end up creating lots of rules to handle complex interactions.

Another idea is to start writing your own custom apps. That way you can see and structure the code the way you want.

On a side note - I did try and install WebCoRE for HE. Had a bunch of hub crashes etc. I would recommend holding off until you get more familiar with HE BUT follow the WC threads here. There are some excellent people working on this.

[csteele]

@zarthan

It's somewhat like looking at random snapshots of someone's vacation. Without the narrative, they are just random images.

Ok, that's just too funny!! Thanks for that

On the other hand, it's all we've got. AND you will not get a lot of argument that it's a poor way to guide/share - there are so many layers, and need a screenshot per layer to really guide a new user.

Beyond that, there's often three ways to do the same thing.

Want to turn on a light by Motion and then have it turn off 5 mins after the last motion?? First is obviously: use Motion lighting, but RM has two ways: Actions for False, delay for 5 mins [cancel] + Light Off and then an all-in-on of Pending Off:

All of that to say.. I get it... yes, wish it was easier to explain. However, clicks win in ease of building, so the web UI is difficult sometimes just because it has layers to walk you thru the creation.

By the way... did you read Bruce say a new version of RuleMachine was in the works?? Maybe it will be an easier explain.

[erktrek]

Huh, did not know about "pending off".. nice to know - thanks.

[csteele]

Don't get confused though.. there's only THREE to choose from.

(ignore switches per mode, not a part of this.)

you can choose to just turn it off after a delay, which is exactly the same as a delay explicitly in the Action on False. The one I now use most often is the 2nd one... Pending cancellation.. .NOT to be confused with the final one, per MODE Pending cancellation

[zarthan]

I wouldn't necessarily care if the interface was the same as it is now. I would just like the option to view / copy what I have created as an exploded structured view of it. I wouldn't even mind if I just had an online view of the rule structure so that I can get it organized in my head or on paper. There are times you just can't beat paper and pencil. With an online copy it would be easy enough to copy it off, make some choices and then post those choices back here for comment. I think everyone could get some benefit from that kind of discussion.

[zarthan]

And Thanks.