I’ve been wanting to migrate from OZW 1.4 Depreciated to Z-Wave JS pretty much since it was announced but wanted to let the smoke settle for a while before pulling the trigger. One of the things I was waiting for was the migration tool that was planned. Is there an update on the possible release of the migration tool?
After spending the last couple of hours reading through the mountains of posts regarding Z-Wave JS and challenges people are encountering, it seems like ZWJS is still a little rough around the edges for a few things that relate directly to my setup:
Battery operated entities may not get automatically “cataloged” in ZWJS and stay as Unknown (I have a lot of these and they are fairly crucial to my automated home)
Problems with older GE/Jasco switches (I only have two, but they seem to be problematic in ZWJS)
Issues getting Wallmotes to work properly
User interface lacking and people having to use the MQTT or ZWAVEJSM (whatever that is, I saw it a handful of times to allow you to do more stuff with ZWJS) to be able to program custom parameters or otherwise tweak the Z-Wave device as many other home automation software platforms allow you to do
Great support for a wide array of devices, but may still require sending your device info in to get it supported (sort of what I hated about Indigo, having to wait for devs to officially support a device)
I’m super excited to dive in (particularly the greatly improved response time of Z-Wave) but it seems like it’s still a bit bleeding edge.
My OZW 1.4 still works pretty great so I’m happy to stick on that if the overall feeling is that ZWJS is “almost there”, but if the feeling is that ZWJS is pretty much wrapped up and ready to go with future enhancements planned but not needed for now then perhaps I’ll take the leap (although renaming 200+ devices is not on my list of favorite activities to spend my weekend on).
I’m in the same boat. + Especially for this I stopped adding Z-Wave devices to the existing setup.
And that already delayed my next big project, as I’m planning to fully automate my whirlpool with a Z-Wave controller. I’m just too afraid, that i get it working now and it broke after the update, and then i will be time pressurized to get it working again. So i decided to stay manual until the Z-Wave story clears out. Looks like this summer there will be no automated bath…
I’m still constantly adding new Z-Wave devices to my OZW 1.4, I don’t see any reason to postpone anything for the sake of ZWJS personally. Worst case scenario is maybe I have an extra 10-15 devices to migrate and in the grand scheme of things that’s nothing.
I also know I could do a full backup of HA and go back to that if things go south, but that will still require a significant time investment to sandbox ZWJS.
You should automated your tub, what’s the harm in doing it now and possibly redoing it later?
Just afraid, that the support stops suddenly and i have to migrate middle of the season, not knowing how and so sitting here with a broken bath. That. I’m just a beginner and those migrations are making me mad. Just learned, how something works and it changes… I need a little more assurance before breaking stuff. But i know, where i started with, i have more open source experiences like this.
Like the migration of OMV from version 4 to version 5, where i made a mistake and lost 4 years of my child’s photo gallery.
Since then I’m not that fast with trying things, i don’t understand fully…
I think ZWJS is here to stay for a while. If you understand the history of Z-Wave integration with HA you would probably feel a bit better about that statement. Things went from the hands of contributors to the hands of the main dev team in ZWJS and it’s widely used on other systems too.
I have no confidence issues with ZWJS honestly, just don’t want to make the switch until I’m sure it’s solid.
To my knowledge no one has even started on one yet. My two cents is that it’s unlikely one gets made and not anytime soon. It’s a one time task that doesn’t take that long. The largest networks take an hour or two. Compare that to the effort it would take to make one (which is much harder than it seems) and a huge number of people have already switched. And you’d still have to fix a bunch of things no doubt due to device types changing. And all of the people that could make one are busy working on the integration itself. I’m certainly not wasting my time on one.
Zwave-is is very stable. Just switch and rename. If you run the script in the migration guide to pull all of the names into a report it takes about 30 seconds per device. Or a max of around 100 minutes for a 200 device network.
OZW dying was an anomaly. You had own dev working on it, who had been doing it for like a decade, and when he burned out it was written in an idiosyncratic way in outmoded languages (c++ and qt). It was also an extremely messy codebase. Few of the people who would have been willing to help could even understand the code.
Zwave-js is written in JavaScript/typescript and has many contributors.
I understand all of of this. The history, the need for change. Yes, it’s clear as the sky. That does not change anyhow, that I’m standing now on a point, where I’m afraid to touch the working stuff and break it, because I just don’t know, how to migrate it seamless…
What are you worried about breaking? Zwavejs is pretty stable and feature rich. At this point we’re passed ozw on command classes supported and such. And the UI is more usable if you use zwavejs2mqtt, which I recommend.
Also to the extent a bug appears, our average time to close them is 2-3 days.
I’m not afraid of Z-Wave JS. I’m afraid of myself knocking the stuff to the ground because of lack of experience.
I’ve Learned about HA just a year a go. In one year i’ve learned to understand Linux, Samba, YAML, Z-Wave, OMV, and who knows what. Everything just enough to get it running.
With my grade of experience this is just a big step and I still don’t believe enough in myself getting it right. I wish, i was a programming guru, but I’m not…
Well, I’m running a house with this stuff, and breaking it is not really, what i want. Like in every room the light switches are Z-Wave. They just have to work.
So I need one of those two things:
A step by step guide to migrate in a way i understand
A seamless migration tool doing the dirty laundry for me
As long, as I do not find the tools to take the step in a way, i fully can trust (even if its trust in my own steps), I’m stuck.
So, i finally managed it. Half day of work, but successful. It was not the migration it self, that was a pain, but re-configuring automation’s of my z-wave devices after the migration.
Luckily i found a blueprint for my 4way button’s, and then I decided to migrate. I still wish, for the less skilled users between us, there will be a migrating tool, as announced time a go… That will be fair enough
