Hello. I still using OpenHAB 1.8 and I’m happy so far.
I mainly using Homematic but also have the MQTT controlled Raspberry PI, an airing and an heating systems.
I still have many ‘to do’ on my OpenHAB-to do list. And before I go further, I like to ask you:
“Stay with OpenHAB 1.8 or migrate to 2.x?”
I have little fear about this migration: what are the effort and the risks?
It is worth at all? What are the benefits?
Thank you in advanced for your suggestions.
I too put back my migration to 2.0 for a while.
Make sure that you save your configuration folder
And the main .cfg file - The one with all the bindings configutations - Can’t remember where that one was stored.
Then follow:
It is straight forward. As an OH 1.8 user you will find it easy.
Familiarise yourself with the new concept of things and channels
Back up your 1.8 machine and have a go. I hesitated but I am not looking back
Held me back from 1.8 -> 2 conversion : lack of login security. OH1 has built-in simple passwording for UI access, which I relied on. OH2 has nothing like that (still).
Security can be added on to OH2 with say a reverse proxy setup, but I didn’t want another separate service to maintain.
Eventually I realized I was a bit naive relying on the simple password anyway, so set up the necessary NGINX and accepted the extra chore.
In the meantime, waiting had allowed new OH2 features to arrive. The most significant for me was addition of “member of” Group triggering features to Rules.
My old rules had a mess of lambda functions for code sharing. Having a proper rules re-write instead of just a review allowed me to really tidy up and eliminate most lambdas.
Biting the bullet and rewriting in this way also caused me to write in new error management rules of my own to exploit new OH2 Thing/channels info e.g. handle external device failure.
All that took some time, but it is possible to have existing OH1 and developing OH2 on same host and swap for a bit of work when time permits.
In reality my old 1.8 setup did what was wanted reliably. So does the OH2 replacement - but it’s now easier to maintain i.e. add new features or hardware.
EDIT postscript - I opted not to try using PaperUI for Items and Things management, and stuck to text files. Easier to manage for large numbers of similar Items.
Nor did I use experimental rules engine, which doesn’t yet look capable of much of what I wanted.
If you are using a Raspberry Pi, i suggest to buy an additional SD card, clone it and give it a try.
You can clone the card with win32diskimager.
Using a clone, you can go back by inserting the old card again.
This was discussed on a number of occasions, see e.g. this thread.
The summary is to start migrating asap but don’t hurry in doing so. Follow the migration tutorial.
I’d recommend migrating.
And to provide just a little bit of elaboration on Markus’ recommendation. If you follow the Migration tutorial, when you get about half way through you should have your existing OH 1.8 configuration using all the same bindings and all your current configs with only the minimal number of changes to your configs and Rules necessary to run. At that point you will be running with your existing OH 1.8 configuration running on an OH 2.x core.
At this point you can start to upgrade to 2.x versions of the bindings you are using or to take advantage of new features in OH 2 like the Member of trigger rossko57 mentions, at your leisure. When you do start to take advantage of the new bindings and features, do so one at at time.
I managed to upgrade over a couple of afternoons of effort, keeping in mind that I was going very slowly and taking lots of notes because I was also writing the Migration Tutorial at the same time. I then managed to migrate to the 2.x version of the bindings, where 2.x versions were available, over an additional couple of afternoons. I probably spent about 8 hours total, again, slowed down because I was writing as I went.
I would say absolutely. Automatic discovery of devices and the Member of/triggeringItem alone are worth the upgrade. HABPanel is awesome! And pretty much all new development is being done on OH 2.
Finally, Java 8 is no longer receiving updates, I believe, and OH 1.8 will not run on any later Java. There is current ongoing work to move OH 2.x to newer Java (11 is the target I think). OH 1.8 will not be similarly updated.