Dealing with numerous Zigbee device quickly gets you to have to complexify your set-up by installing Z2M, simply because devices support is not at the same level in ZHA.
HA should natively work with the vast majority of devices.
Z2M implements those devices faster… why can’t ZHA do the same?
I have met several individuals who didn’t pursue their try of HA because of this complexity and lack of support of some devices.
On the forum, a fair share of reported issues are relate to Z2M installation, support and update. This is too complex for many potential users.
Probably because Z2M is a separate open-source software project with a large dedicated group of users and developers. In other words, a subset of its entire userbase employs Home Assistant.
It’s an open source project. Start contributing to add support for whatever devices are missing instead of complaining that it isn’t what you expect it to be.
Yes and no. Changes in zha first need to be made in zigpy, which the then may need a release cycle or two to get into ha. Z2m happens much faster.
That said, z2m and MQTT are easy to install and run.
This is exactly the problem I am reporting. I am a happy user of HA and try to contribute and help others, especially on the French community when I can.
My point is that if HA aspires at becoming a larger standard with larger installed base, we can not expect all users to participate to the development. Either HA aspires at becoming more mainstream and this would mean things should work better “out of the box” or HA remains a nice tool and game for geeks who can spend time and have skills to contribute.
This should get better with quirks v2. But we are just now seeing it socialized and people submitting quirks using the new format. The biggest issue is, we don’t have enough people who are willing to do the work to add quirks. Tons of issues get opened, but without the device, it’s not easy to help people, especially when they don’t know or are unwilling to find data points and test quirks.