Update README.md to Z-Stack 3 firmware based on 4.40.00.44 SDK as stable

[Hedda]

Update README.md stating that Z-Stack 3 firmware based on 4.40.00.44 SDK is "STABLE" as per https://github.com/zigpy/zigpy-znp/issues/62 as well as https://github.com/zigpy/zigpy/discussions/655

[codecov-io]

Codecov Report

Merging #64 (e1c0f6e) into dev (1ee9c49) will not change coverage. The diff coverage is n/a.

@@           Coverage Diff           @@
##              dev      #64   +/-   ##
=======================================
  Coverage   98.80%   98.80%           
=======================================
  Files          37       37           
  Lines        3255     3255           
=======================================
  Hits         3216     3216           
  Misses         39       39           

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 1ee9c49...e1c0f6e. Read the comment docs.

[MattWestb]

Sorry to saying its can NOT being stable in 24 hours !!

It need being verified on the device by more users for longer time before getting status "stable".

Tasmota was waiting 2 weeks with EZSP 6.7.8.0 that is very short time but its up to the devs.

Beta, Prerelease, release candidate and pending is the right if not being one very serious security or stability bug fix.

I like @puddly and @Adminiuga thinking about it before changing status if it.

[Hedda]

@MattWestb FYI, zigpy-znp developer puddly posted in https://github.com/zigpy/zigpy-znp/issues/62 recommending upgrade to this firmware on all CC2652 and CC1352

Yes, it is very much recommended. I have been reliably able to trigger Z-Stack crashes on older firmwares in about 3-4 hours (by sending requests as fast as possible) that are impossible to reset remotely on both the LAUNCHXL boards and on the ZZH. Slaesh's stick should in theory be able to recover from any lockup with the unreleased reconnect watchdog in the dev branch. New firmware has been running for over 36 hours now without issues.

Other people replied to https://www.reddit.com/r/homeassistant/comments/l9uhvm/psa_zigbee_network_lockup_bug_affecting

[MattWestb]

Its up to hi than the problems is coming like the first serious one https://github.com/zigpy/zigpy-znp/issues/63 Its not professional flagging it stable and roiling it out to all users but its up to the dev team and in the end its can ending with being big problems with the AH team !!

[Hedda]

By the way, Koenkk has also moved the updated firmware image files from the "develop" tree to the "master" (mainline) tree:

https://github.com/Koenkk/Z-Stack-firmware/tree/master/coordinator/Z-Stack_3.x.0/bin

[puddly]

Its not professional flagging it stable and roiling it out to all users but its up to the dev team and in the end its can ending with being big problems with the AH team !!

That user-reported issue was due to them flashing the router firmware by accident.

Other than completely broken firmware, I can't think of anything much worse than weekly lockups that require physically unplugging the coordinator due to hardware reset limitations (no watchdog in Z-Stack). I was stress testing that specific version by sending about 5-10 requests per second for over two days straight, which is what triggered the original UART leak bug in only a few hours. Very little changes between Z-Stack versions but I have had no issues yet so while I wouldn't guarantee that it's the most stable release ever, I would say that it is much better than the previous version :)

[Hedda]

Note! Highly recommend that people backup NVRAM of Texas Instruments adapters with zigpy-znp before upgrading firmware:

https://github.com/zigpy/zigpy-znp#nvram

Tip! If done a backup of NVRAM before then can do a restore of NVRAM after if have any issues or need to downgrade later.