Fix OSRAM Classic A60 RGBW crash

[manup]

The light does crash when ZCL attribute reporting is configured and the binding table is queried. Likely resulting in a watchdog reboot after which the lights turns on in full brightness.

The fix here is to use polling instead of attribute reporting as well as to disable query of the binding table.

Related #7546

[ebaauw]

Is the crash caused by (one time) creating the bindings and reporting configuration to by (multiple times) querying these? In other words: does the crash occur since the introduction of the DDF and final state machine actively checking the binding and reporting? In that case, wouldn’t it be better to do a passive check (see if reports arrive), as is done for the polling as safety net?

[manup]

The creation of the bindings and also the reporting does work, in fact my test light still sends the reports without crashing.
As I understand the sniffer logs the querying of the binding table makes trouble, and also the verification of the ZCL reporting configuration lets the light crash after a while.

With the addition { "supportsMgmtBind": false } to the DDF the binding table queries are suppressed, and the state machine works like the legacy passive checking for reports (which it always does).

However the verification of the ZCL reporting configuration is still done to check that the values are indeed the ones specified in the DDF. This is usually only done once after startup and in case the passive checking detects that something isn't setup. Normally a nice feature that also allows to update a DDF with new saner values if needed.
This can't be deactivated, and currently I don't plan to do so as it would complicate the code just to workaround a few broken devices (so far only Osram and here only a selection of them).

I think using polling here is a ok'ish tradeoff for now. I don't think the old E27 and GU10 lights will be fixed by Osram but current lights on the market, if they show the problems too, should be reported to Osram so they can work on a bug fix firmware.

[Finne75]

I think this will solve a long-standing mistery of lights switching on unsolicitadely :) - apologies for non-technical comment