LiXee ZLinky: fix polling interval + improve option descriptions

[sbraz]

• fix: LiXee ZLinky: make polling interval configurable again
  e30ace4 broke the polling interval configuration. Since then, it
  was always set to 600 seconds.

• fix: LiXee ZLinky: describe actual default polling interval
  The default description ("This device does not support reporting
  electric measurements so it is polled instead. The default poll interval
  is 60 seconds, set to -1 to disable.") mentioned a default of 60 seconds
  but it is actually 600.
  It was also misleading: not all attributes are polled, several of them
  have reporting configured and are automatically updated.

• fix: LiXee ZLinky: improve description of measurement_poll_chunk

[sbraz]

Should I use e.numeric("measurement_poll_interval", instead? I wasn't sure.

[Koenkk]

this is fine

[sbraz]

Ideally, I'd like to move measurement_poll_interval above measurement_poll_chunk but I couldn't find how to do it. Is that possible at all?

[Koenkk]

This should stay as is

[sbraz]

@Koenkk but it doesn't work as-is, the option is called measurement_poll_interval and the poll function checks ${args.key}_poll_interval, so the key must match the option name.

zigbee-herdsman-converters/src/lib/modernExtend.ts

Line 663 in f7323eb

const optionKey = args.optionKey ?? `${args.key}_poll_interval`;

Am I missing something? Before my fix commit, the poll interval was always 600 seconds, no matter the value of measurement_poll_interval.

[sbraz]

Ideally, I'd like to move measurement_poll_interval above measurement_poll_chunk but I couldn't find how to do it. Is that possible at all?

@Koenkk can you help me with that?

[sbraz]

@Koenkk I have another issue with this converter: I believe this never raises an exception on read errors, is that possible?

zigbee-herdsman-converters/src/devices/lixee.ts

Line 1918 in f7323eb

await endpoint.read(cluster, targ.slice(i, i + measurement_poll_chunk), {manufacturerCode: null}).catch((e) => {

I'm not very good with JS/TS but if I replace await endpoint.read(…).catch with a try/except block, I see read errors when measurement_poll_chunk is too high. I believe these silent failures were the reason my ZLinky was acting weird in the first place.
Can you help me find a proper fix for this? Do you want me to create a separate issue?

[Koenkk]

Ideally, I'd like to move measurement_poll_interval above measurement_poll_chunk but I couldn't find how to do it. Is that possible at all?

What do you mean with above?

I'm not very good with JS/TS but if I replace await endpoint.read(…).catch with a try/except block, I see read errors when measurement_poll_chunk is too high

using try/catch is fine here, please update the pr

[sbraz]

What do you mean with above?

I'd like to move the poll interval option above the poll chunk option in the UI:

using try/catch is fine here, please update the pr

OK I'll do that ASAP.

[Koenkk]

Currently it's not possible to define the order, so I would propose to leave this as is.

[sbraz]

using try/catch is fine here, please update the pr

I actually reverted to the current version of the converter, without any change, and I do see the error logs:

[2025-10-25 20:33:20] warning:  zhc:lixee: Failed to read zigbee attributes: Error: ZCL command 0x......./1 liXeePrivate.read(["currentPrice","message1","message2","drawnVAMaxN1"], {"timeout":10000,"disableResponse":false,"disableRecovery":false,"disableDefaultResponse":true,"direction":0,"reservedBits":0,"manufacturerCode":null,"writeUndiv":false}) failed ({"target":24165,"apsFrame":{"profileId":260,"clusterId":65382,"sourceEndpoint":1,"destinationEndpoint":1,"options":4416,"groupId":0,"sequence":200},"zclSequence":14,"commandIdentifier":1} timed out after 10000ms)

I think the reason I didn't see them in the logs is that the poll interval was always 10 minutes, so I missed them. It was not a JavaScript problem.

I'd say this PR is good as-is if you're OK with it.

[Koenkk]

Thanks!