[netatmo] Room-Thing offline after restart

[clinique]

Correcting issue by late loading of the capability.
Resolves #13403
Signed-off-by: clinique gael@lhopital.org

[clinique]

This solves also these warning raised at OH startup:

2022-09-30 08:50:33.091 [WARN ] [.core.thing.binding.BaseThingHandler] - Handler ModuleHandler of thing netatmo:room:home:59959297b4809d1aff8b4726:dining tried accessing its bridge although the handler was already disposed.
2022-09-30 08:50:33.130 [WARN ] [.core.thing.binding.BaseThingHandler] - Handler ModuleHandler of thing netatmo:room:home:59959297b4809d1aff8b4726:office tried accessing its bridge although the handler was already disposed.
2022-09-30 08:50:33.213 [WARN ] [.core.thing.binding.BaseThingHandler] - Handler ModuleHandler of thing netatmo:room:home:59959297b4809d1aff8b4726:parents tried accessing its bridge although the handler was already disposed.
2022-09-30 08:50:33.275 [WARN ] [.core.thing.binding.BaseThingHandler] - Handler ModuleHandler of thing netatmo:room:home:59959297b4809d1aff8b4726:kitchen tried accessing its bridge although the handler was already disposed.

[jlaur]

@clinique - if you change "Solves issue" to "Resolves" in the PR description, the issue will be automatically linked and closed upon PR completion. :)

[jlaur]

@clinique - if you change "Solves issue" to "Resolves" in the PR description, the issue will be automatically linked and closed upon PR completion. :)

To be clear: "Resolves #xxxxx", not "Resolves issue #xxxxx" - still not linked.

[lolodomo]

LGTM, thank you

[lolodomo]

@clinique ! you arre now avoiding to retry the API request in case the API limit is reached but you agree that this is not a fix for #13358 ?

[clinique]

@clinique ! you arre now avoiding to retry the API request in case the API limit is reached but you agree that this is not a fix for #13358 ?

It should fix it. The thing being I did not find how many time we've got to wait when Api limit is reached.

[clinique]

I've got another issue to fix. From time to time API does not answer to camera pings (TimeoutException, most of the time in the middle of the night) and then the binding goes on error. I'll study this once this PR is available to see how it behaves.

[lolodomo]

@clinique ! you arre now avoiding to retry the API request in case the API limit is reached but you agree that this is not a fix for #13358 ?

It should fix it. The thing being I did not find how many time we've got to wait when Api limit is reached.

I see nothing new in that PR that could avoid reaching the API limit.
Please develop.

[clinique]

Sorry. Misunderstood. It not meant to avoid reaching API limit but rather handle it more gracefully

[lolodomo]

Sorry. Misunderstood. It not meant to avoid reaching API limit but rather handle it more gracefully

So we agree.
The issue #13358 is more how to fix the binding to avoid reaching that limit.