[knx] Bugfix problem with DPT 251.600 #15723
Conversation
|
Technically this is correct. The question is what actual problem occurs with the current implementation, because this change is breaking and I can't see any real issue with using 0-255. |
|
When i try to send a new color to the bus i have always an error out of range. See my detailed description in this threat: #https://community.openhab.org/t/rgbw-color-out-of-range-dpt-251-600/150055 |
In the old implementation the datatype was implemented with byte 0-255. The DPT 251.600 is defined with 0-100%, change this from byte to PercentType. Signed-off-by: Marco Mueller <marco@ms.mueller.ch>
|
This pull request has been mentioned on openHAB Community. There might be relevant details there: https://community.openhab.org/t/rgbw-color-out-of-range-dpt-251-600/150055/2 |
I have checked the code again. I don't see that the change is breaking, actual we have a bug. When you send a value over 100 calimero reject the request. That means all value over 100 is not working and will not send to the knx bus. |
|
Correct, I misread the change. I would suggest a minor change though, to get higher accuracy. |
I'm complete open for any suggest, please tell me what we can optimize. |
|
@J-N-K @genesis81 Give me some time, just trying to get a proper test which includes the Calimero part. Until now we have only tests for ValueEncoder decode/encode. With a custom KNXNetworkLink implementation, we should see everything down to the KNX frame level including the coding done in Calimero. |
...org.openhab.binding.knx/src/main/java/org/openhab/binding/knx/internal/dpt/ValueEncoder.java
Outdated
Show resolved
Hide resolved
…ing/knx/internal/dpt/ValueEncoder.java Co-authored-by: J-N-K <github@klug.nrw> Signed-off-by: Marco Müller <marco@ms-mueller.ch>
|
@holgerfriedrich Really nice. I think this can be fixed anyway, I check in calimero code that a float in the range from 0 to 100 is expected here. |
|
I think you need to apply spotless |
Signed-off-by: Marco Müller <marco@ms-mueller.ch>
Signed-off-by: Marco Müller <marco@ms-mueller.ch>
Signed-off-by: Marco Müller <marco@ms-mueller.ch>
|
Test failures are independent, I also see them in my local builds without any changes. |
Looks like changes in the core system that break the tests. |
|
I already commented in the PR that includes the cause of these problems. |
There was a problem hiding this comment.
LGTM, thanks for providing this bugfix!
Just for reference: openhab/openhab-core#3834 (core change currently breaking CI test) and #15727 (WIP new tests)
* Fix problem with DPT 251.600. In the old implementation the datatype was implemented with byte 0-255. The DPT 251.600 is defined with 0-100%, change this from byte to PercentType. Signed-off-by: Marco Mueller <marco@ms.mueller.ch>
* Fix problem with DPT 251.600. In the old implementation the datatype was implemented with byte 0-255. The DPT 251.600 is defined with 0-100%, change this from byte to PercentType. Signed-off-by: Marco Mueller <marco@ms.mueller.ch>
* Fix problem with DPT 251.600. In the old implementation the datatype was implemented with byte 0-255. The DPT 251.600 is defined with 0-100%, change this from byte to PercentType. Signed-off-by: Marco Mueller <marco@ms.mueller.ch> Signed-off-by: querdenker2k <querdenker2k@gmx.de>
* Fix problem with DPT 251.600. In the old implementation the datatype was implemented with byte 0-255. The DPT 251.600 is defined with 0-100%, change this from byte to PercentType. Signed-off-by: Marco Mueller <marco@ms.mueller.ch> Signed-off-by: Jørgen Austvik <jaustvik@acm.org>
In the old implementation the datatype was implemented with byte 0-255.
The DPT 251.600 is defined with 0-100%, change this from byte to PercentType.
Signed-off-by: Marco Müller marco@ms-mueller.ch