Closed MSL-DA closed 3 months ago
Forgive me... I'm new to this. Trying again.
A question about "modelid", is it necessary to enter it twice?
And what about uppercase vs lowercase?
Historically we have a bit of weird format for multiple modelid + manufacturer pairs. Yes it's required that for each modelid the matching manufacturer name must be placed in the array at the same position.
The upper/lower case must match the exact same values as shown in the Basic Cluster since during matching the DDF these are compared without any conversation (in order to be explicit instead of guessing).
@manup as I recall, we've made the manufacturer name case insensitive? That was done since the legacy code is partially trying to unify the upper/lower case zoo we've out there. So in this regard, the proposed change wouldn't even be necessary.
But in this case it is. It was not a problem with version 2.26.3
@manup as I recall, we've made the manufacturer name case insensitive? That was done since the legacy code is partially trying to unify the upper/lower case zoo we've out there. So in this regard, the proposed change wouldn't even be necessary.
Good point, indeed before v2.27.0-beta this was the case. The current code doesn't do this this anymore (after the switch to atom based strings, where each unique string has a 32-bit number). I'm not sure how big the zoo of case variants is.
I'll check how to bring non case based strings for manufacturer names back for DDF matching.
The case sensitive regression is fixed by https://github.com/dresden-elektronik/deconz-rest-plugin/pull/7864
Closing this as the change doesn't matter after this https://github.com/dresden-elektronik/deconz-rest-plugin/pull/7864#issue-2437462856
Hey @MSL-DA, thanks for your pull request!
DDB changes
Modified
heiman/hs1sa_smoke_sensor.json
: HS1SA :heavy_check_mark:Validation
:clock730: Updated for commit 3c987f7c9f4c593594bac20048d6ebdd75ee72d9