Thanks for making us aware of the issue 
BTW do you have an extra ESP32 by any chance?
I will receive 3 new pieces tomorrow
Great, then you can install OpenMQTTGateway version 1.6.0 on one of them.
Give it a unique gateway name like OMG_1_6_0 or something, and leave the decoder URL in platformio.ini as is, as the new Decoder version has also required many changes in OMG since then.
And with OMG 1.6.0 the READ/WRITE command topic actually did have the /config at the end, that’s why I had it in my OpenHAB Thing definition
So your current dev OMG gateway nicely displays the decoded Blind Tilt, but with this 1.6.0 gateway you can use it to only send the commands for the Blind Tilts. The response messages should show on both gateways.
So with only ever using a single command, this shoudl work nicely with the 1.6.0 version gateway.
home/OMG_1_6_0/commands/MQTTtoBT/config
{
"ble_write_address": "D6:48:35:AF:72:BA",
"mac_type": 1,
"ble_write_service": "0d00",
"ble_write_char": "0002",
"ble_write_value": "570F450105FF32",
"value_type": "HEX",
"ttl": 4,
"immediate": true
}
And once we got this regression fixed it’ll all be avaialble within one gateway again
I’ve done a test with a 1.6.0 and it crash too… 
Here is the output on the serial console :
N: [ MQTT->OMG ]: {"ble_write_address":"CA:AA:3E:8F:C7:F4","mac_type":1,"ble_write_service":"0d00","ble_write_char":"0002","ble_write_value":"570f4501010128","value_type":"HEX","ttl":4,"immediate":true}
N: Send on /BTtoMQTT msg {"bleconnect":true,"interval":55555,"adaptivescan":true,"intervalacts":55555,"intervalcnct":3600000,"scanduration":10000,"onlysensors":false,"randommacs":false,"hasspresence":false,"presenceTopic":"presence/","presenceUseBeaconUuid":false,"minrssi":-100,"extDecoderEnable":false,"extDecoderTopic":"undecoded","filterConnectable":false,"pubadvdata":false,"pubBeaconUuidForTopic":false,"ignoreWBlist":false,"presenceawaytimer":120000,"movingtimer":60000,"btqblck":7,"btqsum":4,"btqsnd":1,"btqavg":4,"bletaskstack":3024,"blecoretaskstack":3096,"blestarts":1}
Guru Meditation Error: Core 1 panic'ed (StoreProhibited). Exception was unhandled.
Core 1 register dump:
PC : 0x40090fef PS : 0x00060a30 A0 : 0x801a3e8d A1 : 0x3ffd2160
A2 : 0x00000000 A3 : 0x3ffd222c A4 : 0x0000000e A5 : 0x00000000
A6 : 0x66303735 A7 : 0x31303534 A8 : 0x00000000 A9 : 0x3ffd2120
A10 : 0x3ffd2224 A11 : 0x3fff3a6c A12 : 0x0000000e A13 : 0x0000ff00
A14 : 0x00ff0000 A15 : 0x40404040 SAR : 0x00000018 EXCCAUSE: 0x0000001d
EXCVADDR: 0x00000000 LBEG : 0x40090fd0 LEND : 0x40090fe6 LCOUNT : 0xffffffff
Backtrace: 0x40090fec:0x3ffd2160 0x401a3e8a:0x3ffd2170 0x401a3f29:0x3ffd2190 0x401a4053:0x3ffd21b0 0x400e51dd:0x3ffd21d0 0x400e53e3:0x3ffd22f0 0x400e7e02:0x3ffd2330 0x400e7eab:0x3ffd2700 0x401c2df3:0x3ffd2720 0x400eaca9:0x3ffd2740 0x400e9772:0x3ffd2780 0x4010fe89:0x3ffd27c0
This is very bad 
I had done a READ test with 1.6.0, which had also failed with the latest dev, and it working nicely I just assumed that WRITE would also still work with 1.6.0.
But testing your WRITE command, as well as Sebbo’s Curtian WRITE command, I am reproducing the same crashes.
With Sebbo being really happy with the response time and all, I can only say that he was using version 1.2.0 back then, if you really feel like giving it a test with that version to have a successful result.
What really needs to happen now though is that the READ/WRITE implementation gang of OMG has a good look at this and makes sure that both READing and WRITING commands work nicely again with the current development version and all future releases 
Hi @DigiH,
Do you know if the regression has been solved ?
I did a small test with the last dev version and it still crash when receiving command on config topic
Hi @Daniel615be ,
It should not crash, but you should not use the config topic for commands with the latest dev version, see the associated documentation: