[Datafield] requestWrite with WRITE_TYPE_DEFAULT: onCharacteristicWrite Not Triggering?

Are you doing any queueing?  You can't just do a bunch of reads/writes in a row,  Do a request, wait for a response, do the next request, wait for a response, etc.

The NordicThingy sample in the SDK does this, as do the apps I use with raspberry pi devices (see commqueue.mc)

https://forums.garmin.com/developer/connect-iq/b/news-announcements/posts/would-you-like-some-raspberry-pi-with-your-connect-iq

Thank you for your quick answer.
Yes, I'm queueing my BLE commands while waiting the callback to be triggered and also not sending more than the MTU (20).

Is there a reason you don't use WRITE_TYPE_WITH_RESPONSE all the time?  The callback is triggered when the write has been handled, not when the request is sent

In my use case, it’s not an issue if a BLE command is lost. I would like to send data as frequently as possible.

If you aren't waiting for the response, you could be flooding the BLE device when it's not ready.

I think he does but the problem is the callback "onCharacteristicWrite" does not trigger, so it is impossible to use correctly mode WRITE_TYPE_DEFAULT with last Garmin firmware?

Which device/FW version?  What happens in the sim?

I always use WRITE_TYPE_WITH_RESPONSE, and it works fine....

As it is written above: 
- Vivoactive 5 (v12.10) : problem

- Simulator (CIQ 7.3.1) : no problem


I don't think simulator is relevant versus real behaviour on device... it is not the first time I see different behaviour between reality and simulator https://forums.garmin.com/developer/connect-iq/i/bug-reports/datafield-crashes-when-updating-sdk-from-6-3-1-to-7-3-1

It is MUCH easer to debug/test BLE with the sim.  The HW required is about $12US (the dongle).  What happens if you use WRITE_TYPE_WITH_RESPONSE?  It's worked fine for me for years on different devices.

It works but it is 2x time slower in terms of BLE bandwidth :-(