Thibault Martinez

I think they were the same crash. Not seeing any crash since the `iota_api_http_destroy` one has been fixed.

By the way, `bazel build -c opt --crosstool_top=@iota_toolchains//tools/x86-64-core-i7--glibc--bleeding-edge-2018.07-1:toolchain --cpu='x86_64' --compiler='gcc' --host_crosstool_top=@bazel_tools//tools/cpp:toolchain //mam/api` is equivalent to `bazel build -c opt --config x86_64 //mam/api`

You can use `--config aarch64` `--config armv7` `--config i686` `--config x86_64` `--config esp32_64` Have a look at https://github.com/iotaledger/entangled/blob/develop/tools/toolchains.rc

No ! This is about unpredictability of msg_id to void spam, let's keep it open a little while :)

Hi HowJMay, I already solved this issue in an upcoming PR so no need to work on it. Thanks anyway! Le lun. 27 mai 2019 à 15:14, HowJMay a écrit...

We heavily rely on dynamic memory allocation right now, this won't probably fit some targets

Not yet, missing 3 little issues.

I don't think empty ``` * @file * @brief ``` are necessary, let's only put them if we fill them. Looking at Doxygen documentation, only `@ingroup` is really necessary.

My tests concluded that `@file` is necessary and `@brief` isn't. I highly recommend filling `@brief` or not putting it.

Thanks for the clarification !