can anyone help me:

I get the latest zephyr code, but build fail:

-----Original Message-----
From: users@lists.zephyrproject.org <users@lists.zephyrproject.org> On
Behalf Of Nikos Karamolegkos via Lists.Zephyrproject.Org
Sent: 03 July 2019 13:07
To: users@lists.zephyrproject.org
Cc: users@lists.zephyrproject.org
Subject: [Zephyr-users] DTLS over CoAP for the Zephyr project

Hello everyone,

I have seen that other IoT OS use tinydtls for authentication over CoAP
and IEEE 802.15.4. On the other hand,  the Zephyr RTOS I have seen (in
documentation 1.9) that uses mbed TLS but I confused on how to set it in
IoT end devices. Specifically, I was wondering if there is any tutorial
to start (e.g in reel boards). Also, I would like to know if the DTLS is
based on pre-shared keys or in any elliptic curve cryptography (ECC)
algorithm?

Thank you for your time

--
Nikos Karamolegkos
R & D Engineer at ICS-FORTH
Telecommunications and Networks Lab (TNL)

I'm afraid that's the way CMake's cache works:

https://stackoverflow.com/questions/26145865/confused-about-cmakes-cached-variable-setting-priority

There are at least two "workarounds":

cmake -Be70build -DBOARD=sam_e70_xplained ...

or:

rm -rf build; cmake -Bbuild ...

On a new Ubuntu install, I'm having this issue with Zephyr using West.  It's also referenced in this issue, but the fix mentioned there didn't work for me:

https://github.com/zephyrproject-rtos/zephyr/issues/15246

$ west build -b nrf51_pca10031
-- west build: build configuration:
       source directory: /home/hp1/zephyrproject/zephyr/samples/hello_world
       build directory: /home/hp1/zephyrproject/zephyr/samples/hello_world/build
       BOARD: nrf51_pca10031 (origin: command line)
-- west build: generating a build system
Zephyr version: 1.14.99
-- Selected BOARD nrf51_pca10031
-- Found west: /home/hp1/.local/bin/west (found suitable version "0.5.7", minimum required is "0.5.6")
CMake Error at /home/hp1/zephyrproject/zephyr/cmake/extensions.cmake:1324 (message):
  Assertion failed: The detected dtc version is unsupported.

      The version was found to be 1.4.5                                  
      But the minimum supported version is 1.4.6                       
      See https://docs.zephyrproject.org/latest/getting_started/                    
      for how to use the SDK's dtc alongside a custom toolchain.
Call Stack (most recent call first):
  /home/hp1/zephyrproject/zephyr/cmake/host-tools.cmake:69 (assert)
  /home/hp1/zephyrproject/zephyr/cmake/app/boilerplate.cmake:412 (include)
  CMakeLists.txt:5 (include)


-- Configuring incomplete, errors occurred!


I did try the two environmental variables, which are now part of the Zephyr instructions, but the problem persists.

export ZEPHYR_TOOLCHAIN_VARIANT=zephyr
export ZEPHYR_SDK_INSTALL_DIR=$HOME/zephyr-sdk-0.10.1

I'm unsure how to go about fixing the DTC version issue.

-----Original Message-----
From: users@lists.zephyrproject.org <users@lists.zephyrproject.org> On
Behalf Of Bolivar, Marti via Lists.Zephyrproject.Org
Sent: 03 July 2019 19:20
To: kayli0109@gmail.com; Herbert, Marc <marc.herbert@intel.com>
Cc: users@lists.zephyrproject.org
Subject: Re: [Zephyr-users] cant perform west init

" via Lists.Zephyrproject.Org"
<kayli0109=gmail.com@lists.zephyrproject.org> writes:

Hi Marc,

Thanks for the response.
In that case, can i install multiple version of Python? What version
of Python needed in this project?

Currently, Python 3.4+ is required.

Thanks,
Marti

Regards,
Kay Li

On Wed, 3 Jul 2019 at 10:07, Herbert, Marc <marc.herbert@intel.com>

wrote:

You have a very old Python version in your HOME directory, is that on
purpose? Does some other project require it maybe?



Marc



*From: *<users@lists.zephyrproject.org> on behalf of KAY LI NG <
kayli0109@gmail.com>
*Date: *Tuesday, 2 July 2019 at 18:15
*To: *"users@lists.zephyrproject.org" <users@lists.zephyrproject.org>
*Subject: *[Zephyr-users] cant perform west init



Hi there,



While performing west init, i face this issue.



$ west init zephyrproject



Traceback (most recent call last):

File "/home/gchee2/.local/bin/west", line 11, in <module>

sys.exit(main())

File
"/home/gchee2/.local/lib/python3.4/site-packages/west/_bootstrap/main
.py",
line 501, in main

init(wrap_argv[1:])

File
"/home/gchee2/.local/lib/python3.4/site-packages/west/_bootstrap/main
.py",
line 257, in init

bootstrap(args)

File
"/home/gchee2/.local/lib/python3.4/site-packages/west/_bootstrap/main
.py",
line 323, in bootstrap

exist_ok=True)

File
"/home/gchee2/.local/lib/python3.4/site-packages/west/_bootstrap/main
.py",
line 124, in clone

_banner('Cloning {} from {}, rev. {} into {}'.format(desc, url,
rev,
dest))

File
"/home/gchee2/.local/lib/python3.4/site-packages/west/_bootstrap/main
.py",
line 162, in _banner

print(colorama.Fore.LIGHTGREEN_EX, end='')

AttributeError: 'AnsiCodes' object has no attribute 'LIGHTGREEN_EX'



Some online resources say its the colorama version not up to date,
but i cant seems to upgrade it.



Any idea how to solve it?



Thank you!



Regards,

Kay Li

Hi Johan,

Thanks for the clarification. I am referring to group address. I missed setting up proxy filtering.

Rgds,

Venkat.

On Tue, 18 Jun 2019 at 9:54 PM, Hedberg, Johan <johan.hedberg@...> wrote:

Hi Venkat,

On 17 Jun 2019, at 16.33, Venkat Rao Vallapaneni <vallapaneni@...> wrote:
> I get messages if the destination address is same as proxy node address (like get the status of light message is received on proxy node notification characteristic).
>
> What if the destination address (publish address) is different than proxy node address? Does Zephyr mesh application sends those messages also to proxy notification characteristic? Or does it send messages that are directed to only proxy node?

That depends on what the publish address is. If it’s a group address then any node that subscribes to this address, including the proxy node, will receive it. If it’s a unicast address it needs to belong to one of the elements on the proxy node. I’m not familiar with the implementation of nRF Mesh, however the proxy protocol has a mechanism through which proxy clients (like nRF Mesh) can inform the proxy server of additional group addresses the client is subscribed to. This way the proxy knows to forward messages to the client even though the destination isn’t a unicast address of the client.

Johan

--

Rgds,

Venkat.

Hi,

I am trying BLE mesh on-off app on nrf52840 USB dongle. I have provisioned two devices.
After provisioning, I have connected from my android device to node1 and sending messages to configuration server of node2. Node1, node2 and mobile are next to each other.
I am sending 25 configuration messages (binding app keys, setting pub and adding sub) for all generic onoff server/client models in node2.

Sometimes, I see that I don't get response (No response could be first message or any message in between or sometimes even I am successful in binding app keys to all models, but rarely).
When I debugged, I noticed that node2 receives configuration message (node1 relays the message received by the android device) and it also sends response with mesh message.

This mesh message is missed by node1. I am trying to understand more on why this mesh message advertising packet was missed by node1.

I noticed that in mesh/adv.c, I see that min interval and max interval are same. In the specifications, it is mentioned that: "The Advertising_Interval_Min shall be less than or equal to the Advertising_Interval_Max. The Advertising_Interval_Min and Advertising_Interval_Max should not be the same value to enable the Controller to determine the best advertising interval given other activities."

param.interval_min = ADV_SCAN_UNIT(adv_int);
param.interval_max = param.interval_min;

Why these two values are set same?

I have tried making interval_max as ADV_SCAN_UNIT(adv_int+5) but didn't help.

I have used default settings in mesh on-off sample app:

/* 3 transmissions with 20ms interval */
.net_transmit = BT_MESH_TRANSMIT(2, 20), 

I increased the transmissions to 6 (along with max interval greater than min interval) and it improved (3/5 times successful in sending 25 configuration messages back to back).

Any other recommended approaches to minimize mesh message advertising packet losses?

Thanks,
Venkat.