08:51 | topi` | cbxbiker61: they want to be on the IoT ship, desperately :) looks like the same struggle intel is going through when it tries to gain marketshare in smartphones... |
08:51 | topi` | and even ASUS of all vendors chooses Snapdragons for their new Zenphones ;) |
08:52 | jnettlet | well the Intel subsidies are starting to dry up. They have been loosing money on their embedded stuff for many quarters now. |
08:53 | topi` | jnettlet: what kind of things have you tried on the mikroBUS? I wonder what kind of kernel support it requires? |
08:53 | topi` | if I can present it to my boss as something that "runs even a GPS", then it's all fine |
08:53 | topi` | we probably can find use cases for accelerometers and GPS'es |
08:54 | jnettlet | topi`, nothing yet. Only looked over the specs. I will be getting some when I travel down to Israel in the coming week. |
08:54 | topi` | so it is really a very new thing? |
08:55 | jnettlet | so there is nothing special about mikroBUS except the standardization. |
08:55 | jnettlet | I think they have been around for a couple of years now. They are an eastern european company. |
08:55 | jnettlet | the first year was ramp, and then over the summer they started producing lots of their modules |
08:56 | topi` | can anyone produce compatible modules or is it behind license fees? |
08:56 | jnettlet | everything is accessed over UART, I2C, or SPI |
08:56 | topi` | I just skimmed through the mikroBUS site, the variety of stuff looks neat though |
08:57 | topi` | yeah, standard interfaces :) it all makes sense. |
08:57 | topi` | the issue has always been connecting the buses in a physical way |
08:57 | jnettlet | it is just about customization and rapid prototyping |
08:57 | topi` | every board had different ways of connecting |
08:57 | topi` | I guess a GPS module is good to go with just the UART pins |
08:57 | jnettlet | You are allowed to produce your own mikroBUS™ compatible board, as long as you place mikroBUS™ logo on the board silkscreen. Also, you are not allowed to call it “Click board” because “Click” is the trademark of mikroElektronika. You are also allowed to make commercial version of your board. It would be nice to hear about it, so send us an E-mail to [email protected] |
08:58 | jnettlet | yeah you can buy a Click board with GPS |
08:58 | jnettlet | I plan on adding mikroBUS support to WiringX so we can continue to use a common library |
08:59 | topi` | isn't this WiringX a RPI thingy? |
08:59 | topi` | or was it the one developed for ODROID-C1 |
09:00 | jnettlet | nope it started as a HummingBoard thing |
09:00 | jnettlet | There was WiringPi, which we contacted the author about adding HB support for and he refused to have anything to do with it because it wasn't a RaspberryPi |
09:01 | topi` | hehe |
09:01 | jnettlet | Then a community member, CurlyMoo wrote WiringHB, which I then refused and said this should all be abstracted and generic. |
09:02 | jnettlet | which he hated me for, and then wrote WiringX which is awesome and has pulled in support for the BananaPi and a few other boards. |
09:02 | jnettlet | https://github.com/wiringX/wiringX |
09:03 | jnettlet | So I need to add support for the HB Edge, and then a generic mikroBus support that any board should be able to use. |
09:11 | topi` | I guess there's not much kernelspace stuff in wiringX? |
09:11 | topi` | how does it interface with i2c? |
09:11 | topi` | i2c should be exposed at sysfs nodes |
09:21 | jnettlet | yes |
10:16 | topi` | jnettlet: I have a huawei MU909-521 modem in my HB pro, but it seems jessie's modemmanager doesn't recognize it. Do you have any advice? the kernel is 3.14.14+ from end of April |
10:16 | topi` | when I "modprobe option", the serial ports appear to ttyUSBx as expected |
10:16 | topi` | so, maybe this is not a kernel problem at all... |
10:18 | jnettlet | topi`, does lsusb show that it is in storage mode or modem mode? |
10:20 | jnettlet | or is it pcie? |
10:20 | topi` | this huawei is a mPCIe model, so no Storage mode trick here |
10:20 | topi` | we also have an intel platform with Jessie running on it, and the same model works there |
10:21 | jnettlet | it is mini pcie? many have that form factor but are still usb based |
10:21 | topi` | which is very odd, because basically the version of modemmanager ought to be identical on the armv7 repository as well |
10:21 | topi` | yes, it is usb based (can see it with lsusb) |
10:22 | jnettlet | okay |
10:22 | jnettlet | and mmcli -L gives you nothing? |
10:23 | topi` | nothing, because modemmanager seems to reject it |
10:24 | topi` | there's "couldn't find support for ..." in modemmanager's log |
10:24 | topi` | so maybe none of the plugins wanted to have that particular model. |
10:25 | jnettlet | oh possible. I have some Telit Modems here that work fine. |
10:25 | jnettlet | although Jessie has an older version of MM also |
10:26 | topi` | I guess I'll move this question on to the #modemmanager channel (if exists) :) |
10:26 | topi` | nope :/ |
10:27 | topi` | I wonder if the freedesktop.org guys have their own IRC server somewhere |
10:28 | jnettlet | topi`, I would check the ModemManager commit logs to see if you need a newer version |
12:19 | vpeter | jnettlet: what is the latest openelec image you have? |
12:24 | jnettlet | vpeter, https://www.dropbox.com/sh/9sg6nhrigsrsqwz/AABZuBVqlbMaTwSbMQnb-eISa?dl=0 |
12:25 | jnettlet | you need to patch your .dts files to include dcic support. That is being used to as the video sync source now. |
13:37 | vpeter | jnettlet: You mean node dcic1 and dcic2 with status okay? Or something else. |
13:37 | jnettlet | yes. You can look at the commits I just did for the cbi/hb in the tree |
13:41 | vpeter | Great. |