Netlinx Buffer

jgreer9514

I would like the ability to have a Netlinx module have a dynamic device passed into the module. I tried various ways of doing this but the buffer allocation will always attach to the original device passed into the module. I could decalre modules and have the users the ability to choose the type of device on the port not using Duet.

So an API call to reallocation the buffer to the device in Netlinx.

John

ericmedley

jgreer9514 wrote: »

I would like the ability to have a Netlinx module have a dynamic device passed into the module. I tried various ways of doing this but the buffer allocation will always attach to the original device passed into the module. I could decalre modules and have the users the ability to choose the type of device on the port not using Duet.

So an API call to reallocation the buffer to the device in Netlinx.

John

Variables are passed into a Netlinx module by reference. So, what the value is at startup is whats passed. To get dynamic info in you're going to have to do it by other means.

The easiest way that comes to mind is to create a command protocol and send commands to a virtual device created for the module.

Something along the lines of

send_command vdv_my_vDevice,"NEW_Device_Num:123:1:0

Then in the module you'd parse the string and reference the device with the new values.

It's going to be tricky in that events get ugly when trying to alter devices.

jjames

ericmedley wrote: »

Variables are passed into a Netlinx module by reference. So, what the value is at startup is whats passed. To get dynamic info in you're going to have to do it by other means.

Are you sure? I pass a variable over to my JSON parser and then parse directly on that within the module. It's here on the forum (and I think I posted the source code) if you want to look.

So couldn't in the main program the device number would be a DEV var and when it changes in main code, it changes in the module?

Or maybe, I'm just not quite following . . .

jgreer9514

buffer

If you have a device declared and passed into the module and change it within your program there is no way to make the module re-sync the buffer with the new device. There is for touch panel events but not for the data event.

=(

jjames

Curiously - is this a TPControl feature request? I'm not sure what they'd be able to do about. Again, perhaps posting this in the proper sub-forum would be of help.

AMXJeff

jgreer9514 wrote: »

I would like the ability to have a Netlinx module have a dynamic device passed into the module. I tried various ways of doing this but the buffer allocation will always attach to the original device passed into the module. I could decalre modules and have the users the ability to choose the type of device on the port not using Duet.

So an API call to reallocation the buffer to the device in Netlinx.

John

If you stay way from buffers and use DATA_EVENT this is easily done.

// INSIDE MODULE
DEFINE_EVENT
DATA_EVENT[vdvDeviceID]
{
     STRING:
     {
        // DO BUFFER PARSING HERE...
     }
}

DEFINE_PROGRAM

if (vdvNewDeviceID != vdvDeviceID)
{
    vdvDeviceID = vdvNewDeviceID;

   REBUILD_EVENT();
}

HARMAN_lasthuman

I'm still not 100% sure exactly what you're trying to do but it sounds like you might just want to try using DYNAMIC_VIRTUAL_DEVICE. This will dynamically assign a device number in the 36,864+ address range. You'll guarantee a unique device number

Ex.

DEFINE_DEVICE

vdvDynVirtual = DYNAMIC_VIRTUAL_DEVICE

papadouk

lasthuman wrote: »

I'm still not 100% sure exactly what you're trying to do but it sounds like you might just want to try using DYNAMIC_VIRTUAL_DEVICE. This will dynamically assign a device number in the 36,864+ address range. You'll guarantee a unique device number

Ex.

DEFINE_DEVICE

vdvDynVirtual = DYNAMIC_VIRTUAL_DEVICE

Looking at the Netlinx.axi the DYNAMIC_VIRTUAL_DEVICE is statically assigned with a device number of 36864:1:0. Unless this device number is an alias for assigning a new device number then one will end up declaring many instances of the same device (?).

Kostas