Echelon Neuron C - Manual (Page 59)

Neuron C Programmer’s Guide 47
Note: This discussion uses the terms writer device and reader device . A writer
device is a device that writes to a particular network variable (an output network
variable). A reader device is a device that reads a particular network variable
(an input network variable). In many cases, a device has both input and output
network variables declared in its program, and therefore acts both as a “writer
device” and a “reader device,” depending on the network variable.
When Updates Occur
The new value of a network variable received by a reader device does not take
effect immediately upon reception and processing of the message. Similarly,
assignment of a new value to an output network variable does not cause a
message to be sent immediately. Rather, updates occur at the end of a critical
section in the application program. A
critical section is defined as a set of
application program statements during which network variable updates are not
propagated.
A task is an example of a critical section: once begun, each task runs to
completion. When network variable updates are received or requested, they are
posted by the scheduler at the end of each critical section. An application can use
the post_events( ) function to divide a single task into two or more critical
sections. The post_events( ) function can be used to increase throughput and
improve response time since it forms a boundary at which outgoing network
variable updates are sent and incoming network variable updates are processed.
See Chapter 7, Additional Features , on page 145, for further discussion of
post_events( ).
Declaring Network Variables
The syntax for declaring a network variable is shown below. The first form of the
declaration is for a simple network variable, and the second form is for a network
variable array.
network input | output [
netvar-modifier ] [ class ] type
[ connection-info ] identifier
[ = initial-value ] [ nv-property-list ] ;
network input | output [ netvar-modifier ] [ class ] type
[ connection - info ] identifier [ array - bound ]
[ = initializer - list ] [ nv - property - list ] ;
Note: The brackets around the term array-bound do not, in this case, indicate an
optional field. They are required and must be part of the program.
A Neuron-hosted device can declare up to 254 network variables (including array
elements), but the device’s resource capacity (as defined by its memory map)
might not support the maximum number of network variables.
You can declare an array of network variables using the second form of the
syntax shown above. The array can only be single dimension. The array-bound
must be a constant. Each element of the array is treated as a separate network
variable for purposes of events, transmissions on the network, and so on.
Therefore, each element counts individually towards the maximum number of
network variables on a given device. Each element of the array is then a
separately bindable network variable.