Don't modify register addresses

[phibos]

In the file command_line.c you decrease the given register address by 1.

case 'a':
	rc = parse_int_param(optarg, &params->sad);
	if (params->sad <= 0) {
		ERR("Address should be greater than zero: %s\n", optarg);
		rc = RESULT_WRONG_ARG;
	}
	params->sad--; /* register/bit address starts from 0 */
break;

Is there any reason to modify the register address instead of using the given value? The start address of the devices I use is 1. That's why I have to set the address to 2 on the commandline. This might be very confusing for people comparing the device documentation with my check configuration.

[AndreySV]

Real register and bit addresses start in Modbus protocol from 0.

But unfortunately device manufactures use very different address schemes.
Sometimes registers start from 0, sometimes manufactures think that it's confusing that the first address is zero and start registers from 1.
There are address schemes that include register type (input/holding) in the address. For example, 30001 (first input register), 40001 (first holding register).

The real address is stored internally in params->sad. There is a short description about that in the man page .

[phibos]

Sure. But I think it is a bit confusing. Might be a better idea to add a command line option like --fix-address and decrease the address if set or to keep compatibility add a command line option like --use-raw-address and don't decrease the address if set.

[AndreySV]

What about introducing different addressing schemes as additional parameter: "start from 0", "start from 1", "address space and register number together"? By default, address will be interpreted as before: "start from 1".