ckcvg(input_buffer.i1a.r, start_index.i2.r, num_chars.i2.r, value.r4.r

	This routine decodes the number of decimal ASCII characters
	(in either F or E format) as specified by "num_chars" from the user's
	byte buffer, "input_buffer", to a floating value in "value".  The
	conversion starts at the character position specified by "start_index"
	(first character is position 1).  If a valid conversion has been
	performed, the value of "error" is returned as zero.  Otherwise, the
	position of the first "bad" character is returned in "error" as a
	negative value.

	input_buffer	ASCII buffer to convert to floating point value
	start_index	starting character in "buffer" to convert (starts at 1)
	num_chars	number of characters to convert
			(maximum is MAX_FLOAT_CHARS)
	value		returned single precision floating point value
	[error]		returned conversion status
			(OK -> successful conversion,
			 <0 -> character position of first bad character)

	This function requires the following include files:

	cnsparam_h, clib_h

	Related functions:

	ascii_to_numeric(_c), numeric_to_ascii(_c), cvdec, ckcvd, ckcve, cvg

	C/C++ usage:

	char	input_buffer[NUM_CHARS];
	short	start_index = 1;
	short	num_chars = NUM_CHARS;
	short	error;
	float	value;