Two issues need clarification:
What should be the value range of the BCR.actVal? Meaning in what value it returns back to zero (wraps around).
Should it be the maximum allowed by the data type (Ed1 INT128, Ed2 INT64) or is it a "local issue"?
For Ed.1 one could in principle have an INT32 behind the INT128 type, (documented in the MICS or the PIXIT).
Is the same still be valid for Ed.2 (have INT32 behind the INT64)
15 Aug 13
In Force (green)
for Edition 2, it shall be required to use the INT64 type; that means the value shall wrap at the maximum value and minimum value to zero
19 Apr 13
So based on Thierrys comment, the conclusion would be, to mandate that the value wraps at the maximum value of INT64. This would be an additional explanation required in part 7-3
14 Jan 13
1) value range are defined in 7-2 both for INT32, INT64
the wrap value is not an issue of BCR but of Integer. The same issue happens for INT32.
INT32 wraps to 0
INT64 wraps to 0
Here again, this is no edition 1 / edition 2 problem. This is an Integer problem. INT28 wraps also to 0 and its range was defined in 7-2 Ed1.
2) there are very few CPU that supports Int128. It was then tolerated in Ed1 to have an INT32 to implement the actVal. However, there was never an INT128 behind those implementation.
As for Ed2, INT64 has been retained for BCR, and I suppose that Int64 should be used for the type. Not an Int32.