From: Dirk Gerrits (dirk_at_[hidden])
Date: 2002-09-23 12:14:23
Craig Henderson wrote:
> In this case, the conversion to unsigned char truncates
> the source data type to 8 bits and throws the error. Like you say, this is
> the required behaviour in this situation, and therefore a simple "0xff &
> (...)" is enough to be specific about the required result. I've not tested
> the optimisation of this, but I would not expect any problems.
Shouldn't that be "UCHAR_MAX & (...)"?
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk