Re: [boost] [dynamic_bitset] Endianess and Adaption Vs Copying

Subject: Re: [boost] [dynamic_bitset] Endianess and Adaption Vs Copying
From: Phil Endecott (spam_from_boost_dev_at_[hidden])
Date: 2012-05-04 06:57:32

Joel wrote:
> Bit ordering is completely arbitrary.  If the constructor for reference in
> dynamic_bitset.hpp is changed from:
>
>         // the one and only non-copy ctor
>         reference(block_type & b, block_type pos)
>             :m_block(b),
>              m_mask( (assert(pos < bits_per_block),
>                       block_type(1) << pos )
>                    )
>         { }
>
> to:
>
>         // the one and only non-copy ctor
>         reference(block_type & b, block_type pos)
>             :m_block(b),
>              m_mask( (assert(pos < bits_per_block),
>                       ((block_type(1) << (bits_per_block-1)) >> pos ))
>                    )
>         { }
>
> and the ordering would be reversed.  Now on little-endian machines, the ordering
> nicely goes always left-to-right.

If you're thinking about individual bits being "next to" bits in other 
bytes, and bits or bytes being "left" or "right" of each other, your 
mental model of endianness is over-complicated.  Because bits are not 
individually addressable, the only thing that matters is the order of 
bytes (which are addressable) within words.  Ordering of bits only 
matters when someone draws a picture of something like a peripheral register.

Regards,  Phil.