On 21-Apr-14 12:13, Simon Clubley wrote:
> It's absolutely crazy that in 2014 we are still writing code which requires
> that knowledge to be manually encoded within the program code itself
> (in the form of macros) instead of just placing that information within
> the data type which is where it belongs.

Filly agreed.
In Barns's /Ada 2005/ book there was a CD (which I have unfortunately 
lost) which had several interesting PDFs, one of which dealt with 
handling endian-independent records via specification clauses [attached: 
bitorder.pdf], there's another paper [I don't recall if it was on the CD 
or not] which expand on it to do wholesale byte reversals [attached: 
Endian-Independent Paper] of an outermost record.

So it's not unaddressed in Ada, but it is a little awkward; I think some 
of these could be helped by Ada 2012 aspects though [Size & Bit_Order on 
the component types]. (There's also the question of what happens when a 
component type of the opposite Bit_Order is included... that may be 
answered in the LRM, but I simply don't recall reading it.)