Dat is vreemd. Waarom zouden ze ervoor gekozen hebben een bigendian array te maken?
Volgens
http://forum.synology.com...d_of_CPU_does_my_NAS_have
Heeft (mijn / een) Synology een Marvell Armada XP.
En volgens
https://dev.openwrt.org/wiki/platforms
is dat "gewoon" little-endian.
Zal wel weer iets met backward compatibility te maken hebben (1e processor ooit in een Synology big-endian?). Maar goed, dat is dezelfde reden waarom little-endian nu zo dominant is (in de consumentenmarkt althans).
Leuk leesvoer:
Wikipedia: EndiannessHistorically, byte order distinction was born out of the mainframe vs. microprocessor approach.[dubious – discuss] Until the 1970s virtually all processors were big-endian. The first microprocessor, in the Datapoint 2200, was designed using simpler bit-serial logic where little-endian address and data formats facilitate carry propagation. Datapoint's initial specification was big-endian but to save transistors, they acquiesced to Intel's request for a design change to little-endian. When byte-parallel computation was implemented in later processors (ex: 8080), Intel left in the little-endian format as a compromise for consistency with Datapoint's earlier bit-serial microprocessor. Datapoint never used the 8008 chip (using an MSI equivalent implementation as Intel was unable to deliver it in time), however, and the little-endian format was never actually necessary.[5][6]
We are pentium of borg. Division is futile. You will be approximated.