... and (buffer+8) would not be a valid indexable identifier for an array in a true HLL. And, it's just an another way how pointer arithmetics accomplishes. After all, the first thing a C course teaches, that arrays in C are just a different notation of pointers, and the indices are just a different notation for the pointer arithmetics.

At the end of the day, I see nothing non-C-ish in *(buffer+8+index), and, as you can see on the example above, it's nicely optimised by SDCC (and I don't doubt Keil would optimise it too).

Nevertheless, in a truely HLL approach, it's the task of the compiler's optimiser to take care of these cases, no matter how are they written down. A HLL program is supposed to convey the programmer's intention, rather than give a prescription to the compiler; and it's the compiler's (or, the compiler-maker's) task to figure out how to accomplish the programmer's intention in the given platform.

JW


PS. Of course I see your point and it's a valid one. But a straightforward agreement would not make a discussion, would it? :-P