gwyn@brl-smoke.ARPA (Doug Gwyn ) (03/04/88)
In article <2375@geac.UUCP> daveb@geac.UUCP (David Collier-Brown) writes: -In article <4400@aw.sei.cmu.edu> firth@bd.sei.cmu.edu.UUCP (Robert Firth) writes: ->However, my reading of ANSI X3.9-1978, especially Section 5, on arrays, ->leads me to conclude that array representation in column-major form, ->and array access by chain multiplication-&-addition of subscripts, is ->the only feasible implementation choice. - My reading of the ANSI proposal leads me to the same conclusion. -This is both good (it eliminates an ambiguity) and bad (it removes -the freedom of an compiler-writer to do what is best applicable to -her architecture). - Could someone who is more of a language-lawyer comment on this -interpretation? Was it in fact the intention of the committee? Certainly it is the intention of the proposed standard to continue this requirement, which has always existed in C, recent comments on this newsgroup to the contrary notwithstanding. The argument about "what is best applicable to the architecture" is bogus. I wonder if you guys understand C arrays at all.