News for hackers. Stuff that matters.
Re:Apple's Module proposal to replace headers for C-based la... by cperciva on Monday November 26, 2012 @01:37PM (#4833224)
I don't buy the performance argument: NxM -> N + M only works if every one of the N .c files is including every one of the M .h files.

If you're spamming #includes like that, you need to fix your #includes, not redefine the language.

Re:Apple's Module proposal to replace headers for C-based la... by swift on Monday November 26, 2012 @04:29PM (#4834116)
If a .c file includes U x M .h files on average, where U is a number from 0 to 1, then the work we expect to do at compilation time is proportional to U x (N x M). This is still O(N x M).
Re:Apple's Module proposal to replace headers for C-based la... by cperciva on Monday November 26, 2012 @08:33PM (#4835088)
True, but in my experience U tends to decrease sharply as M increases.