> Even if you use volatile everywhere and use an
>algorithm that avoids atomic RMW like Dekker's, then nothing is guaranteed about hardare memory model.
>

Hmm. You used Dekker's algo to make your point. THat's clever. Just of curiosity, what do you do for living?



>This is a stupid tangent though, you're just hoping to catch me out and make yourself
>feel superior. In reality you've shown not much understanding of the language or your claims yourself.

Gotta admit you're difficult to bait.

Allow me give you an example why C sucks. Say the job is to calculate the predicted realtime trajectory of 5000 online players. You can write less code in pretty much any other language than C to get the job done. And do so in less time. There are several trains of thought and methodoligies to solve the problem. C's language and library facilities are too primitive to support any of them well.

Granted C is popular. But it's 40 years old. And it disregards advances CS made as a (minor) scientific discipline. C, more than anything, doomed many talented young people to wasted million of hours month after month on drudgery that's compeletly pointless, that can be easily removed by a small amount of decent compiler and language work.

But that'd too much to ask. Even Thompson's new language work would get buried by the shear weight existing C code.

It all boil down to how much do code monkeys' life worth? The answer is not much. Just work them longer..