Thou shalt have no other gods before the ANSI C standard 1376

I think, as is typical in human nature, the reaction to the criticism has been overdone. We went from one of the spectrum to the other. Now they're all web developers, with no concept of talking to hardware at all.

What if the job is to work on the OS itself? Write device drivers?

I submit that by taking production threaded software that runs okay on a single processor, and moving it to an SMP box and seeing it deadlock, we can safely argue that your claim is lacking in accuracy.

It is quite common to see what one thinks if thread-safe code run as expected on a uniprocessor systems, and die almost instantly on (known good) SMP hardware.

I don't know of any language that completely prevents errors while writing software with multiple threads. If you do, I'd love to hear about it.

Throw in the lock-free algorithms, and "it just works in the compiler" is completely out the window.

The arrogant polarity that causes "theory = university, superior", "practical knowledge = technical school, inferior" doesn't help. Both theory and application are needed to be competent of course. The program that teaches both would be ideal.

-- Randy Howard (2reply remove FOOBAR) "Making it hard to do stupid things often makes it hard to do smart ones too." -- Andrew Koenig