Richard A. O'Keefe wrote:
[...] arsis/arses (I refer of course to a technical term in poetry)
And in music too, apparently; you learn something new every day.
3. C's bitwise operations are unusually, not to say unwisely, placed.
No kidding. I've been programming in C since 1984, but I still
couldn't write down the precedence rules accurately if my life
depended on it; thankfully, so far, it hasn't. Rather than waste
countless eye-twinklings looking them up, I just parenthesize any
non-trivial expressions, and get on with my life. If this reduces
my standing among studly programming aficionados, the total number
of monkeys I could give tends towards zero, since I'd rather
have working code today than appreciated code eventually.
This lack of desire to memorize random stuff when there is
a perfectly good alternative is, I suspect, a major factor in
Derek's guinea-ppigish programmers not knowing the relative
precedence of '^', '&', '&&' and '>>'.
Derek M Jones wrote:
The ACCU conference is the Association of C and C++ users, and it
also has a Java stream. So I think it is reasonable to assume that
subjects were very familiar with these languages.
If you're going to do research into how well people grasp the
fiddling details of a language, especially one as knobbly as C,
you really need to rule out mental cross-talk from those
differently-knobbled languages such as Java, C++, Javascript,
perl, bash, SQL, as well as many others that it's apparently
still legal to use.
I know from experience that switching from Java to perl to
PL/SQL to C to ruby to Javascript is sometimes like turning
a mental oil tanker. They all use similar subsets of concepts
and symbols, yet all manage to have up-trippingly distinct
semantics creeping out from their dank, unkempt corners.
So, some of your expert C programmers might be hungover
from some recent binge programming in VBA or perl or C++, and
you really need to control for that to make sense of observations
like the '<<' and '^' ones.
Finally, donning my curmudgeon's hat, I note from the conclusion
that "this is the first experimental evidence that software developer
performance is [affected] by the number of times language constructs
are encountered in source code." Pardon me for being thick, but are
you saying here that you've now shown that the more programmers read
something, the more they understand it? Isn't this Obvious (with
a capital 'ooh') from the general case of 'practice makes perfect'?
Do we really need any evidence whatsoever for the specific case of
understanding ambiguous expressions in a dubiously-designed language?
Just asking.
--
Frank Wales [EMAIL PROTECTED]
----------------------------------------------------------------------
PPIG Discuss List (discuss@ppig.org)
Discuss admin: http://limitlessmail.net/mailman/listinfo/discuss
Announce admin: http://limitlessmail.net/mailman/listinfo/announce
PPIG Discuss archive: http://www.mail-archive.com/discuss%40ppig.org/
