W
Wilton Helm
It all depends on what people are going to do. Not many percent of
This is probably true, but possibly sad. It is my claim in the embedded
world, that no one has any business playing there if they aren't capable of
reading a disassembly listing. It is the final word on: a) whether the code
is doing what you expected it to, and, b) whether it is doing it in an
efficient manner.
Obviously for a one-off query that will never be repeated, or an analysis of
an experiment, the effort isn't worth the trouble. The simplest (human)
path to a correct answer is the best one. But for production code (which
encompasses all of the embedded and most of the business market) efficiency
is a useful objective. I've seen so much garbage that is often 10 times a
compute intensive as what it would be if the programmer had had any concept
of how is code was being executed.
I will grant that OO takes a somewhat different mindset than C programming,
and it comes harder for some than others, and some may find it harder if
they start with procedural/sequential programming. On the other hand, for
me, the biggest insight came when I discovered that an object function call
in C++ passed a pointer to the object as the first (hidden) parameter.
Since I think along the lines of hardware, that gave me significant insight
into how OOs work, and thereby into how to use them (and use them
efficiently).
Wilton
the people who gets into programming will ever get a chance to make
a device driver or embedded software or low level stuff like that.
Most will end up making software shuffling SQL data around in one
way or another. Who knows if we really need to learn more than a
scripting language for such a task?
This is probably true, but possibly sad. It is my claim in the embedded
world, that no one has any business playing there if they aren't capable of
reading a disassembly listing. It is the final word on: a) whether the code
is doing what you expected it to, and, b) whether it is doing it in an
efficient manner.
Obviously for a one-off query that will never be repeated, or an analysis of
an experiment, the effort isn't worth the trouble. The simplest (human)
path to a correct answer is the best one. But for production code (which
encompasses all of the embedded and most of the business market) efficiency
is a useful objective. I've seen so much garbage that is often 10 times a
compute intensive as what it would be if the programmer had had any concept
of how is code was being executed.
I will grant that OO takes a somewhat different mindset than C programming,
and it comes harder for some than others, and some may find it harder if
they start with procedural/sequential programming. On the other hand, for
me, the biggest insight came when I discovered that an object function call
in C++ passed a pointer to the object as the first (hidden) parameter.
Since I think along the lines of hardware, that gave me significant insight
into how OOs work, and thereby into how to use them (and use them
efficiently).
Wilton