On Saturday, 1 September 2018 at 05:51:10 UTC, rikki cattermole
wrote:
Then there are polytechnics which I went to for my degree,
where the focus was squarely on Industry and not on academia at
all.
But the teaching is based on research in a good engineering
school...
But in saying that, we had third year students starting out not
understanding how cli arguments work so...
Proper software engineering really takes 5+ years just to get
started, 10+ to become actually good at it. Sadly that won't be
acceptable in our current society.
The root cause of bad software is that many programmers don't
even have an education in CS or software engineering, or didn't
do a good job while getting it!
Another problem is that departments get funding based on how many
students they have and many students are not fit to be
programmers. Then you have the recruitment process and people in
management without a proper theoretical understanding of the
topic looking for "practical programmers" (must have experience
with framework X) which basically means that they get programmers
with low theoretical understanding and therefore fail to build an
environment where people can learn... So building a good team
where people can become experts (based on actual research) is
mostly not happening. It becomes experience based and the
experience is that it isn't broken if customers are willing to
pay.
Basic capitalism. Happens outside programming too. Make
good-looking shit that breaks after the warranty is void.
Anyway, Software Engineering most certainly is a research
discipline separate from CS and there is research and theory for
developing software at different cost levels.
Games are not bug free because that would be extremely expensive,
and cause massive delays in shipping which makes it impossible to
plan marketing. Games are less buggy when they reuse existing
frameworks, but that makes for less exciting designs.
