In article <1992Aug31.195540.13074@ctr.columbia.edu>, shenkin@avogadro.barnard.columbia.edu (Peter S. Shenkin) writes:
> In article <1992Aug31.170849.11927@mprgate.mpr.ca> mcvey@mpr.ca (Iain McVey) writes:
>
> >'Computing Science types' who truly understand accounting should write
> >better accounting packages, shouldn't they?
>
> Well, there's variation person-to-person within all disciplines, and we've
> heard scientists and "computer science types" comment on how well, or
> poorly, scientists write code, but in my experience the implicit notion that
> "computer science types" are automatically good programmers is false.
>
> In many or most CS departments, computer science has more to do with
> proving theorems than writing code, and the experience obtained by
> students getting degrees, at whatever level, in these departments
> reflects the leanings of the department.
>
> I've recently concluded that expecting a computer science graduate to
> be a good programmer is like expecting an English major to be a good
> writer. Both expectations are, unfortunately, false.
>
> No flames intended ....
I fully agree with you.To draw an analogy from Mech engg. a good mechanical
engineer need not necessarily be a good machinist.He can design a good product
and leave the implementation on a lathe or a milling machine or what have you to a skilled machinist.Similarly a good architect or civil engineer need not be a
good mason.So too with Computer Science,computer engineering and programming.I
have come across software engineers who are excellent in designing compilers,
operating systems but are lousy programmers.In fact some of the very good
programmers are B.S. in maths or physics.No amount of language design and research
can make good programmers of every one.As long as there is flexibility there will
be always a choice in the use of code and data structures and the way they are put
together and this will remain an art which only a few will master.
