"Benjamin L. Russell"
Yahoo.com> wrote in message
news:a7fsb4pot97hges75ksn9fr0i8qumg1rbs@4ax.com...
[ SNIP ]> To my astonishment, I once read somewhere that the average number of
> programming language theory research papers that the average
> programmer reads in a month is zero. At first, I couldn't believe
> this. But having worked as a liaison between programmers and
> marketing staff, I think it is quite true.
If by programming language theory research papers you mean something like
http://cs.ioc.ee/~tarmo/papers/essence.pdf (The Essence of Dataflow
Programming) or
http://www.cs.brown.edu/~sk/Publications/Papers/Published/pcmkf-cont-from-gen-stack...
(Continuations from Generalized Stack Inspection), then no, you won't see
anyone except enthusiasts, computer scientists and some very few affected
implementers actually reading stuff like that, no.
I'm guessing (just from assorted conversations over the years) that good
programmers do read a greater or lesser amount of technical articles at an
application level...IOW, how do you actually apply a new technology. As you
know yourself, these can be quite involved (certainly not twinky
material)...for example, one of the better Haskell monad tutorials. Or if
you follow the J programming language programmers' mailing list that's a
good example of fairly technical but useful stuff.
You wouldn't really expect average (that includes good and very good)
programmers to read programming language theory research papers. They don't
exactly have much use for the working programmer. Is that the same thing as
saying that those academic papers are useless? No. But only a very small
group of people translate those academic papers into useful concepts for
developers.
It's besically the same as the difference between scientists, engineering
academics, working engineers, and engineering technologists. All of them
operate at their own levels of abstraction, and each group has people who
are capable of translating stuff from domain A into domain B. But most
working engineers won't be reading pure science research papers.
> Most of the programmers whom I worked with were not interested in
> programming theory, or even in programming per se, and spent most of
> their free time in the office watching giant centipedes eating mice on
> YouTube, chatting in Yahoo! Messenger, or sending e-mail. Once, I
> tried discussing the Towers of Hanoi problem with one of them, and he
> replied that it was "a very hard problem" in programming. I couldn't
> believe this. Towers of Hanoi is a first-year student problem for
> computer science students!
Again from assorted conversations over the years I'm guessing that good
programmers _are_ interested in programming, to the extent that they think
ahead, explore new languages and new ways of doing things, and look for ways
to improve what they do. In a decent shop where the developers are fully
engaged in their work (and I've been lucky enough to mostly work in those) I
have seen very few people watch YouTube!, use a chat program, or send
non-work email. Usually in your free time at the office you have your lunch
or coffee. Most people I know who do personal development don't do any of it
at the office.
As for the Towers of Hanoi example, well, I'm not that surprised. Bear in
mind, most programmers out there don't have a CS degree. The US Department
of Labor has for 2006 that almost 8 out of 10 programmers had an associate
degree, almost half had a bachelor's, and nearly 2 in 10 had an advanced
degree. The NSF found that the most common majors for graduates working as
software developers were engineering (35 percent) and CS (31 percent)...10
percent have a non-science/engineering degree. Very roughly speaking, then,
maybe 15 percent of software developers actually have a full degree in CS.
That's not to discount the value of the non-CS degrees, nor to over-inflate
the value of one, or even to denigrate the value of a person with no degree.
All of us have worked with CS grads that don't know shit, and conversely
worked with someone who has a 2-year diploma, or a degree in English, who
really knows their stuff. But the numbers do show that as a rule you can't
expect typical programmers to be deeply conversant with algorithms or data
structures or...well, lots of things.
> The problem seems to be one of lack of time and lack of interest. Most
> programmers seem to be force-fed programming technologies that they
> are not interested in, in a manner in which they are not allowed to
> explore what is interesting about the topic. Therefore, they learn to
> hate the topic, and proceed to spend all their free time trying to
> forget about programming.
Well, imagine working in a Web shop where you do nothing but PHP and CSS and
Javascript, with an occasional foray into trivial SQL. For a few years as a
junior developer you could probably stay interested, but eventually you'd
have to consider suicide, or do as you said - spend all your free time
forgetting about what you do at work.
[ SNIP]
AHS
