[nylug-talk] Popularity of Programming Languages

[Pete]

--- mbac at netgraft.com wrote:

> 
> On Thu, 11 May 2006 9:37 am, Michael Bacarella
> wrote:
> > On Thu, May 11, 2006 at 01:33:57AM -0400, Monjay
> Settro wrote:
> >>  How are you gauging the use of popular
> programming languages? Does 
> >> your method and observations coincide with
> Tiobe's?
> >>
> >>  http://www.tiobe.com/tpci.htm
> >
> > I'm kind of stunned that decades later, C is still
> very relevant.
> 
> Let me expand now that I've got some time on the
> subway to kill.
> 
> Is C's strong presence simply the weight of the huge
> body of existing 
> code plus maintenance traffic?  Is anyone developing
> new applications 
> today exclusively in C?  If so, why?
> 
> I still write C code, but it is always small pieces,
> in situations where 
> we tried to do it in Java or Python and found that
> we were CPU 
> bottlenecked.  Interestingly enough, because of how
> exceptional these 
> cases are, the usual anti-C rhetoric is irrelevant.
> ;)

some applications are not amenable to the object
paradigm.  for those cases, C provides a faster and
cleaner implementation than C++.

for example, for my phd dissertation, i had to solve a
particularly nasty PDE.  there's really no reason to
cast the solution of a differential equation into
objects.  there's simply no benefit.

numerical computation in general is fairly straight
forward.  the fact that you _can_ write, say, an
eigenvalue/eigenvector solver in C++ is simply the
statement that if you push hard enough, square pegs
really do fit into round holes.

btw, bloomberg (the finance company) is noted to be
mainly a "C house".  most other financial institutions
are "C++ houses".  there's a couple of "java houses"
in the finance community, but not many.

btw, i thought the original article was absurd.  how
could anyone compare the popularity of C/C++/Java to
Javascript or Perl?   to make a popularity comparisons
rely on the issue of choice -- whether you _can_, not
_would_, choose one language over another.

to write a PDE solver in perl, or to write a program
that goes through a text file and reformats data in C,
is simply absurd.  the two languages are fairly
orthogonal to each other.  to compare the popularity
of one to another is vaccuous.

that said, looking at the number of jobs being offered
that require C versus the number of jobs being offered
that require perl IS a valid metric.   but popularity
as a metric?  bleah!

pete

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com