Try to write the following Fortran 77 program (note the
6 space indentation) in Ruby so that it looks as close
to the Fortran as possible without extending Ruby:
program ii
integer i
do i=0,10
print*,i
if (i.gt.5) goto 1
enddo
1 print*,i*i
end
Here’s the Python candidate to date:
for i in range(10):
#begin
print i
if i>5: break
#end
print i*i
Better Python candidate…
from operator import gt
class goto1: pass
try:
for i in range(0,10):
print i
if gt(i,5): raise goto1
#endfor
except goto1: pass
finally:
print i*i
#end
But why is anyone trying to convince a fortran programmer to use a slow,
dynamic language? This reminds me: Too many of the language debates
concentrate on static vs. dynamic typing. The real issue is type
declarations. They waste time during proto-typing. But there is another
way
to eliminate those pesky declarations: type-inferencing.
People are spending a lot of time trying to “optimize” Ruby, Python, and
Perl. They’re all hitting the same problem: dynamic typing. It’s
insurmountable. I was really hoping that Ruby could be THE language, but
last night, browsing the Great Language Shootout page, I noticed a
problem:
Ruby is SLOW! Even compared to Python.
Boo and Ocaml are examples of fast, agile languages, and if more
companies
would adopt them, I’d switch in a heartbeat.