[Techtalk] interpreted vs. compiled languages

Miriam English mim at miriam-english.org
Sat Jun 13 23:12:46 UTC 2009


The way I see it is as a balance between speed of writing and running 
speed. A compiled language can be difficult to write and  maintain 
because the source can be lengthy and always involves you in that darn 
write-compile-debug cycle, whereas an interpreted language can be much 
easier to maintain because of the succinctness of the code (python is a 
great example of this) and because minor alterations let you instantly 
try them out.

For pure running speed I have a feeling no interpreted language can ever 
perform as fast as a compiled language, but often ignored is that 
end-use is only part of the requirements for efficiency. Writing 
something in raw assembly language would be fastest of all, but might 
take months to create when you might take a week to knock out something 
  in perl that performed only half a second slower. And the perl program 
  might take minutes to alter, where assembler might take days.

Ultimately even interpreted languages use compiled libraries. The trick 
is in using the interpreted parts where speed doesn't matter, and the 
libraries of compiled code for inner loops and other places where speed 
is crucial. So the question becomes how you decide to mix compiled and 
interpreted parts to best effect.

On difficulties of using the command line, I think the problem lies 
mainly in the alienating way that "man" pages have traditionally been 
written. They seem clear when you are used to them, but in fact they are 
extremely difficult for newbies. They should have many examples near the 
top of the page and be written in less intimidating bureaucratese ("the 
blah command shall take..." instead of "blah uses"). I have been trying 
to gradually rewrite some documentation for shell commands for myself. 
It is slow because I have too much else to do. I never thought of making 
my stuff available on the net. I'll clean it up and do that.

Cheers,

	- Miriam

-- 
If you don't have any failures then you're not trying hard enough.
  - Dr. Charles Elachi, director of NASA's Jet Propulsion Laboratory
-----
Website: http://miriam-english.org
Blog: http://miriam_e.livejournal.com



More information about the Techtalk mailing list