Books for the NOC guys...

Nick Hilliard nick at foobar.org
Fri Apr 2 17:43:29 UTC 2010


On 02/04/2010 14:39, Valdis.Kletnieks at vt.edu wrote:
> On Fri, 02 Apr 2010 13:48:48 BST, Michael Dillon said:
>>> So, what are you having your up-and-coming NOC staff read?
>>
>> In an attempt to wean them off of unmanageable PERL scripts
> 
> There is not, and there never will be, a useful programming language that
> makes it the least bit difficult to write totally abominable creeping-horror
> unmaintainable code in.

+n

"Whatever language you write in, your task as a programmer is to do the
best you can with the tools at hand. A good programmer can overcome a poor
language or a clumsy operating system, but even a great programming
environment will not rescue a bad programmer". (Kernighan and Pike)

Although language zealots are loth to admit it, certain languages are
better suited to some things than to others.  Perl's syntactical support
for text processing are second to none, but for web stuff, it's hideous.
PHP stinks on the command line and text processing, yet its web integration
makes it a good candidate for small web projects.  Shell scripts are
designed specifically for command line tool management, pipes and all that
sort of thing, and just because other languages support popen() and system,
that doesn't necessarily make them good choices for stuff which involves
unix scripting.  I write readable and maintainable code in all three.

Java elicits strong reactions.  No doubt, you can use Java plumbing
libraries to scale to impressively large systems. On the other hand, Cisco
Configuration Professional (the new SDM) provides an excellent example of
how badly you can screw up a good idea by using the wrong tool - I'm not
interesting in using or recommending a desktop tool which takes 2 minutes
to start on a fast computer.

You can write unimaginably awful code in python and ruby, and irrtoolset's
code is a prime example of what you can do with c++.  Yet, we all
acknowledge that python, ruby and c++ are high quality languages.

In short: less zealotry, more pragmatism and a realisation that each
language has its own strengths and weaknesses.  Bad code is bad code in any
language.

Nick




More information about the NANOG mailing list