Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uwm.edu!linac,att!ucbvax!dog.ee.lbl.gov!elf.ee.lbl.gov!torek
From: torek@elf.ee.lbl.gov (Chris Torek)
Newsgroups: comp.misc
Subject: Re: WANTED: "C" code line counter program
Message-ID: <11078@dog.ee.lbl.gov>
Date: 19 Mar 91 00:40:04 GMT
References: <1991Mar6.214157.18633@ntpal.uucp> <2628@p4tustin.UUCP>
Reply-To: torek@elf.ee.lbl.gov (Chris Torek)
Organization: Lawrence Berkeley Laboratory, Berkeley
Lines: 25
X-Local-Date: Mon, 18 Mar 91 16:40:04 PST

Although these observations have been made before, I will make them
again.

Code `productivity' metrics are hard to define, but line counts are
not even a good approximation.

>Once you have the comments removed, you can use the wc program ...

If you remove comments, such a line count would call a day wasted if
it were spent commenting some code so that it could be maintained.

Whether or not you remove comments, deleting lines is not necessarily
`negative productivity'.  For instance, in order to make TCP work over
actual (lossy) networks, you need good estimates of the round trip time
and its variance.  Van Jacobson did this for Berkeley Unix; when he was
done, the final code was several lines shorter, but it was much faster
over real Internet connections.  He has recently removed even more
lines (and added just a few in a few critical places) and obtained
large performance improvements.

What is really needed is a way to measure the amount of `useful thought'
put in, rather than the number of characters typed.
-- 
In-Real-Life: Chris Torek, Lawrence Berkeley Lab CSE/EE (+1 415 486 5427)
Berkeley, CA		Domain:	torek@ee.lbl.gov