Re: OT: Lines of code per day...

There are many times where code has to be completely reworked based upon new
information or new ideas for how it should be done. This is extremely common
in OOP coding where we 'refactor' on a regular basis. So counting number of
lines doesn't do much good since you are counting work that was done before
but now being done in fewer (hopefully) lines. That is merely 1 example.

There are loads of reasons why lines coded is terrible. Despite the desire
to put a measurement on programmers somehow, I haven't seen an effective way
yet other than just results.

Alan

Alan Churchill
Savian "Bridging SAS and Microsoft Technologies"
www.savian.net

From: ben.powell@CLA.CO.UK
Reply-To: ben.powell@CLA.CO.UK
To: SAS-L@LISTSERV.UGA.EDU
Subject: OT: Lines of code per day...
Date: Tue, 9 May 2006 11:15:56 -0400

Dear SAS-L

This is wholly inappropriate benchmark for performance and I fully expect to
be flamed, but what are your sage thoughts as to a good amount of code to
produce in a day....? Bear in mind this could be good/bad/average code as
befits the reputation of the programmer, and indeed the style of
programming.

For instance a weak programmer with a penchant for copying and pasting
irrelvent comments might easily produce 250 lines of code in a morning and
yet still be a weak programmer. Whereas in solving a hashing algorithm a
strong programmer could be satisfied with a meagre 50 lines of code...

Apparently it took 4 months for LT to code Linux to a point where Bash would
run, and this took 10,000 lines of code, which at 20 days per month would be
125 lines per day... Which is to say the average is probably most useful.

Rgds.