Recoding pros and cons (was: RE: [thelist] RE: PHP 'proof of conc ept' virus developed)

Steve Cook sck at
Thu Jan 11 02:43:28 CST 2001

While David's comment has a lot of truth, it's not always detrimental to
code in a way that encourages rewriting.

Well, that's a controversial statement, but I think I can back it up! Of
course project planning is a "Good Thing(tm)", but it's not always possible
to plan in such a way that perfect code will be written from the word go. I
remember reading about a study that looked into a way of coding that
encouraged programmers to code a solution to a problem and then throw it
away and recode for inclusion in a finished application. (Unfortunately I
have no supporting links for this).

They found that as long as this was planned into the project from the
beginning, it led to better code quality at the cost of a slightly increased
development cycle. The development time was only slightly increased because
time was saved from the end of the project in reduced bug chasing.

I've personally found that in certain applications that I've repeated
several times, I've ended up finding new and much more efficient ways to
code them. However this does require at least as much forward planning as
any other programming project.


   WapWarp -
 Wap-Dev -
 Cookstour -

> -----Original Message-----
> From: David Wagner [mailto:dave at]
> Sent: den 11 januari 2001 08:35
> To: thelist at
> Subject: Re: [thelist] RE: PHP 'proof of concept' virus developed
> > Experienced jumpers pack their own parachutes for the same reason 
> > experienced developers write their own code.  ;-)
> I'd like to point out, for those coders who are considering skydiving,
> that one should NOT attempt to re-pack a chute on the fly. :)
> Come to think of it, this is a good rule for coding, too... 
> write it well
> the first time, or prepare to hit the ground hard. (This is why I've
> stopped coding.)

More information about the thelist mailing list