PS: oh and yes, like mentioned in the Martin Fowler article, when I
say of measuring and comparing productivity, I am thinking in terms of
business value, not lines of code (which would be incredibly silly).
On Wed, Jan 30, 2013 at 12:25 PM, Bruno Medeiros
<[address removed]> wrote:
> Just because productivity cannot be objectively measured, it doesn't mean
> one can't have a feel for his own productivity (a subjective, possibly
> non-liner, measure).
> In fact, I would say I suspect it's because of the fact productivity can't
> be objectively measured that Pair Programming is still considered in several
> circles to be a good thing (and debated in other circles, like this one). If
> you could measure it objectively, there would be no discussion, we would
> know for sure in which situation PP would be useful (if any, other than a
> few minor niche cases).
> You see, I remain highly skeptical of PP. The fact is, if you have two
> programmers, of roughly the same skill and talent, if they are doing PP,
> they would have to be *twice* as productive as each on their own. *Twice*!
> Can you imagine that?
> When I look at a days or weeks worth of development and work, I can't see
> how I could be twice as productive by merely having one other person helping
> me. A good way to think of this is roughly doing the same work in half the
> amount of time. To achieve this, there would need to be a combination of:
> writing all the code correctly at first so I write twice as fast (no
> refactoring, no redesign, not much time thinking); Or, never writing those
> bugs that take a long time to debug and fix. I don't see this happening with
> just the cooperation of another person. In fact, I would estimate the
> productivity gain would be something between 0% to 50%, depending on the
> task, and the other programmer.
> Maybe (and even this case it's just a maybe), if one is a developer that
> writes undisciplined, cowboy approach code when alone (badly structured and
> thought out code, no tests, no consideration for reuse or future changes),
> then the cooperation with another programmer might make it more structured
> and thought out, good enough to have a 100% productivity gain vs single
> developers. (But still, then the best approach would be if both developers
> would just be more disciplined on their own.)
> - Bruno
> On Tue, Jan 29, 2013 at 5:06 PM, Abraham Marín Pérez
> <[address removed]> wrote:
>> The key aspect of the subject could be in the sentence "feel that I am
>> much more productive working on my own". What do you mean by "productive"?
>> One could feel more productive because he/she produces more lines of code,
>> but what's the quality of that code?
>> I think pair programming, like productivity (see
>> http://martinfowl...), is a
>> qualitative aspect of programming, and hence depends on many personal
>> Abraham Marín Pérez
>> Blogs: http://www.gatogo...
>> Twitter: @AbrahamMarin
>> LinkedIn: http://lnkd.in/TT...
>> Sent from my iPhone
>> On 29 Jan 2013, at 09:40, Jim Collins <[address removed]> wrote:
> Bruno Medeiros