I am agnostic about pair programming. I believe that having strong programmers and a good code review process offers at least as much benefit as (if not more than) full-fledged pair programming, but if it is working for someone, I don’t see any reason to object. Researching which brought me to Obie Fernandez’s strange self-congratulatory article on pair programming.
Apart from the evidence-less accusations about what “most” software developers, managers and firms seem to do, it also has some silly ideas about hiring good developers. Read:
The best way to hire talent is to drop them in actual work situations for at least a week and see what happens. […] Our candidates spend their interview week rotating on actual production code, pairing with the same people they’ll be working with if they’re hired. Everyone gets a say in whether to hire a new recruit, and hesitance from a developer that actually paired with them means they do not get hired.
I don’t understand in what universe a talented programmer is going to take a week’s worth of leave to work as part of an interviewing process for an obscure company. Maybe a desperate, unemployed programmer? Since the best programmers are already attracted and hired by the big firms (Google, Microsoft, Facebook, etc.), is this any way to recruit them away?
This is simply a case of “don’t know what is missing“. The company simply doesn’t realize how drastically it has shrunk the pool from which it can hire and assumes that it has the best possible developers. There is some benefit to doing your homework before hiring someone, but this is a practice that has gone too far into the other direction.