NYCPHP Meetup

[Paul A Houle]

gary066 at yahoo.com wrote:
> A lot of the 'high paying' jobs are going unfilled because they are in reality a big pay cut from consulting, not due to a lack of people. It's the free market at work.  I have run my own consulting company for the past 5 years - generally we have anywhere between 2 and 10 consultants on projects depending on business cycle;  I personally have two CS degrees from Carnegie Mellon and 15 years experience in Java/J2EE/SOA.  I have been an employee for 9 of those years, a consultant for 6 (I still bill myself out). The 150K base positions generally result in 60 hour work weeks; no comparison when compared to rates that are in the 90-125hr range for every hour.  I have a client in another state with a high level architect position paying 150-175 base.  Getting resumes has been extremely difficult - the problem is not a lack of interested people, the problem is they require full time status.  That kills the interest.
>
> I have dealt with a lot of PHPers.  I've never had one ask for $95.  They know their value in general.
>
> Unions are not the answer (I don't see outsourcing as a problem).  Unions are for industries where there is little differentiation in productivity.  The difference in productivity between a good and bad developer is orders of magnitude.
>
>   
    I used to believe that thing about developer productivity,  but now 
I don't.

    The environment in which you are working has a big effect on your 
productivity.  If you're developing command-line batch processing 
scripts,  and the requirements are well known,  you can be 10x more 
productive than you can be in a project where you get useless 
requirements documents,  need to argue with your boss to get to talk to 
the customers to revisit the spec,  then leave ten phone messages for 
the customers,  finally get to talk to them,  have a good conversation 
about the specs -- then they call you back an hour later and you need to 
explain the business process model that you worked out again.  Then 
there's the week it takes to go back and forth with the GUI designer,  etc.

    Early studies in software engineering showed that "coding" is about 
1/5 of the time it takes to design an app:  you've also got testing, 
deployment,  requirements work, and design.  Even if you could eliminate 
the coding time to zero with some supercompiler or by hiring a 
superdeveloper,  you're still left with 4/5 of the time consumed by 
other activities.

    We've all encountered "superproductive" developers whose MO is to do 
the 20% of the work that gets 80% of the results,  and then split town.  
A developer who's got a lot more skill and really understands what it 
takes to get things into production ends up looking like a plodding 
idiot in comparison,  because they're the ones who have to normalize the 
database tables,  add the "htmlspecialchars()" so that the app isn't 
busted when somebody types a quote into a form field,  deal with 
cross-browser issues and all the other stuff that takes a lot of time.  
The "plodder" knows perfectly well that the app would have been better 
off it building it in a framework,  but got involved too late in the 
game to really make that move.