[plt-scheme] Re: Programming for non-programmers

Rohan NIcholls <rohan.nicholls at pareto.nl> writes:

>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-scheme
>
> On Thu, 14 Oct 2004 11:44:17 -0500
> Robby Findler <robby at cs.uchicago.edu> wrote:
>
>> I suppose it's a theoretical worry to some (esp. until Nov...) but
>> anecdotally the information I've come across seems to suggest that
>> outsourcing overseas doesn't work very well at all.
>
> I have first hand experience with the outsourcing disaster, and it was
> not pretty. You are absolutely right, I don't know if the project was
> typical, but it was for a big bank, who had hired a development
> company that worked with "virtual teams" ie. a couple of people on
> site who then communicated with a programming sweat shop in India.

I have heard several similar anecdotes of outsourcing disasters.

The management of one company I worked for was trying to persuade the
engineering dept. to find some things to outsource.  They didn't
expect to outsource anything important, but they were hoping that we
could find enough scutwork to make it worthwhile.  I wouldn't have
minding outsourcing a bunch of non-critical code.  If you *have* a
million monkeys and you don't exactly need Shakespeare, you can apply
the `Monte-carlo' method.  One one-thousandth of the productivity at
one ten-thousandth of the cost is a gain if you have the time.

But we weren't stupid.  Credulous, perhaps, but not stupid.  Before
committing anything to anyone, we asked for code samples from some of
their `star' programmers.  Nasty stuff, indeed.  It was literally
nauseating to read.  Fortunately, all we wasted was a few hours of
time; no money changed hands.

After that experience, I have become completely unconcerned about
software outsourcing.  Yes, it may be the case that a lot of people
who write VBScript kludges will find it hard to compete against their
counterparts in the third world, but there is no reason whatsoever to
believe that *serious* product development will migrate overseas. 

> This also was a major problem, not only for obvious language reasons,
> but even more because of communication bugs due to cultural
> differences.  One example is that the word "no" has a different
> meaning and weight in different cultures, as well as the unwillingness
> to disappoint (in the indian subcontinent this tends to be much
> stronger than in the west), and the list goes on.

A friend of mine who manages the Boston office of a multinational
software support firm mentioned this exact problem.  They specialize
in critical IT support --- the kind where someone will, if necessary,
fly to the customer's site at 4:00 a.m., replace the hardware,
hand-patch the software, or even rewire the local Telecom office to
get things back on line.  Of course, 99.9% of the work can be handled
by Perl scripts running on a server farm, and the management has been
outsourcing a lot of that.

His biggest problem is in getting simple `yes or no' answers from
subordinates who have been raised in a culture where one *never*
*ever* says `no' to a superior, even if it is a neutral question about
an observable fact.  He spends a lot of time telling people that he
is not passing judgements on their ability but simply attempting to
gather information.  He wants to know if he can depend upon a task
being finished at some time, or whether he should start planning
alternatives, but invariably he is told ``Yes, sir.  It *will* be
finished.''  

This isn't an attempt to prevaricate.  For them, it would be
considered unspeakably rude to answer `no'.  (Sort of like answering
yes to `Does this dress make me look fat?')  But Westerners (and
Americans in particular, and *engineers* in the extreme case) have a
cultural bias that places the value of truth above that of social
class.