Are most programmers male?
Frithiof Andreas Jensen
frithiof.jensen at removethis.ted.ericsson.dk
Tue Aug 13 12:58:59 CEST 2002
"Opus" <opus at value.net> wrote in message
news:mailman.1029143358.29003.python-list at python.org...
> Or quite the opposite. I have found that the ones that don't
> document their code are the more social engineers.
All of the above behaviours would mean that the organisation floats and
there is lack of control with the development and the staff - by management
and by quality assurance - and that the product probably suck too!
If somebody does not willingly do the job, be it to document or unit test,
that somebody can get a boot in his/her arse!
> Most of the hermits know how to test their code.
Yep - THEIR(!) code - not the organisations, the customers or, god forbid,
fellow programmers - which are all incompetent, blithering fools (We are
talking the true, Real Programmer, here ;-).
> to speak, you should try being nice to one sometime and learn
I have actually experienced a few pathological cases, where all you would
learn would be a litany of abuse ;-)
...But of course one does talk to people...even politely.
> But that decaying moose might just be unique in the organisation in
> that it knows the whole system. How many engineers know the workings
> of the whole system in a large one?
Again a Management Problem:
The majority of the value of most software businesses lies in the knowledge
the organistation creates while building it's products - it is totally
ludicrous to rely on a trusted few to preserve the "corporate knowlede
capital", people may leave, get ill and even die (Quite apart from creating
a scalability problem when the same trusted few has to answer numerous
questions from hordes of newly hired developers).
Hence critical information such as architecture and design must be
documented and distributed within the organistation, backup if you like.
Yet management often belives that the product is the only value produced and
allows the business to be put - Microsoft is not worth what it is today
because it wrote Windows and Office, it's value is high because it can
produce *new* designs of it's software based on the knowledge inside the
I have seen quite a few development projects in both hardware and software
design fail utterly because of overreliance on a single "expert", where the
expert either is wrong from the beginning and the fault is never caught
because he works in isolation or the guy leaves halfway through the project,
leaving no discernible clues on what his ideas were.
So, to summarise:
I do not dispute that there are key persons that essentially drives all the
work and that they are neccessary - My problem is that allowing them to
become Hermits is suicidal for the project/organisation; and in any event I
would prefer a Guru to the Hermit given the choice.
More information about the Python-list