[Courses] [Careers] Almut's meanderings among electronics, psychology and programming

[Almut]

On Fri, Feb 18, 2005 at 09:21:29AM +1100, Mary wrote:
> On Mon, Feb 14, 2005, Almut wrote:
> > So, I ended up maintaining what grew into a complete data evaluation
> > package -- I'd always been willingly integrating ideas and wishes
> > people had uttered.
> 
> Do you know if the package is still around? Was this before or after
> Free Software got mindshare? Did you ever consider releasing it? :)

Actually, I don't know what became of it.  The situation was the
following:  I had a passage in my contract saying that everything I
produce during work hours automatically becomes the property of my
employer (not too unusual, I think).  And so, nothing was keeping them
from licensing the package to other research institutes -- for rather
high fees.  Of course, I'd been investing quite a lot of volunteer
efforts and private time in addition to the regular work hours.
To honor this fact, I had come to an agreement with the institute that
I get 25% of the profit made through selling the licenses.

In short, I have to confess it was not only closed source, but I also
had gone to great lengths to develop sophisticated copy protection
and licensing mechanisms to enforce "proper" usage.  Although this
was undoubtedly interesting from a purely technical perspective, in
hindsight, I'm no longer particularly proud of those endeavours...
(but - like a converted non-smoker - I have a clearer picture of what
to fight against now ;)

Later, after having converted to the Free Software point of view, I had
played with the idea of releasing it as open source, but didn't really
find a good handle on making it happen.  When I left the institute, I
had handed on the package to a new programmer they had hired for that.
As I wanted to have my mind free to fully concentrate on my new work,
I gave up all my rights as the original author -- no rights, no
obligations.  About two years later, I tried to contact that guy out
of plain curiosity, but he had already left the institute for a better
paid job in the industry...  That was 5 years ago -- and in the
meantime, I've almost completely lost contact.

> 
> What kind of psychologists get to stay in one place? Only the kind who
> do clinical work?

I think it's mainly the scientific area that's subject to this
always-on-the-move problem.  Actually, I think it's not confined to
psychology.  At least over here, it seems to be a more general issue
with working in research -- but I can't tell for sure.  The declared
rationale is that the few available positions are not being clogged up
by people sitting there permanently, getting rusty.  In other words,
it's meant to keep "fresh blood" and new ideas flowing into science,
and foster exchange of expertise among research groups.
(OTOH - though generally a good idea - if people leave science just
because of that, one might presume an inherent flaw in the strategy.)

In clinical psychology, and in the area of work and organisational
psychology (covering things like human resources, public relations
work, advertising and marketing, etc.), I guess it's not such a
fundamental problem to get a long-term position (well, if you find one
at all).  In clinical work, though, it can still be a rather long path
before you may settle down and do psychological counseling or offer
similar services. (After Uni you'd typically have to complete a number
of additional specialised expensive trainings -- without which you're
not entitled to do the work that you could pay the trainings from...)

> 
> What counts as medium in this area? How many people would work on a
> medium project and how long does it take to get it done?
> 
> Also, does your company have an ongoing relationship with clients, or do
> you go in, build a workflow, and that's the end of the relationship?

It varies between roughly 100 and 3000 hours accounted for the whole
project.  I usually work on my own, or together with one or two of my
co-workers.  On the client side, about the same number of people are
involved (those I'm in direct contact with, not counting all potential
users of some application).

The shorter projects run for about a month, while others may extend
over a period of several years, with intermittend breaks of varying
lengths. Those are the ones where adaptations have to be made from time
to time, because the requirements keep changing (new tools, new data
formats, previous workarounds no longer needed, modified workflow...).
Some never get finished officially, but end in some support contract
(fix number of hours per time period) to cover whatever future change
requests might come up.

(From a software design perspective, those moving target projects are
the most problematic ones, because sometimes, after two years or so,
requirements no longer have any resemblance whatsoever to what had been
requested originally -- i.e. internal program/class/object structure no
longer matches current reality.  I haven't yet found an optimal way to
deal with this mess^h^h^h^h kind of challenge...)

I typically work on 3-4 projects in parallel.  As my brain is not
as good at task switching as current CPUs / OSes, I always try to
continuously work on one task for a few days before I switch to the
next.  My scheduling policies are not very elaborated -- those who
shout the loudest or most often usually get assigned a slot :)
(BTW, they don't shout literally... most of my clients in fact are
very nice and polite.)

With smaller projects, even end users mostly call me directly when
they encounter problems.  This guarantees them more or less immediate
handling, but it can be rather distracting for me, if it happens five
times a day, and I'm supposed to concentrate on a completely different
project.  So, for longer running projects, my company attempts to set
up some more formalised handling of support calls.  

Along similar lines, efforts are being made to migrate older projects
away from senior developers, so they get free to acquire and launch
new projects.  Thus, part of my work involves inner house transfer of
projects and specialised knowledge to newly employed collegues or to
wherever capacity has become available.

Transitions like these are not always as easy as it might seem, once
your clients have gotten used to being in direct contact with you as
the chief developer.  The latter seems to be of paramount importance
to quite a few, meaning that, often, you're in for a tightrope walk,
delicately balancing your own and your client's interests.  Ideally,
they'd like to have someone they know and can trust to be there to
discuss and take care of whatever problem they need solved -- as soon
as possible, of course.  Equally, they honor it, if you proactively
invest some thoughts into improving their overall work experience and
efficiency.  Most of these things only do happen, if you closely work
together for a longer period of time.

In fact, we have an ongoing relationship with most of our clients.
One-time cooperations are typically with clients who have a decent
working knowledge of programming themselves.  They want professional
assistance mainly to get them started with a well designed framework.
Once you have delivered a working prototype (with source code), they
are able to make minor changes and extensions on their own.

Some would prefer if they could "buy" you exclusively...  Occasionally,
things even get a bit tricky to handle, if you're working for several
mutually competing departments within the same company  (it may sound
funny, but some appear to get kind of jealous when they find out).

OK, I realise I'm starting to ramble, so I'll just leave it at that...

> 
> How many women are there among the engineering clients?

I don't have any precise numbers, but I'd estimate around 3%.  In any
case less than 5% at the places I got to look around -- not counting
secretary and purely administrative work. (Not all that positive, but
still better than at the time I studied electrical engineering, where
there were just 0.5% women...)

On a side note:  for some strange reason, in business relationships
I currently get along better with men.  I'm not sure why this is.  I
tried to get into contact with some of the women to identify some way
to cooperate, but it never has worked out so far.  I figured it might
have to do with some kind of envy or misunderstood competitiveness.
It would be a real pity if this was true, i.e. if we women are making
life harder for each other than it need be.  If anyone has any ideas,
let me know.  (In case you think "well, no surprise, it's because you
come across like the biggest <negative word of your choice> ever", I'd
still be interested in your input, but then please tell me off-list ;)

> 
> Did you ever wish you did have an expert around to guide you (or to
> compete with)? Do you have one in your current work?

I'd say I never really felt I was missing an expert -- as long as there
were good books around.  At least not with respect to having a teacher.
Even when there were no docs around, having to find out things myself
by experimentation or trial and error, made me rather creative and
trained my heuristic competences...
Maybe some guidance might have been beneficial, in the sense of having
orientation or advice on what is essential to look into, or what is
the Right Way to do things.

Generally, I think I'm the prototypical autodidact. I learn best, if I
can determine for myself exactly how much time to assign to individual
subtasks.  I guess my learning habits and preferred speed deviate from
the average.  While getting into something new, I can't help making
lots of connections in my head to other stuff I already know.  So, my
first time around with new material, this takes its time, but it pays
off in the end, because I do remember rather well, and, by having
"drifted off" previously, I often have already solved questions along
the way, that have not even occurred to other people, yet :)  
Admittedly, this technique requires sufficient motivation, so it works
less well, if something doesn't interest me really...

In my current workplace, there are many experts around, specialising in
all kinds of topics.  This is very convenient, if you need quick help.
Also, talking to them, or simply listening to them getting into geeky
conversations at lunch or so, serves as a perfect general guidance in
the long run...  So, in short: although I was never missing experts
before, it feels good to have them around now.

> 
> > [1] Not sure whether "tutorial" is the correct English term. (...)
> 
> That would be called a "tutorial" in Australian universities. However,
> there are tutorials all the way through undergraduate. In more senior
> tutorials it might be more of a discussion between equals.

Thanks for confirming. (My notion of "tutorial" centered around written
introductions, howtos, and the like...)

Almut