[MUD-Dev] Design iterations (Was: R & D)

Sean Kelly sean at hoth.ffwd.cx
Fri Jun 7 11:42:41 New Zealand Standard Time 2002

On Fri, 7 Jun 2002, Brian Bilek wrote:
> "Koster, Raph" wrote:
>> There is a general disdain among many game developers for
>> academically trained programmers usually because games are
>>   - technically not held to the same standard as other types of
>>   software
>>   - on much stricter schedules than code done for academic
>>   applications
>> ...but I have not generally seen a similar disdain for
>> programmers from the business world.

What about academically trained programmers who have since worked in
either the business world or the sciences?  Frankly, I think the
game industry's apparent disdain for established engineering
practices is somewhat foolish.  It's as if the perception that the
game is art somehow nullifies the need to produce quality software.

> Thank you for this, it explains quite a bit.  That bit about
> stricter schedules jogged my memory.  On a slight tangent - I
> actually followed the development of a recently released game,
> having a close friend at this particular game company (no names in
> case I'm giving away a trade secret, hehe).  One of the things
> which interested me is the way the development house treated
> 'milestones.'
> In the project management processes I am familiar with, a
> 'milestone' would be the completion of a major deliverable or a
> critical piece of code, with associated initial scheduling
> estimates.  Progress is tracked against those milestones.  Due to
> the nature of code development, schedules are not 'strict,' but
> rather the schedule is adjusted on a regular basis to reflect the
> project's true progress and updated estimates of work.  There are
> goals and objectives to be shot for, but a project manager would
> rather have a true understanding of where the project is at than
> making developers bust their butts trying to meet arbitrary
> deadlines.

Sounds good so far.
> However, this particular game company treated milestones as timed
> events - it was crunch time before a 'milestone,' everyone worked
> late and worked weekends trying to complete their bits of code,
> art, or design prior to reaching the 'milestone' date.  Is this
> just a word used in a different way, what I would call 'staged
> releases,' with those releases being held to a deadline?  Maybe
> this was unique to that company?  Or is this a difference between
> game companies and other software development groups that you've
> seen throughout the industry?

>From what I've seen, this practice of crunch-time is primarily
restricted to the gaming software world.  First, unrealistic
deadlines are generally established by marketing or management.
Second, it's the mark of a competent programmer to foresee not only
how long a particular task will take, but the potential issues that
may arise, and any potential political changes that may occur in
that timeframe.  That is, programmers typically have to think 6
months or more ahead as they work.  With the mercurial nature of
many decision-makers, the programmer should also be able to
anticipate any issues that may arise and plan accordingly.  Part of
that is by making easily maintainable software, the other is to
assess the importance of each task and to balance this with an
estimate of how likely that task is to change or be eliminated, and
when.  IMO competent programmers are as aware of the dynamics of
their work environment as the dynamics of their code.  The more
adept that they are at this, the less their efforts will be wasted
on things that hindsight may prove unnecessary.

So I think the issue comes down to a matter of poor planning (which
hearkens back to lack of established engineering practices) and an
overzealous marketing department.  Other companies may have crunch
time at times, but in my experience this is generally the result of
last-minute changes that the programmers did not anticipate or the
discovery of bugs that must be fixed as soon as possible.

>> There is a general disdain for managers, marketers, execs from
>> other businesses because they tend not to be knowledgeable about
>> game-specific management issues (particularly the need for
>> iteration, but also market issues, etc), but these are simply
>> areas where a good manager will learn the ropes. So I view this
>> as a shortsighted position. There definitely have been many
>> mistakes made by executives new to the industry who try to treat
>> it like the movies or like baked goods. But they can get better
>> at it.

The comparison to movies does somewhat hold -- it's still an
entertainment industry.  And Hollywood is slowly learning that
following the three-ring-binder of movie design does not gurantee
that the movie will be a success.  Further, the past few years have
seen a shift from blockbusters towards more independent films, in
terms of audience preference.  It seems there have been similar
issues in the game industry.

> When it comes to business types, why do you think the disdain
> exists?  I understand what you are saying, but it seems to me that
> adjustments have to be made when a manager moves from any industry
> to any other, and each industry has its own unique problems.  I
> have rarely, if ever, seen an industry with so much disdain for
> outside managers.

Outside anyone, for that matter.  It's like the gaming industry is
an underground rock movement that holds all the "sellouts" in
contempt, then wonders why it is not more successful.  I grant that
things are changing (signified by the focus shift in "Game Developer
Magazine") but it's taking a while for the ideas to seep in.
> A friend of mine who recently joined SOE suggested to me that it
> may be because a lot of the employees in the industry like the
> very laid-back atmosphere of a game company, and fear that a more
> 'business' oriented manager would want to whip the employees back
> into 'line.'  The sort of hard management that many people joined
> the industry to escape.  Do you think this is close to the mark?

There is probably a fear of being turned into Dilbert's office, but
IMO this is misplaced.  A work environemnt is the product of the
people that create it.  If you hire someone who turns out to be a
pointy-haired boss, find someone else who works better with the
team.  But in my experience, a more structured environment gives the
people that work there a more regular schedule and more resulting
free time.  If this happened, the game industry might see more
competent professionals willing to work at game companies, as they
would not be trading away their lives as a result.  Of course the
salary level would also have to become more competitive as new hires
might not necessarily be straight out of highschool with no leverage
to speak of.  This would in turn decrease profit margins, etc.

Gone are the days of garage game shops.  When will the business
practices follow suit?


MUD-Dev mailing list
MUD-Dev at kanga.nu

More information about the MUD-Dev mailing list