Re: Part 2: System Structure

[Michal Suchanek]

On 5/17/06, Marcus Brinkmann <address@hidden> wrote:
> At Wed, 17 May 2006 11:47:44 +0200,
> "Michal Suchanek" <address@hidden> wrote:

> Or, to put it the other way round: Being able to cheat _ensures_ that
> I can always play for entertainment, and not because I want to achieve
> something else, because it leaves me the option to skip parts of the
> activity that I find tedious.  So, cheating in this case helps a lot
> to prevent or contain addiction.

This is an interesting view. But I am not sure that cheating is that important.
And realiable services that should logically run on user resources are
not limited to games or competitions.

Maybe a more general question should be asked: Is it feasible to write
a service that runs on user resources and is simple and reliable? What
properties will such service have? Which use cases it can support and
which not?

For one, how do you implement ping? If the user runs ping in his own
session it is not restricted at all. He can modify it in any way he
wishes. If ping is a service that uses a constant amount of resources
there are only so many addresses it can ping at once. It would be
subject to DoS. How do you receive the pong? Does the ping service do
it or do you receive it through a different service? What if you do
another ping before you receive the pong? How do you make sure that
the user does not fill the network connection with pings (ie enforce
some minimum time interval between pings)?

Note that a constructor that guarantees that a ping program cannot be
modified solves this neatly. You can spawn as many pings as you wish.
They will use some constant place to store the time of last ping to
enforce the interval, and receive the pong for you. They will last
until pong arrives or until you get bored and kill them.

If you have some plan for network stack that makes writing ping simple
I would like to hear about it. Sure, if the accounting of bandwidth
and handling of replies is done in the networking stack the only
priviledge the service provides would be the injecting of the packet
on your behalf. But then the networking service has to solve a more
general problem than ping had.

How would device drivers transfer data? I guess that the device can
handle only finite number of buffers (or it will be limited for the
purpose of writing the driver). The driver could pay them and only
exchange pages with clients. Suboptimal (ie the driver would keep the
pages allocated even while unused) but doable. I guess it is pretty
much the way it is done now.

> > > What you say is correct.  But competitions have others affect as well.
> > > They nurture and exploit addictive strains in people, and in fact they
> > > frequently kill people.  Without even investigating into these issues,
> > > I know from news reports about one person killed by playing Starcraft
> > > and one person killed by playing Everquest (another committed
> > > suicide).  There has also been at least one murder over online game
> > > "property."  Online game addiction is widely recognized now, and the
> > > social cost, carried by society, is enormous, in terms of health care,
> > > broken up relationships, neglected children etc.  I want to coin the
> > > term "digital drugs" for these types of infrastructure (Everquest was
> > > coined "Evercrack" by its players).
> > >
> > > So, no, it is not at all obvious to me that we should support this.
> > > In fact, I am quite convinced that personally, I do not want to
> > > support this.
> >
> > The fact that people may get addicted to something does not mean we
> > should not support it.
>
> But this is not my argument.

What is your argument then? I got the impression that exactly this is
(one of) your argument.

> > People can get addicted to wide variety of
> > things including sweets, cocolade, coffee, work, all kinds of games
> > (computer or not), adrenaline sports, ...
> >
> > Banning all such things would be very restrictive and ineffective.
> > There are many drugs that are actually banned, and people still get
> > addicted to them. Yet people who would make good use of them cannot do
> > so legally.
>
> Yes, this is a problem.  It can be solved by legalization and
> monitoring and control.  However, the monitoring and control part is
> crucial.

hmm, this is getting way off topic. But I will try to answer anyway.

Well, my general idea is that what is illegal is unmonitored and
uncontrolled by definition. You may fail monitoring and controlling
but prohibition is obviously not a substitute for either.

To put it other way, it is the people that get addicted what causes
the problem. They will get addicted to something eventually. And you
have to live with that.

At the very least, if the industry that produces the addictive goods
is legal it can produce taxes that can be used to reduce the damage it
causes. Plus the production and use is legal so the collateral damage
caused by forced criminatily can be substantially reduced as well.

> Currently, there is absolutely no oversight and control to what online
> game producers can do.  It's in fact very astonishing that they not
> only develop their games very explicitely and carefully according to
> the psychological science of addiction, but also advertise them as
> such, and the addictivity of a game is hailed in computer games
> magazines as a virtue.

It may be a hyperbole, at least in part. Anyway, most games have
limited amount of campaigns, the players get bored eventually, and
want a new game.  So they may be addicted to gaming but probably not
to a particular game.
Online games try to get around the lack of content by bringing
multiple human players together.

But you should look at non-computer games as well. People can get
obsessed with chess, card games (for money or not), role playing
games, and what not.
I suspect that here applies the Murphy's law "Coputers allow a human
to make far more errors in shorter time than would be possible without
them."
They just give more gaming possibilities. You can play over net or
against ai instead of meeting other plaers in person. Computers can
simulate many traditional games, and add many new frameworks for games
that would be otherwise impossible.

> This is even more surprising given that one of the main target group
> are kids, and whenever you mix kids and addiction in one sentence, you
> should make an appointment with your lawyers.  The industry is
> actually very aware of the risks, but for some reason it has not yet
> entered public awareness.  Computer addicition (in general) is not yet
> taken very seriously by many, but it is only a matter of time until it
> will become a main focus.  In China, there is already legislation to
> contain it, but there is a conflict of interest because the
> politicians also want to protect a growing industry for economic
> interests.

Well, perhaps they can get some sensible metric of addictiveness of a
game, and mark games that are too addictive as 'unsuitable for
children' much like alcohol or tobacco. The impact of that will
probably depend on the metric and the way the law is enforced..
> If you want to further explore this subject matter, I would recommend
> making an experiment.  Try yourself to design a game that is as
> addictive as it can be.  Then compare your game with the games that
> are currently on the market.  Then try yourself to design a game that
> is just as entertaining, but not so addictive.  Again, compare.  If
> your result is that a game can be just as entertaining but less
> addictive, do you think that there is a moral obligation by the game
> industry to explore these possibilities?  If your result is that
> entertainment is coupled with addictiveness, do you think there should
> be warnings and other methods of monitoring and control to reduce harm?

There is a technical problem in writing a new game. Basically I could:
a) come with a novel cathing idea. This way one could write a quite
simple yet successful game.
b) take one of the widespread types of games (like strategic
simulation, FPS, ..), and recycle. This requires a new or modified
engine (not too hard), and content. Creating content (for a game
comparable to what is on the market) requires at least a few
individuals skilled in different areas and lots of man-hours.

The nearest I could do is to perform a design phase for (b) without
doing the actual work.
It looks more or less equal to answering the question : What features
make a game entertaining, and what features make it addictive?

I am not sure about the answer. And do not know any research done in
answering this question. The most I know of is some reserch in
answering "Are (some) computer games addictive?".
If I meant to do it right the research itself would be quite a bit of
work I guess.
Although it is an interesting problem it is not so interesting for me
that I would do the research myself.

Thanks

Michal