[MUD-Dev] Is database access a bottleneck?

[Ted L. Chen]

Koster, Raph
> From: Ted L. Chen
>> bradley newton haug
>>> Raph Wrote:

>>>> I frequently see military projects using game-like technology
>>>> that took years to make and cost millions of dollars and
>>>> probably could have been done in a few weeks by any competent
>>>> FPS team in the industry.

>>> Goes both ways, I agree.  Flight Sims are a good example here.

>> 'game-like' I would say is the operative word.  A lot of things
>> can seem similar on top while varying drastically underneath.
>> That said, I wouldn't trust using Microsoft Flight Sim in almost
>> any research project (yes, I know there are attempts from certain
>> people to bridge that gap).  And conversely, I wouldn't want to
>> run the latest CAE B747 simulator in my living room.

> I am not talking about heavy-duty sims. I am talking about
> projects like "we're going to take the Half-Life engine, project
> it onto a rooms wall, hire live actors to send the trainee orders
> via radio, and build a small interactive scenario where you have
> to figure out how to deal with a military jeep car accident in
> downtown Mogadishu." Frankly, any of the Tom Clancy games would
> probably have been better training. :)

Hehe, I can't argue against that example ;) I hope you weren't
referring to that one when you said "cost millions of dollars".  The
ones I hear mainly have to do with linking up tank trainers and use
HLA or some sort of large-scale multi-user thing.  Sorta like a MUD,
but they have interactive executive control over the simulation.
Those are pretty big engineering feets, so I'd probably throw that
into the heavy-duty sim category as well.

> That is just one example. I am not saying that there isn't
> important and valuable work and knowledge available there--quite
> the opposite. I am saying that the reluctance to find solutions
> outside of one's own backyard extends everywhere.

I'm not sure if it's reluctance or whether it has more to do with
ignorance, applicability, and available resources.

Short of the GDC, the games industry seems pretty closed.  From a
researcher's standpoint, it's mostly still DOOM.  And there's really
no place to see what the latest development is aside from the most
superficial glance in your local EB.  Only if you spend time (an
inane amount I might add), surfing all the gaming sites, and lists
such as this, would you realize that someone out there might have an
applicable theory or product.  Academic journals and conference
papers provide this 'collected' resource the other way around.

Once you pass that hurdle, you hit applicabilty.  A lot of research
oriented code revolves around data collection.  Believe it or not, I
actually wanted to incorporate MS Flight Sim in a piloted study.
Sadly, MSFS, like most games, is closed systems and make quite a lot
of assumptions when they tackled specific problems.  Precise state
information is usually the first to go.

So you've finally found something that's useful to you and can
actually be applicable.  How do you go about incorporating it?  For
research projects that aren't of the 'heavy-duty' category, getting
good programmers is hard (a harder one than the game industry
faces).  Funding is too.  DOOM just happens to fit the nice feature
+ free + easy coding *cough* entry that fits the bill.  Going up the
FPS chain, you start getting more advanced techniques which puts
more onus on the programmer having that knowledge.

>From a product manager's standpoint in the games industry,
middleware might save you a lot of money.  From the research
manager's standpoint, that same thing might just as well cost you a
lot more in terms of getting someone up to speed to do the necessary
code changes.  It might be cheapest to use DOOM or roll your own
simple engine.

Don't get me wrong, I'm in the happy place on top of the fence.  I
just like playing devil's advocate and giving reasons why it's not
the most comfortable place to sit on.. did I mention it was a picket
fence?  =)

TLC


_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev