Newsgroups: rec.arts.int-fiction
Path: gmd.de!nntp.gmd.de!xlink.net!howland.reston.ans.net!europa.eng.gtefsd.com!news.umbc.edu!eff!news.kei.com!world!pie
From: pie@world.std.com (Carl Muckenhoupt)
Subject: Re: Choosing your IF setting / genre
Message-ID: <CquA3p.6Jq@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
References: <2rqp40$k1r@nexus.uiowa.edu>
Date: Fri, 3 Jun 1994 20:55:00 GMT
Lines: 51

booth@grant.cs.uiowa.edu (Michael Booth) writes:

>In article <1994May21.194107.17053@cs.tcd.ie>,
>Russell Wallace <rwallace@cs.tcd.ie> wrote:
>>scythe@u.washington.edu (The Grim Reaper) writes:
>>
>>>In article <1994May19.013111.4382@cs.tcd.ie>,
>>>Russell Wallace <rwallace@cs.tcd.ie> wrote:
>>>>[Comments about being able to blow things up deleted...]
>>>>
>>>[deletia]
>>>>(even if it was only provided for one puzzle and is only useful there)
>>>>or cutting things up with an axe (ditto) would do far more to make the
>>>>game believable, than the ability to type TELL THE ROBOT TO BURN ALL THE
>>>>BOOKS EXCEPT THE BLACK AND RED ONES or suchlike; if people are prepared
>>>>to spend huge amounts of effort on the latter, why not the former?  It
>>>>needn't be *that* difficult... just record what material everything is
>>>>made of, and have a section of code that says objects made of paper will
>>>>burn easily (e.g. with a match), objects made of wood will burn with
>>>>difficulty (e.g. with a flamethrower), metal and stone won't burn at
>>>>all...

>>>Sheesh, you have got to be kidding.  Implementing any sort of vaguely
>>>realistic physics into an i-f game is difficult, if not impossible.  Just
>>>saying that wood objects will burn under certain conditions isn't enough.
>>>We'll have to change the description, possibly create ashes, 
>>>alter the object's
>>>weight, change NPC's reactions to the object, possibly have certain puzzles
>>>change (a bucket with the bottom burnt out isn't going to be much use), check
>>>if the object's containing anything so that we can see if those things burn,
>>>etc, etc.  Physics is a much harder problem than parsers, which is why we
>>>see more of the latter than the former.

Take a look at NetHack sometime.  The data structure for an object has a
field indicating its material, which can be paper, wood, stone, glass,
iron, gold, leather, and probably a few others I'm leaving out.  Since the
material only governs the physics of the object, approximate fits are fine -
gems are considered stone, steel is iron, etc.  For any general action,
all the game has to do is switch on the material.  Of course, NetHack has
the power to generate many identical objects, so simply replacing the
wooden object you just burnt with another copy of the "pile of ashes"
object is not a problem. (And needn't be a problem in a text adventure,
either, as long as we can create new objects at runtime.  Heck, if the
development system is flexible enough, you could set the weight of the
pile of ashes according to the weight of the burnt object.)
At any rate, I suggest it not as the end-all of game physics, but a
relatively simple system that yields relatively reasonable results,
and would certainly be an improvement over the systems found in typical
text adventures, which are simply masses of special cases.

Carl Muckenhoupt
