>(Note: I looked at Curses, my g-d, the complexity. If I
This starts me on a different train of thought. Recently I've acquired and
(after suitable agony writing a format conversion program) started to play
the old Scott Adams games. Compared with Curses or the Adventions games, or
even most Infocom, the Adams games have no prose at all, no size, a simple
two word parser that looks only at the first three letters of a word, etc.
Then, why do I, and many others it seems, continue to find these games
charming and well worth playing? I'm inviting discussion on this... is it
pure nostalgia? Curiosity? The desire to "cruise" with something smaller
and simpler from time to time?
Or did Scott really just do a fine job with what he had to work with,
setting puzzles that were elegant for their time, and coming up with themes
and stories that were good by any standard? Perhaps so. There are other
little games in what might be called the "cheesy" category, but after a few
minutes you find yourself deleting them.
Today's text games are great; they offer unparalleled depth and
sophistication, with more improvements ever on the horizon. But some of the
early ones retain a surprising amount of value (IMHO).
how about, because prose, size, and parser are not really essential to
gameplay? also, tight constraints encourage careful design.
About the parser. Call me a heretic, but I'm not particularly
attached to the idea of sophisticated parsers. I rarely type more
than verb-noun, maybe an indirect object, and adjectives as necessary.
It's mostly stuff that would be easier to do with point+click, at
least for nouns. Verbs need slightly different strategies..
I'd like to see a text adventure that takes advantage of a window
system interface. Is this significantly different from a graphic
adventure? Maybe not..
I think that the simplicity, in terms of size, length of text, number
of objects, and parsing, all combined to make it much easier to focus
on what was important to the game. Scott didn't often rely on a single
phrase to do what was needed (like in Lurking Horror when I was stuck
for half an hour trying to come up with the phrase "lower ladder").
He didn't have characters, which are a problem in the Infocom games
because you don't know what the capabilities of the characters are
and you just have to hope that there isn't some more sophisticated
way of interacting with them that you haven't thought of (this stopped
me cold in Witness).
I won't swear by this, but I think he played fair with the player more
than Infocom did. I completed lots of Scott Adams adventures. I have
NEVER completed an Infocom adventure without help or a hint book.
(AND, he used the 3-line status line:
Location, Objects, Obvious Exits. I NEED the obvious exits line.
Somehow I miss text saying "and a path running northeast" when it's
embedded in a 2-paragraph description; I had to give up on Lurking
Horror and read the hint book to discover there was a stairway going
up in the middle of a hall.)
> >Size is a red herring. The Scott Adams games were (on average) about as
> >large as the Infocom ones. I'm measuring in number of rooms, number of
> >objects, and number of commands needed to win.
> Are you sure? I've only played about 4 of the Scott Adams games, but
> I remember they were all much smaller than Zork I.
Mmm, I guess. It's been a while for me. I would put it in the same order
of magnitude, though.
Note also that Zork I had more rooms than the average late-model Infocom
game, because there was much less description per room.
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
How about the Legend system (Gateway, Eric the Unready etc) which doesn't
really work, but is a good try (I just turn off the silly windows and
revert back to typing commands), or the wonderful Magnetic Scrolls
interface (used to best effect in Alice in Wonderland), which almost
out-windowed Windows(tm) ? (< rhetorical question mark)
I found it at gmd.de. Haven't sat down at a PC to play with it yet.
But offhand, I want to say, if it really is easier to type commands,
then it wasn't done right. :)
For one, using icons for verbs is probably silly, except for something
that's naturally visual, like the compass directions.
>I found it at gmd.de. Haven't sat down at a PC to play with it yet.
>But offhand, I want to say, if it really is easier to type commands,
>then it wasn't done right. :)
Based on the tip, I checked this game out. It had the bizarre effect of
drawing *yellow* lines on my mono LCD screen--even if I tried to force
regular VGA mode. Now THAT's bizarre behavior.
For one, it's just layered on top of the traditional text adventure
framework. It's basically a cut+paste interface: click on a word, it
gets pasted at the prompt. You still have to type any word you can't
see. Most of the verbs have to be typed, and the few that you can
click on are the ones that are easy to type anyway: "i", "n", etc.
A well-integrated mouse interface is going to be much different. I
have some concrete ideas on what it can look like, and I may try to
prototype something in tcl/tk.
With careful design, I don't think you need a parser at all, which is
good, because I don't want to write one. :)
I *like* the idea of a good mouse interface layered on top of a good
Myst was a good start at a zero-text, no-parsing mouse interface, but
it wasn't nearly as flexible as the simplest two-word parser. Darkseed
was actually better; it had mouse equivalents of "get" and "twiddle",
just as Myst did, but it also had an idiom for "apply X to Y".
Can you describe what your ideas are?
Scott Adams adventures usually had about 20 locations.
>I found it at gmd.de. Haven't sat down at a PC to play with it yet.
>But offhand, I want to say, if it really is easier to type commands,
>then it wasn't done right. :)
Well, first of all realize that Infocom did typing commands entirely
wrong. They forced you to go out to 6 or 7 letters, ALL THE TIME,
even for words like "examine" that you needed constantly.
Better to be like VMS (can't believe I said that) and require the
minimum number of letters to distinguish words (usually 3 or 4).
Given that, I can type a 2-word command in about 1 second.
No way could I move a mouse around that fast.
Maybe, but I think you can do better if you put the mouse interface on
the same level as the parser, or underneath the parser.
Most good mouse interfaces are built on the principle of direct
manipulation. Using the mouse to manipulate the language to interact
with the game is probably going to be more awkward than using the
mouse to interact with the game world directly.
>Myst was a good start at a zero-text, no-parsing mouse interface, but
>it wasn't nearly as flexible as the simplest two-word parser. Darkseed
>was actually better; it had mouse equivalents of "get" and "twiddle",
>just as Myst did, but it also had an idiom for "apply X to Y".
I haven't seen those. (They're not free, are they?)
>Can you describe what your ideas are?
Let me see...
First, there should be separate windows for your inventory, your
current location, and maybe other stuff if necessary.
Referring to nouns (actually, objects) is easy. Any object that you
can address should be visible somewhere, like in the inventory window,
or the here-you-are window.
Well, it's awkward if you need to refer to a noun that isn't attached
to a concrete object or isn't in your immediate vicinity, but
situations like this should be rare.
Click on an object to examine it. Click+drag to pick it up, drop it,
wear it, or otherwise move it.
For other verbs, every object usually has a fairly small number of
"plausible actions" associated with it. Click on the object to bring
up a list of these actions, either in a pop-up menu, or in a "current
object" window. This object-oriented interface should be easier to
use than an interface that puts all the game's verbs in a single list
or a single control panel.
So you should be able to do most things with just mouse clicks. Click
"horse", then "ride". Click "rope", "tie to", "rail".
It's object-verb order, not the verb-object order of English, but
interaction like this is very common in other mousy interfaces, so it
shouldn't be hard for people to adapt to it.
This structure makes it awkward to have hidden, unobvious actions, but
I'm not really fond of guess-the-verb puzzles anyway. And making all
the verbs explicit this way will ensure you don't have unintentional
guess-the-verb puzzles. (I spent a long time trying to "roll die" in
Navigation should also be simple clicks, probably in the here-you-are
window. Kind of like navigating a hypertext interface. If you want
to be more user-friendly, provide a clickable map too.
Is this too different from traditional text adventures? I could see
this becoming a GUI v. command-line religious war. :)
It's really only a small step from this to a full-fledged graphic
adventure. Just use pictures for the here-you-are window, and icons
Actually, there aren't really any hard ideas in what I've described,
so I suspect there are already graphic adventures with a structure
like this. I don't really know what's out there. The most recent
thing I've seen is The 7th Guest, which isn't really an adventure
game. Before that, I think I played some of the King's-Quest and
Leisure-Suit-Larry series. I must be way out of date.
>Click on an object to examine it. Click+drag to pick it up, drop it,
>wear it, or otherwise move it.
>For other verbs, every object usually has a fairly small number of
>"plausible actions" associated with it. Click on the object to bring
>up a list of these actions, either in a pop-up menu, or in a "current
>object" window. This object-oriented interface should be easier to
>use than an interface that puts all the game's verbs in a single list
>or a single control panel.
Actually, Return to Zork has this type of thing. The entire screen is
graphic. To use something, you click it (right button to call up
inventory), and you will get a small window of "things you can do with
this object". Click one, and zippety-doo away you go.
Ahem... All Infocom games (from 85 onwards, I think) had "x" for
examine; that's as short as it gets.
>Better to be like VMS (can't believe I said that) and require the
>minimum number of letters to distinguish words (usually 3 or 4).
In fact, this is something that could be implemented into any Z-code
interpreter. Word completion on the command line would also be
> >Myst was a good start at a zero-text, no-parsing mouse interface, but
> >it wasn't nearly as flexible as the simplest two-word parser. Darkseed
> >was actually better; it had mouse equivalents of "get" and "twiddle",
> >just as Myst did, but it also had an idiom for "apply X to Y".
> I haven't seen those. (They're not free, are they?)
No, they're commercial games. Summaries:
Myst: the screen displays a first-person view. Click on part of the
scene to walk there or turn to face it. Click on an object to move
closer to it; if you're already close to it, click to push, pull, open,
close, or take it. In some cases you have to click-and-drag to (for
example) slide a lever up or down. The cursor changes to tell you
whether you're going to turn, move, or do something else. (Only those
three cursors.) If you pick something up, the cursor indicates that, and
then clicking on something applies the object to it. You can only carry
one thing at a time.
Darkseed: The screen displays a third-person view. There are three
modes: walk-to, examine, or use. At the top of the screen is your
inventory, as a sequence of icons. If you "use" something, you may have
the option of clicking on another thing, to apply the first thing to it.
> Click on an object to examine it. Click+drag to pick it up, drop it,
> wear it, or otherwise move it.
> For other verbs, every object usually has a fairly small number of
> "plausible actions" associated with it. Click on the object to bring
> up a list of these actions, either in a pop-up menu, or in a "current
> object" window. This object-oriented interface should be easier to
> use than an interface that puts all the game's verbs in a single list
> or a single control panel.
This solves the biggest problem with Myst, which is that each object
only has one verb associated with it. However (as you note) it makes it
very obvious what each object is for. I don't like "guess-the-verb"
puzzles either, but I *do* like finding an object that I don't know what
it's for. I find a glowing jewel -- do I rub it, look into it, throw it
at someone, or eat it? If all the "correct" verbs are spelled out for
me, the basic flexibility of the text game is lost.
I actually prefer the idea of a fixed list of verbs, which all can be
used with every object (even if most attempts just get "You can't tie
your sword to anything, dope.") It doesn't take too many verbs to cover
an interesting game.
Alternatively, spend a *lot* of time making red herrings. If the pop-up
menu for the glowing jewel really does include "rub", "throw at...",
"stare at", and "eat", it's an improvement. Hrm. I still prefer a
general list -- it gives me much more of a sense of control. Dunno, it's
just a gut reaction. Even limiting me to the generic "apply X to Y"
feels preferable to having a limited list of verbs for each object.
Well, here are some approaches to unusual actions:
- Once you discover an unusual action, maybe by reading a note that
describes it directly, then that action gets added to the object's
action menu. This doesn't work for indirect information, like: decode
this cipher to find out what to do with the jewel.
- Let the player add and delete actions in an object's action menu.
The point is to optimize for the common case. Any object is usually
used in just a few ways, and a small menu is much easier to use than a
>I actually prefer the idea of a fixed list of verbs, which all can be
>used with every object (even if most attempts just get "You can't tie
>your sword to anything, dope.") It doesn't take too many verbs to cover
>an interesting game.
Well, Curses has about 100 distinct verbs. This could probably be
reduced to maybe 50 if you prune away the motion and status verbs.
This is still large for a menu that's supposed to be used every few
But it's not a big deal to provide a menu of all actions in addition
to the per-object menus. I think the per-object menus should be a lot
more useful though.
Of course, there's the problem of actions that are unobvious in
general, like the magic words in Adventure. Some of this can be
handled with the revealed-action strategy. Maybe the rest of the
cases can be handled be letting the user type in arbitrary actions,
but this seems to me another guess-the-word puzzle. The plover room
in Adventure is a particularly annoying one.
Another way of handling unobvious actions would be to let the user try
to "activate" any word they see in the narrative or the descriptions.
This is a no-typing approach, but offhand I don't think it's any
better than the typing approach...
How was it? Was it easy to use? Did you have to do any typing at
all? Was it ever easier to type than to mouse? Did the mouse
interface impose any noticeable constraints on the game design?
Not quite correct. Here is the list:
Adventure # 1: 33 locations
So the average number of rooms would be 30, which is still less than in
Infocom's games (the two smallest Infocom games, Witness and Seastalker,
had exactly 30 rooms). "Zork I", by the way, has more than 100 rooms.
On RTZ, my complaint about the point + click interface was that in order
to spot an object that was useful, all you had to do was move the pointer
around the picture until you got the name of an object. There was no way
to disguse a useful object (the lighthouse railing comes to mind), as you
can in text-based systems, i.e., the "scenery" flag in Inform.
Admittedly, I'm digressing from the point (but what else is new?). I
think that part of the romance of a text adventure is lost without
typing. Point + click interfaces are essential for things like desktop
publishing, GUI's, and the like. A text adventure should be more
"organic", you and your keyboard. No GUI's, no graphics, no sound
effects. The true power of a text adventure is the writing, the words of
the author, coupled intimately with the imagination of the player. Once
we trust that mental image to one person (a graphics artist), the major
effect is lost. I had a very, very distinct image of the caverns of the
Great Underground Empire for a decade before I played RTZ. What I saw on
my screen just didn't match.
----------------------| S.P. Harvey |--------------------------
"Most of the world was mad. And the part that wasn't mad was angry.
And the part that wasn't mad or angry was just stupid.
I had no chance. I had no choice." - Charles Bukowski, 'Pulp'
----------------------| sha...@interaccess.com |--------------------------
> How was it? Was it easy to use? Did you have to do any typing at
> all? Was it ever easier to type than to mouse? Did the mouse
> interface impose any noticeable constraints on the game design?
It was a pain. (I haven't finished) You can't type anything
at all. There is NO text. You MUST use sound to listen to
the NPC's (yeesh). There is no equivalent to "again". If
something needs to be done repetitively, you trudge through
And it's SLOW. You can't just bypass someone's long speech
if you've heard it already, or skip quickly past some
animation, etc. (of course, Monkey Island II was annoying
in this regard, but since I enjoyed that one, I'll forgive
- Luxury! In MY day, we had to make do with 5 bytes of swap...
However, idea for verb/noun:
Click a verb, click an object. That verb is applied to the object, and if
it gets anything other than the "You can't tie the sword to anything,
dope" reaction, that verb is added to the list of verbs for that object.
Note that sometimes the first use of a verb with an object may be in an
inappropriate place or otherwise ineffective, so a simple search through
the verb list would not help. Also, for commands which require Direct and
Indirect Objects, these can be prompted for "What do you want to tie the
sword to?" and have the response be a point-and-click at the object setup.
but, in a text adventure, you can just say "get all" to find out what
all the objects are anyway, right? If "get all" doesn't work, then
just type "get X" for every noun you see. I don't think this is any
different from waving the mouse around to find potentially useful
objects. It's just more tedious.
Anyway, I'm pretty fond of text adventures as a form of fiction, so
I'm trying to see how a good GUI can simplify the mechanics of
gameplay without detracting from the fiction experience.
I think the way out of this difficulty is to only
have generic or obvious verbs, and design the puzzles
so that they require ingenious combinations of
ordinary actions, rather than a single unusual
action with the right object.
For instance, the things you might be able to do with
the jewel are take it, examine it, wear it, put it in
something else - i.e. all of the usual things you would
Somewhere else you might find a camera with a flash.
The generic verbs apply to this, too, plus a few special
but obvious ones such as looking through the viewfinder
and taking a photo.
What's not so obvious is that if you open up the flash
unit and insert the jewel, you end up with a laser!
With this kind of puzzle, all of the possible actions
for an object can be displayed in a menu without
giving anything away.
I even suspect that a system which forced the author
to declare all the possible actions in this way
would result in better puzzles.
|> "And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
Greg Ewing, Computer Science Dept, +--------------------------------------+
University of Canterbury, | A citizen of NewZealandCorp, a |
Christchurch, New Zealand | wholly-owned subsidiary of Japan Inc.|
Hey cool, LucasArts without the spiffy graphics! No offense, Greg. Your
idea just doesn't appeal to me. The problem is not the unusual actions
you need to take to solve certain puzzles. The problem usually lies in
poor implementation or betatesting. If there aren't enough clues or
synonyms to solve the puzzle, then players are gonna get stuck. The
trick is to let some people beat your game up for extended periods of
time prior to releasing it. Well, I mean, if you really love your mouse,
be my guest to write such adventures. Still, I have yet to see a good
one done in this manner that lasts me more than a night without some damn
maze to get lost in.
<silly sarcasm on, insert smilies as needed>
"I mean hey, we've already got auto-mapping tools for our games...Let's
add in auto-verb-finding. Hey, while we're at it, let's just make some
auto-playing tools and script the result to file, then go back and read
it later. :)"
Seriously. You can't make a hard game with that sort of interface.
Here, look at these playing times:
Monkey Island - 10 hours (hey, I was a lot younger then.)
Sam n' Max - 4 hours.
Ringworld - 1 hour.
Day of the Tentacle - 6 hours.
Last Half of Darkness - 35 minutes.
Compared to the Infocom games, these were unsatisfyingly short and easy.
I don't pretend to know why, but I feel that their interface has a lot to
do with it. The only puzzles I got stuck on were 'hunt the pixel'
puzzles, that special idiocy reserved for Graphical games.
Now, probably you could rewrite Suspended to use such an interface, sans
pictures and all, but the difficulty would still be diminished.
Admittedly, it's not always a bad thing to make a game easier, but it's
not valid to state that easier is better, either. Hell, I LIKE not
having a list of verbs, just so long as 'charm snake' and 'play flute'
mean the same damn thing. And you guys have heard my tirades against
mice so much I'm sure you're no longer interested. So put a mouse
interface in, with a nice list of verbs and nouns, or whatnot. Then, in
the lower right hand corner, put an icon of a typewriter that turns
everything OFF and lets me play it my way.
Return to Zork spoilers lie ahead, BEWARE:
Consider Return to Zork, then. I still contend that every puzzle was
either bafflingly trivial, or terribly obscure. Someone please tell me
why I would burn a bra box? Now explain why milk cures blindness.
Lastly, explain to me why I can't climb a tree to grab some coins out of
it but must instead hit it with my sword? These are the type of things
that show it doesn't matter how few verbs you use, poorly written is
poorly written. All things being equal, I prefer ye olde keyboarde.
< V R I O Software. We bring words to life! | ~~\ >
< T "We at Vertigo apologize for the delay. Sorry." | /~\ | >
<_WATCH for Avalon in early AUGUST!___wh...@uclink.berkeley.edu__|_\__/__>
> With all the digitized audio and video, you simply cannot afford extra
> objects to help flesh out the world. It's simply a question of
> resources. Text-based IF is much more forgiving here.
> I realise we've strayed from the original topic. Most of these
> complaints would not be inherent in a point + click textual interface,
> but until we can afford hundreds of extra megabytes of sound and video
> for background objects, we're going to be left with a group of clearly
> usable objects against a backdrop of pretty digitized video.
That's the author's decision. I am now embroiled in plans for a game
with *no* digitized video, or as little as possible. When you go back to
still frames, a 600 meg CD suddenly becomes big again. Big! Big, I tell
Half the things you find should be not necessary to winning the game. Of
the things that are, half the things you can do with them should not be
relevant to the plot. Of the things you can do that are relevant to the
plot, half should turn out to be wrong, or at least can be done in the
Gerry Kevin Wilson:
>So put a mouse
>interface in, with a nice list of verbs and nouns, or whatnot. Then, in
>the lower right hand corner, put an icon of a typewriter that turns
>everything OFF and lets me play it my way.
heh. I could say, you just haven't met the *right* mouse interface,
but that would be silly and probably not true. However, I know that
*I* haven't met the right mouse interface.
The Multidimensional Thief is actually a game that could probably have
a good mouse interface. Most of the puzzles in it are a matter of
working out the implications of a few "simple" objects, like the
portable hole. There are a few unobvious actions that would be
awkward to hide though..
Let me see if I can build a mouse interface for Balances, and I'll get
back to you. If this works out, I might even sit down and write the
Great American Text Adventure that I have in mind.
However, if you want to follow this rule, make sure that you can always go
from one room to any other room. I HATE having to restart the game because
I couldn't carry enough in the first half, and so I miss some vital item, and
I can't return to the first half to get it. And even so, half of all items
being red herrings seems a little excessive (unless, of course, you're counting
scenery items, in which case I'm in more-than-complete agreement with you)
>"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
| One .sig to rule them all, one .sig to find them... |
| One .sig to bring them all and in the darkness bind them |
| The Grim Reaper (Reaper of Souls, Stealer of .sigs) |
| scy...@u.washington.edu |
[Return to Zork spoilers]
: Consider Return to Zork, then. I still contend that every puzzle was
: either bafflingly trivial, or terribly obscure...
: ...Now explain why milk cures blindness.
...I think I finally figured this one out. Carrots are good for the
eyes. The cow only eats carrots. Therefore, the milk is made from
processed carrots and is good for the eyesight. A clue is given to
you if you drink the milk; it says you can see better... but I agree,
this is totally obscure. My first thought was just to eat (or give)
(RTZ suffered for me, big-time, at the end, where I dearly wished for
love brings with it strange currencies domi...@ucunix.san.uc.edu
> >Half the things you find should be not necessary to winning the game. Of
> >the things that are, half the things you can do with them should not be
> >relevant to the plot. Of the things you can do that are relevant to the
> >plot, half should turn out to be wrong, or at least can be done in the
> >wrong order.
> However, if you want to follow this rule, make sure that you can always go
> from one room to any other room. I HATE having to restart the game because
> I couldn't carry enough in the first half, and so I miss some vital item, and
> I can't return to the first half to get it.
Sorry; having to restart the game is exactly what I mean.
It's unlikely that I would use one-way doors and being unable to carry
enough stuff; that's kind of tacky. But I very much like objects that
can only be used once, where you have several possible places to use it.
That comes to the same thing; you have to try it several ways, and you
*will* have to restore the game (unless you're incredibly lucky.)
I will make it obvious that you're doing something irrevocable. (ie,
using an object that can only be used once.) OTOH, if I used a one-way
door, that would be obvious too. ("There is a long slippery slide
leading down from here. You will not be able to make it back up...")
I assume that my players are willing to save the game before doing
anything irrevocable. If you're not, um, play some other game.
Hmm, okay, I guess. IMHO, it should always be possible to play the game
straight through, no save/restore, and win, without any luck being involved.
However, since most players aren't perfect reasoners like Sherlock Holmes
or something :P, and they can't look at the source to figure out what
I was thinking, they'll probably have to restore quite a lot. But in theory,
I think you should always be able to figure out all the puzzles by skill
and logic, and not need to guess. I don't like one-way situations. They're
hard to avoid (goodness knows... Curses must have over a dozen situations where
you have to restart... very annoying!), but ideally possible. The game I'm
working on now will be linear, but I don't think (the stuff I've got planned
now, anyway), it'll be possible to work yourself into a hole where you have
to restart. Death is possible, yes. But I prefer dying to having to restart
the game! Usually, when going someplace deadly, you are warned beforehand
and can save. Whereas if I miss the bottle of weed killer in Curses, and
walk into the area where I'm supposed to use it, and I don't have it, I'm
stuck! Not only that, but I won't realize I'm stuck for a really long time,
probably, and so I won't have saved beforehand. But if there's sufficient
warning to save, and if it doesn't just depend on luck, then cool, go ahead
and be irrevocable. (This *is* all MHO, of course. YMMV, and all that,
so don't take this as the game design bible or anything :P).
>I assume that my players are willing to save the game before doing
>anything irrevocable. If you're not, um, play some other game.
Well, that's true.
>"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
I could see this as a deliberate design decision. It's easy enough to
give the message "Can't do anything with that right now" if you click
anywhere at all, but this would probably make the game much harder.
For the "tie rope to railing" situation, what I'd probably do is
attach an action to the rope but not the railing. I can't think of
any situation where I'd want an object that has "tie rope to it" as
one of its actions. This is basically the rope's action, not anything
else's. This keeps the puzzle unobvious, but solvable.
>With all the digitized audio and video, you simply cannot afford extra
>objects to help flesh out the world.
I'm not sure this is relevant. Audio and animation are expensive, but
a still picture is pretty cheap. If you want red herrings, it's not
too much work to delineate a number of useless objects in the scenery;
this is just bounding-box and bitmap.