WARNING! THE FOLLOWING IS FAIRLY LONG (BUT NOT NEARLY SO LONG AS A
PSYCH PROFESSOR'S PERORATION, SO TAKE HEART)...
Having settled on the plot, puzzles, and etc., I've learned that (at
least in my case; if anyone disagrees about this method, I'd be
interested in hearing their comments) its best to code the story in
"passes." The aim of the first pass is to get a complete, working game,
with all the somewhat extraneous elements like menus and such in place,
but with only the bare minimum of commands required to solve it
included. In other words, I don't try to code for every possible action
the player might take with respect to this or that object, nor do I
provide for any alternate solutions at this point. Anyhow, after I have
this "skeleton" game in hand, I will try to subject it to every kind of
abuse I can dream up in order to anticipate and correct problems which
might crop up during play; I want to make sure that, no matter what I
do later on, the game will be solvable and, for lack of a better word,
"consistent" in its gameplay.
(Whoops! Back up; the above should really be step two. Step one can be
considered a distinct step: I first code up the map, as completely as I
can, with perhaps a few necessary objects but for the most part empty,
so that I can walk through it, critiquing and perfecting the room
descriptions and checking that all the movement directions provide the
proper passage or yield appropriate messages--note: this can also be a
good way to see how best to break up my files, since I can't bear
trying to edit one ultra-long game file and I tend to separate it out
according to some geographical principle. *Then* I move on to what I
described above).
Sorry about that. Ok, after the skeleton game is done and I've beat up
on it a little, I move on to pass two (or three, depending on how
you're counting at this point), in which I go back and treat every
object as completely as I can, with the sole exception of NPCs'
dialogue. I pull out my printed copy of the verb library--as well as
considering any new verbs and commands I have added, of course--and run
down the list for each object, trying to imagine the various ways an
item might be used and giving due attention to the various locations in
the game, since it may seem entirely reasonable to the player to try
using an object in some very particular way in some particular location
(or in combination with some other object) which I hadn't considered
when I first conceived the game (certain objects are particularly prone
to this kind of usage; like ladders, for example). But this doesn't
mean that I actually provide special routines for every possible use of
every possible item; that would be practically an impossible task in
all but the shortest of games. Yet I do try to anticipate and consider
each possibility, so that I (hopefully) don't miss the really important
and reasonable things (or the many opportunities for a good joke which
arise from this fine-toothed-comb treatment). "Total coverage" is for
me the ideal to be strived for in the second pass, even if it is never
likely to be achieved.
Finally, after all this, I will make the last pass, in which the major
task is the consideration of NPC dialogue, encyclopedia entries, and
the like. In the case of NPC dialogue, my experience has been that if I
try to "make it up as I go along," I invariably end up with either a
very dull character or a decidedly schizophrenic one (since I will have
forgotten by the end many of the particulars of his/her personality
which I had in mind in the beginning). Also, after the game is more or
less complete and writ in stone, it is much clearer exactly what the
character's actual role in the game is and what the proper scope of
his/her "knowledge" about things. This isn't to say that I consider it
a cardinal sin to put in a dialogue response during one of the earlier
"passes"; if I come up with a good line, I want to include it before I
forget it. But all things considered, I've found it preferable to code
up my characters' dialogue and behavior (as far as possible) at a
single sitting.
The case of encyclopedia entries (and other similar things) is somewhat
the same, yet the reason for saving the bulk of these for the third
pass is more like the case of any other object in the game. I don't
have a problem with adding entries as I go along, and this is even
preferable, really, since the text of any given entry will be far more
complete and appropriate (both in terms of what is said and what is
*not* said), when I have clearly in mind the actual atmosphere and
conditions of the game which exist at the time the player will be
consulting the encyclopedia--this is always most clear when one is
actually coding and doing the heavy lifting of the game development.
But I think it's advisable to add a separate, dedicated consideration
to this element of the game for completeness' (and again, humor's)
sake, and there's no very compelling reason which ought to prevent this
from being saved until the last.
All that having been said (hey! you still awake out there?), it is a
fact that I have never yet released a game. Why, then, should anyone
pay any attention to what I've said? Well, perhaps they shouldn't. I'm
very far from claiming that my way is best for everyone (or even for
me; I'm open to being persuaded otherwise). On the other hand, I have
four separate games at fairly advanced stages of completion which I
have grown frustrated with and, in essence, junked, because I violated
the "rules" I set out above--or, to put it more accurately, I have
learned this mode of procedure in the school of hard knocks. Now that
I am proceeding on my fifth game in what is, for me, a more orderly and
consistent way, the going is proving very much smoother.
But I'm with Whizzard. I'm interested in hearing the opinions of you
others out there. C'mon, guys, spill it!
Jim Newland
76461.2144@compuserve.com