[StoryHarp] Detailed Review
Bob Newell
The authoring system 'Story Harp' has been much advertised by its
author on this newsgroup. I found myself being initially critical of
the product without having had a thorough look. This was certainly
unfair, and so, I downloaded version 1.1 and spent some hours
evaluating and testing, in order to be fair about the whole thing.
I'd like to present a summary of my findings. These findings are
purely my own opinion. You are encouraged to make your own
evaluations before finalizing your opinions about Story Harp.
Story Harp presents itself as a certain type of text-based story
authoring and playing system. (The player and composer are part of an
integrated package.) Story Harp runs only under Windows 95; it is a
shareware product with a $79.95 license fee, payable if you use the
authoring system for more than 24 hours of on-line time. There is at
this point no limitations on the unpaid use of the player portion of
the system.
Story Harp allows rapid assembly of "choose your own adventure" (CYOA)
style text games. A much touted feature of Story Harp is its
incorporation of text-to-speech and speech recognition features in the
player component. I did not test these features, but as will be seen,
this would not much affect my evaluation of the authoring component.
Story Harp comes with a few sample "worlds", or stories, and a
Windows-help based tutorial. My first action was to play a few of
the sample stories.
The player interface is very intuitive. I had no difficulty at all in
starting up the stories and going through them. The visual interface
is a small split window. Story text appears in the top window, and a
multiple-choice style list of options appear in the lower window.
Clicking the mouse on one of the options selects it and moves the
story forward. Nothing could be simpler.
Indeed, that was my first impression. Nothing could be simpler than
playing a Story Harp CYOA; in fact, it was so simple as to be
incredibly dull. This would surely be improved by the presence of
extended text (not part of the sample worlds) with more interesting
themes. It might be improved had I used text-to-speech and speech
recognition; or it might actually be worse, since then controlling the
pace would have been difficult (something like the annoyance of books
on tape). I have, in the past, enjoyed numerous printed CYOAs;
however, the printed book experience seemed far superior to the Story
Harp experience.
Especially annoying was the lack of a clean termination to the
stories. When the end of a story is reached, the program continues to
offer the last frame of the story until you actually exit the program,
independently of the story.
I then moved on to looking at the authoring system. I first read
through the tutorial. I must say that the tutorial is of very high
quality. It is very easy to follow, quite complete, and sequenced in
a logical manner. 15 minutes with the tutorial and I felt fully
qualified to start working with the authoring system.
The authoring system presents several "views" of the story space.
Story locations are referred to as "contexts" and "rules" or choices
appear for these contexts. Each choice in a multiple choice set is
expressed by a separate rule under a given context. Rules may have
conditions which refer to simple boolean-valued variables; rules also
may define a new context (move) to which control is transferred.
These principles are the sum and substance of Story Harp. (I have
omitted playing sounds and music.)
I immediately started to compose a simple CYOA story. I had no
difficulty at all; the learning curve is really flat. I suspect
someone with some programming experience could become a Story Harp
expert in about an hour.
There are some annoyances in the text interface to the authoring
system. Multiple choice options are forced to lower case, giving an
illiterate appearance to the list of choices presented to the player.
This is really detrimental to the CYOA style. For instance, if one of
the options is, "Would you like to ask Fred a question?" it would
appear as "would you like to ask fred a question". Reply text can't
be seen all at once; the input form has a height of one line, and text
just continues off the end of that line.
There are alternate text-input methods. There are "wizards" which
supposedly help create contexts, rules, etc. However, they do so in a
piecemeal manner that makes it hard to keep things together. The
"wizards" use a text-editor like interface, which is actually superior
to the form-based interface, but not as well integrated.
I was most disappointed by the "map" view. I had expected a graphical
input capability that might have made producing a CYOA location
skeleton very rapid. Instead, the default mapping of a story is a
real mess. My prototype story, which had about 6 contexts or
locations, was strewn seemingly at random around the screen. I had to
spend a little while rearranging the graphical display just to make it
legible. And, then, I found there was precious little I was able to
do with it. In the Windows 95 environment, I would expect to be able
to right-click on an item in the graphical display and pop-up a menu
of choices. This would be very useful; for instance, clicking on a
context should give me the option of adding or editing rules, etc.
All that happens is that left-clicking on a context selects that
context for use by the forms-based text input interface. Improvement
of the graphical capabilities should be a high priority for the
authors; they should refer to the GUEmap shareware package for an
example of how this can be done well. For instance, it would be nice
to be able to click on a rule, drag the mouse, and have it create a
"move" to a new or existing context. Alas.....it isn't so.
Story Harp has extensive testing and debugging features. I didn't
spend a lot of time with these, but they appear to be very helpful and
rather easy to use.
So, finally, we come to some sort of evaluation. Story Harp should
not be evaluated one-to-one against the broad authoring systems such
as TADS and Inform (except for suitability of intended purpose).
Story Harp should be compared to other means of generating CYOA
stories, however.
To do this I took a look at the TADS contributed library for CYOA
style stories. (I believe there is a similar one for Inform but I
didn't look for it.) The TADS library works in conjunction with the
TADS compiler and runtime and thus shares all the advantages (and
drawbacks) of TADS. Input is via text editor; I prefer this but my
preference may not be universal. The player is not integrated;
however, run-times exist for a wide variety of computer systems.
TADS-CYOA is also very easy to learn and use. It is a bit more
"syntactical" than Story Harp, but if I were to use it, I would write
a simple Perl preprocessor which would do the syntactical bits for me,
and would make it a very rapid authoring tool. There are no graphical
representations for TADS-CYOA, but on the other hand, the entire text
file is in front of you at all times. (In any event, no authoring
system will obviate the need to pre-plan and pre-diagram your story.)
The TADS debugger is very powerful and reasonably easy to use.
In addition, the added features of TADS are available should you
decide to expand CYOA into something larger or more complex--- or, if
you should simply need an advanced feature at a critical location.
(Story Harp deals with the random number issue by simply not having
them.)
I would be hard-pressed to recommend Story Harp over TADS-CYOA. Here
is a brief summary of my findings.
Advantages of Story Harp:
-Very easy to learn
-Excellent tutorial
-Truly suitable for non-programmers as an introduction to CYOA
authoring
Complaints about Story Harp:
-"illiterate" and boring feel to the player component
-Very limited capabilities
-Unpleasant text interfaces to authoring system
-Disappointing graphical interface (which, if developed, would
markedly improve the product)
-Limited "target" platforms (Windows 95 only) for potential story
players
-Very expensive license for a system this limited
At this point there is simply no way I can recommend Story Harp over
TADS (or Inform) even for CYOA stories. There is just nothing
compelling enough about it to justify the license fee. In fact, even
if it were freeware, it wouldn't be my choice for CYOA authoring.
Bob Newell
Paul Fernhout
Bob -
Thanks for taking the time to create a detailed and fair review of
StoryHarp. Here are a few minor comments.
With version 1.2 (released around May 14), the price for registration is
now $39.95 for a regular registrations, and half that for an educational
registration (available to students, teachers, and parents buying it for
students at qualified educational institutions). The review listed the
earlier higher price.
StoryHarp is useable without Agent (as in your review). And we should do
everything we can to make it more interesting in that mode, so thanks
for the comments on that. However, we find that text-to-speech and sound
effects (which both require Agent) as well as music all tend to spice up
games (both when playing and when editing) without taking away too much
from the imagination (as graphics can sometimes do). Adding sound to an
adventure takes only a fraction of the effort required to add
significant graphics, thus still keeping an audio adventure within the
scope of what one person can do in a short amount of time. StoryHarp's
design to make sound easy to put into IF is, of course, why we call it
an "audioventure" authoring system, as opposed to an "adventure"
authoring system.
Also, because of licensing limitations, one needs a certain sound
effects CD in the CD drive to hear some of the GarTrek sounds, and also
some of the music is only available under WinNT (in the media
directory). We're hoping to have better examples in the future that do
not require relying on availability of third party sounds and music. If
we went to shipping the product on a CD (an eventual goal once we've
satisfied its critics enough to be comfortable stamping out 1000 copies,
and then raise the funding), we would ideally do so with a large
standard library of music and sounds anyone could integrate into their
adventures, with the presumption StoryHarp users would be likely to have
that CD. Of course, I'm not sure where to get those sounds or music yet,
since I'm concerned about the true pedigree of most "royalty free"
sounds and music I see on the net.
Sounds cannot substitute for a compelling plot, and I would be the first
to admit that the examples are not stellar works of IF by any means.
Although I do think "House and Yard", written by my wife, is pretty good
for a short work, and "Max the Computer" (also by her) is fun for a
while when using text-to-speech and speech recognition. The largest
example I made, GarTrek, is really more a thrown together demo related
to space habitats than something with literary aspirations.
I have to admit here I'm a much better programmer than IF author. While
I've played Zork, Adventure, Hitchhiker's Guide, Starcross, and several
other IF pieces, my largest IF work before GarTrek was a rip-off of
"Land of the Lost" done in PET Basic eighteen years ago (and never
distributed).
Several people have pointed out that StoryHarp needs at least one really
good long story written in it for people to take it seriously. Sadly, I
don't think that's going to be something I can do on my own. Maybe I
should ask David Cornelson if he'd let me take a crack at putting his
project into StoryHarp (although I'm not sure if I have the time for
that right now, and I know he wants an Inform version). Mostly I wanted
to use StoryHarp to make many short stories related to space habitats,
but I find having an easy-to-use authoring system still doesn't make me
capable of writing great characterizations, suspenseful situations,
great descriptions, and so on required to make such stories compelling.
I appreciated your comparison of using StoryHarp to using TADS with a
library. Thanks for making it. One thing you did not point out however,
is that StoryHarp's browser and map may provide a big advantage when
maintaining a very large adventure. When one makes a test adventure with
six or so locations, issues of scale are not that apparent. Once you
have hundreds of locations, a browser and a map allow you to much more
easily manage complexity than just using a text editor and some Perl
scripts.
While I know you were disappointed with the map's limited graphical
programming abilities, the purpose of the map right now is more to
provide a means of navigating a complex story as opposed to creating one
(which is done as you point out primarily by filling out textual forms).
I guess we focused so much on creating a system one could rapidly enter
text into without using the mouse too much, that we lost sight of people
who preferred doing more mouse work. However, that does not excuse the
map raising expectations that lead to disappointment, especially since
many of the suggestions you make would not be that difficult to
implement.
One thing StoryHarp provides is a single unified system for doing CYOA
games. If you use a Perl script to generate TADS code, you can't then go
back from the TADS code to the original source when you want to make
major modifications (without losing modifications to the generated TADS
code). At some point we would like to add to StoryHarp the ability to
generate skeleton code for IF programming systems like TADS and Inform,
which would at least save you the trouble of writing Perl scripts.
although the "round trip" problem would still exists in that you
couldn't then go back from TADS to StoryHarp to modify generated code.
Thanks for the great suggestions for future features to improve the map
and other areas of the program. We will be working on the issues you
raised. Hopefully, if we can address most of them in new versions, you
or someone else might consider doing another review.
-Paul Fernhout
Kurtz-Fernhout Software
=========================================================
Developers of custom software and educational simulations
Creators of the PlantStudion(TM) Botanical Illustration Software
http://www.kurtz-fernhout.com