Open source IF development
David Fisher
community and how it might apply to IF in his blog at
http://gamingphilosopher.blogspot.com/2006/04/three-kinds-of-detrimental-closedness.html
...
Here is a portion of it:
... it seems to be 'not done' to discuss pieces
you are working on in any detail in public.
Technical questions can be asked, yes; but how
is it that the rec.arts.int-fiction newsgroups
contains no substantive discussion by authors of
the pieces they are working on?
What keeps authors from, say, publishing their
initial 'story board' and asking for comments?
What keeps them from asking for advice on how to
best incorporate the theme of conflicting loyalties
in their new superheroes game?
I can answer those questions - it is a practice of
being very closed because of the possibility of
'spoiling' the piece for your future audience, a
substantial part of which hangs out on the newsgroup.
But this practice is detrimental to the quality of
works, most of which would benefit from continuous
discussion with peers throughout the process of
creation.
What do people think of the idea of effectively collaborating with the whole
of RAIF on a project ? What would you feel about playing a game where you
already knew the basic plot, most of the puzzles, etc ?
Does the enjoyability of IF depend mostly on the element of surprise ?
David Fisher
Andrew Plotkin
>
> What do people think of the idea of effectively collaborating with the whole
> of RAIF on a project ? What would you feel about playing a game where you
> already knew the basic plot, most of the puzzles, etc ?
>
> Does the enjoyability of IF depend mostly on the element of surprise ?
No, but the enjoyability of fiction depends partly on the element of
surprise. This is true of IF to at least the same extent it is of
other fiction.
It is *more* true of *puzzle-heavy* IF, because figuring out a puzzle
is not fungible with designing the puzzle.
(Yes, I re-read novels. It is an experience of, well, re-experiencing
(via prompted recall) how I felt the first time I read it. This works
both ways: if I know what will be in a novel before I read it, it will
never have the impact it would have had if I'd read it fresh, because
what I'm re-experiencing is learning about the book, not reading it.
This makes me one of those people who avoids spoilers.)
--Z
--
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
*
Just because you vote for the Republicans, doesn't mean they let you be one.
Emily Short
David Fisher wrote:
> Victor Gijsbers makes some interesting comments about the "open source"
> community and how it might apply to IF in his blog at
> http://gamingphilosopher.blogspot.com/2006/04/three-kinds-of-detrimental-closedness.html
> ...
>
> Here is a portion of it:
> initial 'story board' and asking for comments?
> What keeps them from asking for advice on how to
> best incorporate the theme of conflicting loyalties
> in their new superheroes game?
I do discuss such problems with specific people (and I usually credit
this sort of thing as "alpha-testing" or "design consultation" in game
credits, depending on how much in-progress content the person saw).
But, spoilage problems aside, I don't like the idea of design by
committee; I find it more useful to talk such things over in detail
with a couple of genuinely interested parties than to collect general
feedback from a host of onlookers.
Jim Aikin
> initial 'story board' and asking for comments?
I think I'd go even further than Emily on this one. I'm not sure how useful
it would be to brainstorm with others during the initial stages.
My initial conception of a work may be quite specific (and if so, why would
I want to change it at that stage?), but it often mutates as I go along.
Things that I thought would be cool start to look stupid. Things that I
thought would be a cakewalk turn out to be very difficult to code. New ideas
for enhancements occur to me.
Maybe I need a letter between alpha and beta. It would, I think, be
genuinely useful to get some feedback on my current project as it nears
completion but before beta-testing. Beta is about finding bugs (though
beta-testers do sometimes suggest useful enhancements). Once I have a
completed alpha version, it would be great to have someone go through the
whole thing and say, "Have you ever considered X, or Y, or Z?"
--Jim Aikin
ChicagoDave
> What do people think of the idea of effectively collaborating with the whole
> of RAIF on a project ? What would you feel about playing a game where you
> already knew the basic plot, most of the puzzles, etc ?
I view IF development the same as the relationship that a journalist
has with an editor. I think the writer sets the tone and chooses the
bulk of what to say in an article, but the editor has the
responsibility of adding or suggestion the bigger picture. There might
be fact-checkers that review the article before it's finalized.
In IF, the writer has the editor and the writing role and then uses
fact-checkers for verification. But they may also use a small group of
people to discuss ideas with.
Obviously we have some strong editor/writers in the IF community. I
think this is an anomoly and I believe as a whole, IF might be better
served if those roles were separate. The writer comes up with an idea
and has a friend who they trust to edit their storylines. They also
have a group of people that might add depth, do research, or bring new
ideas to the concept.
In this scenario, I think open-source IF would be more efficiently
developed.
I'm sure those in the community that view themselves as strong
writer/editors would feel extremely uncomfortable with this scenario
and that's understandable. If you know you're capable of doing both
jobs well, why would you enlist outside help?
But some or many of us might feel that we're not that strong of an
editor or writer, but we do have great imaginations, research skills,
testing skills. Or we are strong writers, but prefer someone else to
verify things and offer suggestions on a regular basis.
I think there has been a lot of assumption about how IF _should_ be
written. I think that stems from the way it has been successful to
date. It has never been tested or proven to work otherwise, but that
does not mean that a more collaborative approach would not work.
It just means it hasn't been seriously attempted.
David C.
Adam Thornton
"David Fisher" <da...@hsa.com.au> wrote:
> I can answer those questions - it is a practice of
> being very closed because of the possibility of
> 'spoiling' the piece for your future audience, a
> substantial part of which hangs out on the newsgroup.
> But this practice is detrimental to the quality of
> works, most of which would benefit from continuous
> discussion with peers throughout the process of
> creation.
>
>
> What do people think of the idea of effectively collaborating with the whole
> of RAIF on a project ? What would you feel about playing a game where you
> already knew the basic plot, most of the puzzles, etc ?
>
> Does the enjoyability of IF depend mostly on the element of surprise ?
Surprise? No. Novelty? Probably.
But what proportion of your novels do you re-read? What proportion do
you read six or more times?
A tester is necessarily going to see the same old stuff over and over
and over again, and it gets very boring after a while. Certainly one of
my problems in terms of retaining testers is that they get bored with
the areas I've implemented or that they know how to get to, and really
don't have much interest in going back and proving that I've fixed
behavior in some part of the game they consider over and done with.
If *everyone* is a tester, then everyone will get burned out on the game
before it is in what I, the author, consider releasable shape.
Adam
Andrew Plotkin
>
> I view IF development the same as the relationship that a journalist
> has with an editor. I think the writer sets the tone and chooses the
> bulk of what to say in an article, but the editor has the
> responsibility of adding or suggestion the bigger picture. There might
> be fact-checkers that review the article before it's finalized.
>
> In IF, the writer has the editor and the writing role and then uses
> fact-checkers for verification. But they may also use a small group of
> people to discuss ideas with.
>
> Obviously we have some strong editor/writers in the IF community. I
> think this is an anomoly and I believe as a whole, IF might be better
> served if those roles were separate.
In other contexts, you've talked about a "writer/implementor" model,
where one artist invents a world and the story events, and someone
else takes that work and writes the IF source code.
Is that what you're talking about here? Or are you thinking of the
"writer" as a writer/implementer (like we do currently), with the
"editor" having the traditional book-editor role of feedback,
suggestions, and management?
--Z
--
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
*
Making a saint out of Reagan is sad. Making an idol out of Nixon ("If the
President does it then it's legal") is contemptible.
Benjamin Caplan
> > What keeps authors from, say, publishing their
> > initial 'story board' and asking for comments?
>
> I want to change it at that stage?), but it often mutates as I go along.
> Things that I thought would be cool start to look stupid. Things that I
> thought would be a cakewalk turn out to be very difficult to code. New ideas
> for enhancements occur to me.
My impression (not having actually completed any games either way) is
that the main value of open-source development would be an increase in
feasible *scale*. An OS'd IF game could reach the size and complexity
of a modern commercial game that enjoys the hired full-time attention
of dozens or hundreds of developers. (Not the _file_ size. But I
wouldn't be surprised if it reached 3 to 5 megs, even if I7 wasn't
involved.)
The trick is finding a game that justifies this much (collective)
effort. (People seem to be doing quite well without OS development.) I
am thinking of Hack-->Nethack as a model for this project; therefore I
propose that an existing game be selected, one that is generally
well-thought-of and that would benefit from a major 'expansion,' and
its author persuaded to release the source.
One of the biggest obstacles, I think, will be that the current
prevalent style in IF is towards smaller, tightly integrated games that
really wouldn't 'expand' well. Still, only one suitable specimen is
needed.
Just my two cents.
Emily Short
Benjamin Caplan wrote:
> The trick is finding a game that justifies this much (collective)
> effort. (People seem to be doing quite well without OS development.) I
> am thinking of Hack-->Nethack as a model for this project; therefore I
> propose that an existing game be selected, one that is generally
> well-thought-of and that would benefit from a major 'expansion,' and
> its author persuaded to release the source.
I think there was a project to expand Anchorhead, in fact, though this
was with a specific research goal in mind (search for "Search-Based
Drama Management in the Interactive Fiction Anchorhead"). And someone
is doing a graphical version of Vespers
(http://www.rampantgames.com/articles/IFInterview.html). I realize
neither of these is exactly what you're describing, but there is some
precedent.
For that matter, there are lots of versions of Adventure out there.
Mike Rozak
> What do people think of the idea of effectively collaborating with the
> whole of RAIF on a project ? What would you feel about playing a game
> where you already knew the basic plot, most of the puzzles, etc ?
>
> Does the enjoyability of IF depend mostly on the element of surprise ?
Some random comments:
- As was pointed out, "more people" implies "larger worlds". Is bigger
better?
- For many people, the "fun" is creating the game, not playing it. There's
no reason why such people wouldn't team up and create a larger game.
- Of course, it'd be like hearding cats.
- The text MUD (www.mudconnect.com) and amateur MMORPG
(http://update.multiverse.net/forum/) communities rely heavily on forming
teams. This is particularly important for graphics-based worlds where the
eye candy is a lot of work. (In MUDs/MMORPGs, for various reasons, bigger is
considered better. I disagree with many of the reasons though.)
- My own multi/single-player IF development kit, http://www.mxac.com.au/mif,
includes graphics eye-candy, although not as much as Multiverse; Teams may
prove helpful (but not necessary) when producing content for it.
- Cats aren't heard animals; you can easily spend more time hearding and
fighting over direction than producing.
--
Mike Rozak
http://www.mxac.com.au
ChicagoDave
> In other contexts, you've talked about a "writer/implementor" model,
> where one artist invents a world and the story events, and someone
> else takes that work and writes the IF source code.
>
> Is that what you're talking about here? Or are you thinking of the
> "writer" as a writer/implementer (like we do currently), with the
> "editor" having the traditional book-editor role of feedback,
> suggestions, and management?
This is different than those "other contexts", which is more of a
methodology than simple collaborative development.
Yes, I believe the traditional book-editor role might be a good one for
some writers to have access to, including myself.
The open source model has one key component that seems difficult to
transfer to IF development. When someone starts an open source
development project, there is usually one person or a small group of
people that are "in charge" of the scope and direction of the project.
After that, people are asked to attack a list of requirements and as
long as those requirements pass unit and integration testing, they can
get checked in.
So lets say that you and I want to create a three-part space opera that
has puzzles and involves a complex set of puzzles regarding space
travel, communicating with alien cultures, and various laser
technologies.
Like a D&D Dungeon Master, you and I develop the backstory and the
outline for all three episodes.
We then build out that outline to have rough scenes that must
accomplish certain tasks within the scope of the episode and the
trilogy.
Once this is complete, we can advertise our project and wait for people
to sign on. People will tackle various scenes and submit them for us to
review. We can edit them or accept them as is.
Some of the writing might not be game specific. We might have a task
that says, "We need a system of lasers that have the ability to attack
aliens that can phase-shift into alternate dimensions." or "We need the
backstory for an alien culture that has the ability to transport
through space without technology." This writing will then get
incorporated into the base project and tasks adjusted accordingly.
As scenes are added, maybe we go through a review session and "firm-up"
the whole project or possibly make alterations.
Hopefully, we garner enough interest to get enough writers, testers,
and coders to make the whole thing come together.
David C.
Michael Martin
> ... you and I develop the backstory and the
> outline for all three episodes.
>
> We then build out that outline to have rough scenes that must
>
> Once this is complete, we can advertise our project and wait for people
Didn't the Snowman Sextet basically do this? (Or, in a less
deliberately connected manner, Narcolepsy?)
That said, it seems like the real model here is less "collaborative
story writing" and more "shared-world anthology". But that moves it
away from what your original goal seems to have been.
--Michael
John
wrote:
>Victor Gijsbers makes some interesting comments about the "open source"
>community and how it might apply to IF in his blog at
>http://gamingphilosopher.blogspot.com/2006/04/three-kinds-of-detrimental-closedness.html
How about "Committee Comp" or something similar? Teams of at least
five or so people collaborate on a piece of IF, and submit it for
judging.
Just an idea...
James Mitchelhill
And this is why it won't work. ;)
You have to be fairly... odd... to want to spend significant amounts of
time working in a medium that is, depending on your point of view, either
all but dead or just in ascendency, with a global audience of a few
thousand at most, and realisticly, an informed audience of maybe a hundred
people, of which perhaps half will play your game.
(Quick aside: Does AIF actually have a larger audience than mainstream IF?)
This requires a fairly strong streak of independence. While there are
occasional works that are produced in collaboration, the majority of IF
games seem to come about as the result of one author's singular vision.
And given that most non-IFComp competitions seem to attract a single
entrant, the chances of finding even one committee to take part are very,
very close to zero.
--
James Mitchelhill
ja...@disorderfeed.net
http://disorderfeed.net