Non-english IF?!

0 views
Skip to first unread message

ler...@classic11.rz.uni-duesseldorf.de

unread,
Dec 11, 1995, 3:00:00 AM12/11/95
to
Saluton!

Well, ''as the subject says'': What about non-english IF?

Up to now, I only know very few examples: this German Zork I,
one about the dutch island Texel in Esperanto(!) and onother
piece mentioned on the Inform web-page. I myself did write
a small adventure in German in ZX-Spectrum and Amiga-Basic,
but it was more of a test for a German parser than a full
blown adventure.

Are some languages more suited to computer-IF than others?
It seems to me that it is more complicated to write a German
parser that ''understands'' language considered relatively
natural than an English one. And Esperanto is claimed to be
easy to process with computers (I think that originally means
'easy to translate automatically). How about other languages?

Also, many IF-writing systems seem to contain inbuilt messages
in English that make them quite useless for writing in other
languages.

What are your opinions on these points?

Ad Astra!
JuL

ler...@sunserver1.rz.uni-duesseldorf.de / Never disturb a dragon, for you will
J"urgen ''JuL'' Lerch / be crunchy and taste good with ketchup!

Mikko Vuorinen

unread,
Dec 11, 1995, 3:00:00 AM12/11/95
to
ler...@classic11.rz.uni-duesseldorf.de () writes:

>Also, many IF-writing systems seem to contain inbuilt messages
>in English that make them quite useless for writing in other
>languages.

Exactly. That's why I develop my own (sucky) game system since I want to
write a Finnish game. That is, I write the whole damn program in C. I can
implement anything I want (if I know how to do it anyway since I'm a
crappy programmer :)

--
)))) (((( ********************************
)) OO `oo'((( * E-mail: vuor...@lut.fi * "Yes."
6 (_) ( ((( * IRC: Dilbon * - William Shakespeare
`____c 8__/((( ********************************

Julian Arnold

unread,
Dec 12, 1995, 3:00:00 AM12/12/95
to
(ler...@classic11.rz.uni-duesseldorf.de) wrote:
> Also, many IF-writing systems seem to contain inbuilt messages
> in English that make them quite useless for writing in other
> languages.

The Aventuro authoring system at gmd is apparently in Esperanto, and I
believe Alan has both English and Swedish libraries (and I seem to remember
there was planned support for other languages). Inform's default messages
should be quite easy to put into another language, by changing
LibraryMessages and Parser() (You would have to change quite a lot in
parser.h if you wanted to catch all the debugging messages though).
--
Jools Arnold jo...@arnod.demon.co.uk


Stephen van Egmond

unread,
Dec 12, 1995, 3:00:00 AM12/12/95
to
Stephen van Egmond <svan...@undergrad.math.uwaterloo.ca> wrote:
>example, Inform allows you to output text to reflect determinism*: ("the

Oops, forgot my footnote.

* I know this word doesn't apply. I made it up.

/Steve

Bjorn Gustavsson

unread,
Dec 12, 1995, 3:00:00 AM12/12/95
to

>
>Are some languages more suited to computer-IF than others?
>It seems to me that it is more complicated to write a German
>parser that ''understands'' language considered relatively
>natural than an English one. And Esperanto is claimed to be
>easy to process with computers (I think that originally means
>'easy to translate automatically). How about other languages?
>

Esperanto would be easy because there is the one ending for
nouns, another for adjectives, and so on, without exceptions.
There is one gender (like English), and one definite article
(again like English), and no indefinite article (easier than
"a" or "an" for English). There is an accusative case, however,
but it is easily derived from the word stem.

Most other languages have more inflexions. But languages
like Swedish or German wouldn't be so hard to do, I think.
What is needed is more properties for each object, for instance
one property to indicate which form of the definite article to use,
one to indicate how the plural is formed, etc. Highly inflected
languages, say Russian or Finnish, would be harder to do well.
Russian has three genders and six cases, conjugates verbs according
to both person and tense, and has a lot of exceptions to the rules.

>Also, many IF-writing systems seem to contain inbuilt messages
>in English that make them quite useless for writing in other
>languages.

Z-Machine interpreters don't have much English messages in them;
only the prompts for filenames when you do save or restore.
It would be easy to change them.

Björn


Stephen van Egmond

unread,
Dec 12, 1995, 3:00:00 AM12/12/95
to
In article <4ahp3d$h...@selene.rz.uni-duesseldorf.de>,

<ler...@classic11.rz.uni-duesseldorf.de> wrote:
>Also, many IF-writing systems seem to contain inbuilt messages
>in English that make them quite useless for writing in other
>languages.

This is indeed true. I don't know anything about German, but for

example, Inform allows you to output text to reflect determinism*: ("the

object" vs. "an object" vs. "a mouse") as well as plural vs. singular
("objects" vs. "object" or "mouse" vs. "mice"). Clearly, these concepts
don't port well to other languages like, say, French where you would need
the masculine and feminine constructs, for example.

Since the code to all aspects of Inform (library and compiler) is
available, they could be implemented, though. Someone just has to do it.

/Steve


Steven E. Hugg

unread,
Dec 14, 1995, 3:00:00 AM12/14/95
to
In article <4almd4$s...@newsbf02.news.aol.com>,
ersat...@aol.com (ErsatzPogo) wrote:
>Other than parser troubles, the most time-consuming task, it would seem to
>me, would be translating all the libraries and message files -- which is
>doable in both TADS and Inform.

So what we need, then, is a language-independent IF-authoring system with an
interlingula at its core, so that every sentence is translated/detranslated on
the fly. Any takers? :-)

---------------------------------------------------------------------
Steven E. Hugg
HAMCO Software


Bjorn Gustavsson

unread,
Dec 14, 1995, 3:00:00 AM12/14/95
to
In article s...@newsbf02.news.aol.com, ersat...@aol.com (ErsatzPogo) writes:
>>Most other languages have more inflexions. But languages
>>like Swedish or German wouldn't be so hard to do, I think.
>>What is needed is more properties for each object, for instance
>>one property to indicate which form of the definite article to use,
>>one to indicate how the plural is formed, etc. Highly inflected
>>languages, say Russian or Finnish, would be harder to do well.
>>Russian has three genders and six cases, conjugates verbs according
>>to both person and tense, and has a lot of exceptions to the rules.
>
>And where in i-f is it necessary to conjugate verbs to anything other
>than the imperative (or whatever that case is called)?
>

When the game describes for the the player what happens. IMO, most of
the complexity in inflected languages would be on the output side, to
ensure that all messages the game prints are grammatically correct.

> throw ball at monkeys
> The monkeys play with the ball for a while.

It that would be translated to Russian, the game has to know the
instrumental case of the object thrown (the ball in this example).
Verb conjugation would in pratice not be much of a problem; only in
past tense if the subject isn't known beforehand.

BTW, Jigsaw, with the unknown genders of the player and Black, would
be difficult to translate into Russian. Any sentence in the past tense
with Black (or the player) as the grammatical subject would reveal
the gender (the verb ending would tell).


-- Bjorn

Reply all
Reply to author
Forward
0 new messages