I’m quite impressed with the tables feature in I7 Inform for handling
conversations.  I’d like to add a parsing feature which trys "show
something to someone" as a default in cases where the something is not
listed in the Table of Replies.  I’m having trouble finding the
correct syntax for this.  Here’s what I’ve tried.

Report asking nikolai about:
                say "Nikolai ignores you.".

Before asking Nikolai about something:
        if second noun is not a Topic listed in the Table of Replies:
                try showing second noun to nikolai.

After asking Nikolai about a Topic listed in the Table of Replies, say
"[Nikolai reply entry][paragraph break]".

Any suggestions?

Doug

Consider the following:

————————————————————–
Disposition is a kind.

happy is a disposition. [not capitalized so that the printed name is
usually not capitalized]
angry is a disposition.

A person has a disposition called current mood.
The current mood of the player is happy.

When play begins:
        change the right hand status line to "([Current mood of the
player])".

Rule for printing the name of a disposition while constructing the
status line:
        say "[printed name]" in title case.
————————————————————–

The point is capitalizing the current mood of the player in the status
bar, but not capitalizing it in other contexts.

When running, I get the following printed in the status bar:

Run-Time Problem P10: Since  Is Not Allowed The Property "Printed
Name", It Is Against The Rules To Try To Use It.

The problem is solved if I drop the last rule in the above code, but
then the mood is not capitalized in the status bar (even though I did
capitalize the text substitution – isn’t that supposed to do the
trick? would something else do it?).

It is also solved if I make disposition a kind of thing (instead of
object), but I’d rather not make "things" of my abstract concepts,
because I have quite a few such concepts, and I’m using some various-
various relations over all things.

Speaking of which, is there a way to limit my various-various
relations to containers and supporters only?

Thanks for any input you may have,
Yaron

Hello there,

I need your help to solve a problem related to disambiguation : how
can we "force" the interpreter to ask for a disambiguation ? For
example, I’ve noticed that when a noun is referring both to an object
carried by the player and to an object which is in the room, the
interpreter always "understands" that the player is referring to the
object carried by the player : how can I do to make the computer ask
the player which object he is referring to ?

Thanks for your help !
Mule hollandaise

Hi,

I’ve posted a few times in my attempts to get a build environment working
to allow me to develop a version 3 Z-machine game, which can be used with
Z-Machine interpreters with limited resources, such as those for retro
platforms like the old 8-bit computers.

Anyway, I’ve put together an archive containing a complete build
environment for Windows, along the lines of the one Roger Firth provides
alongside the Inform Beginner’s Guide.

http://www.retrosoftware.co.uk/informz3/

Basically, extract the archive available on the site above and there’s an
example v3 game you can modify and then compile and execute by running a
single batch file.

This version includes the BBC Micro Z-machine interpreter our community
has been developing, but it shouldn’t take too much effort to drop in an
alternative Windows interpreter, should you prefer.  You may still
appreciate the archive’s compiled older Inform 6.15 executable which
still compiles v3 games.

If anyone makes any use of this and starts to develop a v3 game, we’d
love to hear about here:

http://www.retrosoftware.co.uk/forum/viewforum.php?f=30

And, indeed, can provide hosting and feedback, if required.

I’ve started my own v3 game, and will post back here when I’ve made any
significant progress.

Cheers,

Sam.

"Test" by Emiliano Short

The Area is a room.

A book is a kind of thing.

In the Area are ten books and the shelf.

At 9:01 AM: try taking the shelf.

[At 9:02 AM: try taking one book.]

———————————————–

The last line of that code prevents it from compiling (if it isn’t
commented out), because "one book" is not a specific thing.  However,
the player is able to type "TAKE ONE BOOK", no problemo.  What work-
around is there that would allow me, as the in-game god, to do what
the player can do?

This is a bit of a vague issue at this point and I’m just throwing
this out there in case others have noticed it.

The issue I’m seeing seems to be prevalent on Windows Vista, not so
much on Windows XP (at least from what I can see). Here are the specs
of one such machine:

Intel Core Duo CPU 2.00 GHz; 4 GB of RAM; Windows Vista SP1

Here is the source text:

[I7 Source Start]
"Initial Screenplay" by Jeff Nyman

The Stage is a room.

Rule for printing the banner text: do nothing.
Rule for printing the name of the Stage: do nothing.
[I7 Source End]

Just to compile that on the above spec’d machine takes about 10
seconds. The more rules I add, the worse this gets. ("Patient Zero"
from the manual takes about 15 seconds to compile. Larger games, like
the source for "Child’s Play" or games that use a lot of extensions,
can take up to 45 seconds.) This is as opposed to Windows XP, where
processing is much quicker. (This problem occurs on all Windows Vista
configurations that I’ve seen.)

This is making it extremely painful to do anything once we get works
of a large enough size, particularly as we often do a lot of "add/
refactor-then-compile" steps.

I’ve also run the compile steps directly through a command line and
approximately the same delay occurs, at least from what I can see. I
wasn’t sure if this was an artifact of the IDE itself or with how
ni.exe is executing. It appears to be the latter.

So the problem seems to be relatively localized to ni.exe executing on
Windows Vista.

The only thing I can say is that when I ran a profiler during ni.exe
execution, I notice it would be *slightly* worse in processing time if
an instance of iexplore.exe (Internet Explorer) was also open. If that
instance was closed, the "slightly worse" just went back to being
"worse." I’m not sure at all why that would be the case. (This would
happen with an Internet Explorer that was just opened — thus a
smaller memory footprint — as well as one that had been left open for
some time — thus with a larger memory footprint.) This same effect
would not happen with other applications, like Word or Acrobat being
open.

I’m not asking for a solution as I don’t necessarily think there is
one; I just wanted to point out this potential problem, at least for
people with Windows Vista.

- Jeff

5% paypal handling charge supports the online payment! PayPal
cheap  Guess wallet
Dear friend
welcome to shopping on www.Shoes-paypal.cn
1.5% paypal handling charge supports the online payment!
2.Use your intergla replacement more good gift!
3.notes by email and website of deliver each package at first time.

I’m wondering how much like an enumeratio a new kind of value is.

say I define heat as

Heat is a new Kind of Value. Heats are frozen, cold, luke warm, warm,
very warm, hot, scorching.

I know from the examples that I can reffer to the heat before cold

or the heat after warm.

but is there a way to compare two heats. How would I define hotter and
colder ?

so I could say

The ice is a thing. The heat of the ice is frozen.
The water is a thing. The heat of the water is warm.

ever turn when the Ice is colder then the water:
    <<do something >>

Again thinking of kinds of values as enumarations I would expect to be
able to use normal equality operators but this does not work so I’m
wondering if there any other phrasing that will allow me to treat a
kind of value as an ordinal scale.

regards

Konrad.

hallo there can you all comment about my site.. and what should i add
to
make my site looking goods. thanks- butterflygirl- http://butterfly-friends.page.tl

I’m trying to build a system in my I7 game where a large number of
things exist which are related to each other, but only in small
numbers. I’ve decided to do this by using a table (to define the
things) and lists, rather than relationships, since I don’t need to
worry about the possibility that Thing A might suddenly become related
to Thing Z. Whenever I try to test this, though, I get an error.
Here’s some simple code which exemplifies the problem:

#begin code

"Test" by Tiefling

A monarch is a kind of thing. Some monarchs are defined by the Table
of Monarchy.

Table of Monarchy
monarch offspring
John    {}
Richard I       {}
Henry II        {Richard I, John}
Matilda {Henry II}
Henry I {Matilda}
William II      {}
William I       {William II, Henry I}

#end code

And here’s the error messages:

Problem. In ‘Some monarchs are defined by the Table of Monarchy’   you
give a value of the offspring property as ‘{ Richard I, John }’, but
this ought to be a list of a complex value of some sort, not a list of
monarchs.

Problem. In ‘Some monarchs are defined by the Table of Monarchy’   you
give a value of the offspring property as ‘{ Henry II }’, but this
ought to be a list of a complex value of some sort, not a list of
monarchs.

Problem. In ‘Some monarchs are defined by the Table of Monarchy’   you
give a value of the offspring property as ‘{ Matilda }’, but this
ought to be a list of a complex value of some sort, not a list of
monarchs.

Problem. In ‘Some monarchs are defined by the Table of Monarchy’   you
give a value of the offspring property as ‘{ William II, Henry I }’,
but this ought to be a list of a complex value of some sort, not a
list of monarchs.

I’ve no idea what ‘a list of a complex value’ looks like. The
intention is very much that every entry in the ‘offspring’ column
should be a (possibly null) list of monarchs.

Can anyone help me untangle this?

Many thanks,

T.