Prepare to release sgt-puzzles (20170606.272beef-1).

[sgt-puzzles.git] / puzzles.but

\title Simon Tatham's Portable Puzzle Collection

\cfg{winhelp-filename}{puzzles.hlp}
\cfg{winhelp-contents-titlepage}{Contents}

\cfg{text-filename}{puzzles.txt}

\cfg{html-contents-filename}{index.html}
\cfg{html-template-filename}{%k.html}
\cfg{html-index-filename}{docindex.html}
\cfg{html-leaf-level}{1}
\cfg{html-contents-depth-0}{1}
\cfg{html-contents-depth-1}{2}
\cfg{html-leaf-contains-contents}{true}

\cfg{chm-filename}{puzzles.chm}
\cfg{chm-contents-filename}{index.html}
\cfg{chm-template-filename}{%k.html}
\cfg{chm-head-end}{<link rel="stylesheet" type="text/css" href="chm.css">}
\cfg{chm-extra-file}{chm.css}

\cfg{info-filename}{puzzles.info}

\cfg{ps-filename}{puzzles.ps}
\cfg{pdf-filename}{puzzles.pdf}

\define{by} \u00D7{x}

\define{dash} \u2013{-}

\define{times} \u00D7{*}

\define{divide} \u00F7{/}

\define{minus} \u2212{-}

This is a collection of small one-player puzzle games.

\copyright This manual is copyright 2004-2014 Simon Tatham. All rights
reserved. You may distribute this documentation under the MIT licence.
See \k{licence} for the licence text in full.

\cfg{html-local-head}{<meta name="AppleTitle" content="Puzzles Help">}

\C{intro} Introduction

I wrote this collection because I thought there should be more small
desktop toys available: little games you can pop up in a window and
play for two or three minutes while you take a break from whatever
else you were doing. And I was also annoyed that every time I found
a good game on (say) \i{Unix}, it wasn't available the next time I
was sitting at a \i{Windows} machine, or vice versa; so I arranged
that everything in my personal puzzle collection will happily run on
both, and have more recently done a port to \i{Mac OS X} as well. When I
find (or perhaps invent) further puzzle games that I like, they'll
be added to this collection and will immediately be available on
both platforms. And if anyone feels like writing any other front
ends \dash PocketPC, Mac OS pre-10, or whatever it might be \dash
then all the games in this framework will immediately become
available on another platform as well.

The actual games in this collection were mostly not my invention; they
are re-implementations of existing game concepts within my portable
puzzle framework. I do not claim credit, in general, for inventing the
rules of any of these puzzles. (I don't even claim authorship of all
the code; some of the puzzles have been submitted by other authors.)

This collection is distributed under the \i{MIT licence} (see
\k{licence}). This means that you can do pretty much anything you like
with the game binaries or the code, except pretending you wrote them
yourself, or suing me if anything goes wrong. 

The most recent versions, and \i{source code}, can be found at
\I{website}\W{https://www.chiark.greenend.org.uk/~sgtatham/puzzles/}\cw{https://www.chiark.greenend.org.uk/~sgtatham/puzzles/}.

Please report \I{feedback}\i{bugs} to
\W{mailto:anakin@pobox.com}\cw{anakin@pobox.com}.
You might find it helpful to read this article before reporting a bug:

\W{https://www.chiark.greenend.org.uk/~sgtatham/bugs.html}\cw{https://www.chiark.greenend.org.uk/~sgtatham/bugs.html}

\ii{Patches} are welcome. Especially if they provide a new front end
(to make all these games run on another platform), or a new game.


\C{common} \ii{Common features}

This chapter describes features that are common to all the games.

\H{common-actions} \I{controls}Common actions

These actions are all available from the \I{Game menu}\q{Game} menu
and via \I{keys}keyboard shortcuts, in addition to any game-specific
actions.

(On \i{Mac OS X}, to conform with local user interface standards, these
actions are situated on the \I{File menu}\q{File} and \I{Edit
menu}\q{Edit} menus instead.)

\dt \ii\e{New game} (\q{N}, Ctrl+\q{N})

\dd Starts a new game, with a random initial state.

\dt \ii\e{Restart game}

\dd Resets the current game to its initial state. (This can be undone.)

\dt \ii\e{Load}

\dd Loads a saved game from a file on disk.

\dt \ii\e{Save}

\dd Saves the current state of your game to a file on disk.

\lcont{

The Load and Save operations preserve your entire game
history (so you can save, reload, and still Undo and Redo things you
had done before saving).

}

\dt \I{printing, on Windows}\e{Print}

\dd Where supported (currently only on Windows), brings up a dialog
allowing you to print an arbitrary number of puzzles randomly
generated from the current parameters, optionally including the
current puzzle. (Only for puzzles which make sense to print, of
course \dash it's hard to think of a sensible printable representation
of Fifteen!)

\dt \ii\e{Undo} (\q{U}, Ctrl+\q{Z}, Ctrl+\q{_})

\dd Undoes a single move. (You can undo moves back to the start of the
session.)

\dt \ii\e{Redo} (\q{R}, Ctrl+\q{R})

\dd Redoes a previously undone move.

\dt \ii\e{Copy}

\dd Copies the current state of your game to the clipboard in text
format, so that you can paste it into (say) an e-mail client or a
web message board if you're discussing the game with someone else.
(Not all games support this feature.)

\dt \ii\e{Solve}

\dd Transforms the puzzle instantly into its solved state. For some
games (Cube) this feature is not supported at all because it is of
no particular use. For other games (such as Pattern), the solved
state can be used to give you information, if you can't see how a
solution can exist at all or you want to know where you made a
mistake. For still other games (such as Sixteen), automatic solution
tells you nothing about how to \e{get} to the solution, but it does
provide a useful way to get there quickly so that you can experiment
with set-piece moves and transformations.

\lcont{

Some games (such as Solo) are capable of solving a game ID you have
typed in from elsewhere. Other games (such as Rectangles) cannot
solve a game ID they didn't invent themself, but when they did
invent the game ID they know what the solution is already. Still
other games (Pattern) can solve \e{some} external game IDs, but only
if they aren't too difficult.

The \q{Solve} command adds the solved state to the end of the undo
chain for the puzzle. In other words, if you want to go back to
solving it yourself after seeing the answer, you can just press Undo.

}

\dt \I{exit}\ii\e{Quit} (\q{Q}, Ctrl+\q{Q})

\dd Closes the application entirely.

\H{common-id} Specifying games with the \ii{game ID}

There are two ways to save a game specification out of a puzzle and
recreate it later, or recreate it in somebody else's copy of the
same puzzle.

The \q{\i{Specific}} and \q{\i{Random Seed}} options from the
\I{Game menu}\q{Game} menu (or the \q{File} menu, on \i{Mac OS X}) each
show a piece of text (a \q{game ID}) which is sufficient to
reconstruct precisely the same game at a later date.

You can enter either of these pieces of text back into the program
(via the same \q{Specific} or \q{Random Seed} menu options) at a
later point, and it will recreate the same game. You can also use
either one as a \i{command line} argument (on Windows or Unix); see
\k{common-cmdline} for more detail.

The difference between the two forms is that a descriptive game ID
is a literal \e{description} of the \i{initial state} of the game,
whereas a random seed is just a piece of arbitrary text which was
provided as input to the random number generator used to create the
puzzle. This means that:

\b Descriptive game IDs tend to be longer in many puzzles (although
some, such as Cube (\k{cube}), only need very short descriptions).
So a random seed is often a \e{quicker} way to note down the puzzle
you're currently playing, or to tell it to somebody else so they can
play the same one as you.

\b Any text at all is a valid random seed. The automatically
generated ones are fifteen-digit numbers, but anything will do; you
can type in your full name, or a word you just made up, and a valid
puzzle will be generated from it. This provides a way for two or
more people to race to complete the same puzzle: you think of a
random seed, then everybody types it in at the same time, and nobody
has an advantage due to having seen the generated puzzle before
anybody else.

\b It is often possible to convert puzzles from other sources (such
as \q{nonograms} or \q{sudoku} from newspapers) into descriptive
game IDs suitable for use with these programs.

\b Random seeds are not guaranteed to produce the same result if you
use them with a different \i\e{version} of the puzzle program. This
is because the generation algorithm might have been improved or
modified in later versions of the code, and will therefore produce a
different result when given the same sequence of random numbers. Use
a descriptive game ID if you aren't sure that it will be used on the
same version of the program as yours.

\lcont{(Use the \q{About} menu option to find out the version number
of the program. Programs with the same version number running on
different platforms should still be random-seed compatible.)}

\I{ID format}A descriptive game ID starts with a piece of text which
encodes the \i\e{parameters} of the current game (such as grid
size). Then there is a colon, and after that is the description of
the game's initial state. A random seed starts with a similar string
of parameters, but then it contains a hash sign followed by
arbitrary data.

If you enter a descriptive game ID, the program will not be able to
show you the random seed which generated it, since it wasn't
generated \e{from} a random seed. If you \e{enter} a random seed,
however, the program will be able to show you the descriptive game
ID derived from that random seed.

Note that the game parameter strings are not always identical
between the two forms. For some games, there will be parameter data
provided with the random seed which is not included in the
descriptive game ID. This is because that parameter information is
only relevant when \e{generating} puzzle grids, and is not important
when playing them. Thus, for example, the difficulty level in Solo
(\k{solo}) is not mentioned in the descriptive game ID.

These additional parameters are also not set permanently if you type
in a game ID. For example, suppose you have Solo set to \q{Advanced}
difficulty level, and then a friend wants your help with a
\q{Trivial} puzzle; so the friend reads out a random seed specifying
\q{Trivial} difficulty, and you type it in. The program will
generate you the same \q{Trivial} grid which your friend was having
trouble with, but once you have finished playing it, when you ask
for a new game it will automatically go back to the \q{Advanced}
difficulty which it was previously set on.

\H{common-type} The \q{Type} menu

The \I{Type menu}\q{Type} menu, if present, may contain a list of
\i{preset} game settings. Selecting one of these will start a new
random game with the parameters specified.

The \q{Type} menu may also contain a \q{\i{Custom}} option which
allows you to fine-tune game \i{parameters}. The parameters
available are specific to each game and are described in the
following sections.

\H{common-cmdline} Specifying game parameters on the \i{command line}

(This section does not apply to the \i{Mac OS X} version.)

The games in this collection deliberately do not ever save
information on to the computer they run on: they have no high score
tables and no saved preferences. (This is because I expect at least
some people to play them at work, and those people will probably
appreciate leaving as little evidence as possible!)

However, if you do want to arrange for one of these games to
\I{default parameters, specifying}default to a particular set of
parameters, you can specify them on the command line.

The easiest way to do this is to set up the parameters you want
using the \q{Type} menu (see \k{common-type}), and then to select
\q{Random Seed} from the \q{Game} or \q{File} menu (see
\k{common-id}). The text in the \q{Game ID} box will be composed of
two parts, separated by a hash. The first of these parts represents
the game parameters (the size of the playing area, for example, and
anything else you set using the \q{Type} menu).

If you run the game with just that parameter text on the command
line, it will start up with the settings you specified.

For example: if you run Cube (see \k{cube}), select \q{Octahedron}
from the \q{Type} menu, and then go to the game ID selection, you
will see a string of the form \cq{o2x2#338686542711620}. Take only
the part before the hash (\cq{o2x2}), and start Cube with that text
on the command line: \cq{PREFIX-cube o2x2}.

If you copy the \e{entire} game ID on to the command line, the game
will start up in the specific game that was described. This is
occasionally a more convenient way to start a particular game ID
than by pasting it into the game ID selection box.

(You could also retrieve the encoded game parameters using the
\q{Specific} menu option instead of \q{Random Seed}, but if you do
then some options, such as the difficulty level in Solo, will be
missing. See \k{common-id} for more details on this.)

\H{common-unix-cmdline} \i{Unix} \i{command-line} options

(This section only applies to the Unix port.)

In addition to being able to specify game parameters on the command
line (see \k{common-cmdline}), there are various other options:

\dt \cw{--game}

\dt \cw{--load}

\dd These options respectively determine whether the command-line
argument is treated as specifying game parameters or a \i{save} file
to \i{load}. Only one should be specified. If neither of these options
is specified, a guess is made based on the format of the argument.

\dt \cw{--generate }\e{n}

\dd If this option is specified, instead of a puzzle being displayed,
a number of descriptive game IDs will be \I{generating game IDs}invented
and printed on standard output. This is useful for gaining access to
the game generation algorithms without necessarily using the frontend.

\lcont{

If game parameters are specified on the command-line, they will be
used to generate the game IDs; otherwise a default set of parameters
will be used.

The most common use of this option is in conjunction with \c{--print},
in which case its behaviour is slightly different; see below.

}

\dt \I{printing, on Unix}\cw{--print }\e{w}\cw{x}\e{h}

\dd If this option is specified, instead of a puzzle being displayed,
a printed representation of one or more unsolved puzzles is sent to
standard output, in \i{PostScript} format.

\lcont{

On each page of puzzles, there will be \e{w} across and \e{h} down. If
there are more puzzles than \e{w}\by\e{h}, more than one page will be
printed.

If \c{--generate} has also been specified, the invented game IDs will
be used to generate the printed output. Otherwise, a list of game IDs
is expected on standard input (which can be descriptive or random
seeds; see \k{common-id}), in the same format produced by
\c{--generate}.

For example:

\c PREFIX-net --generate 12 --print 2x3 7x7w | lpr

will generate two pages of printed Net puzzles (each of which will
have a 7\by\.7 wrapping grid), and pipe the output to the \c{lpr}
command, which on many systems will send them to an actual printer.

There are various other options which affect printing; see below.

}

\dt \cw{--save }\e{file-prefix} [ \cw{--save-suffix }\e{file-suffix} ]

\dd If this option is specified, instead of a puzzle being
displayed, saved-game files for one or more unsolved puzzles are
written to files constructed from the supplied prefix and/or suffix.

\lcont{

If \c{--generate} has also been specified, the invented game IDs will
be used to generate the printed output. Otherwise, a list of game IDs
is expected on standard input (which can be descriptive or random
seeds; see \k{common-id}), in the same format produced by
\c{--generate}.

For example:

\c PREFIX-net --generate 12 --save game --save-suffix .sav

will generate twelve Net saved-game files with the names
\cw{game0.sav} to \cw{game11.sav}.

}

\dt \cw{--version}

\dd Prints version information about the game, and then quits.

The following options are only meaningful if \c{--print} is also
specified:

\dt \cw{--with-solutions}

\dd The set of pages filled with unsolved puzzles will be followed by
the solutions to those puzzles.

\dt \cw{--scale }\e{n}

\dd Adjusts how big each puzzle is when printed. Larger numbers make
puzzles bigger; the default is 1.0.

\dt \cw{--colour}

\dd Puzzles will be printed in colour, rather than in black and white
(if supported by the puzzle).


\C{net} \i{Net}

\cfg{winhelp-topic}{games.net}

(\e{Note:} the \i{Windows} version of this game is called
\i\cw{NETGAME.EXE} to avoid clashing with Windows's own \cw{NET.EXE}.)

I originally saw this in the form of a Flash game called \i{FreeNet}
\k{FreeNet}, written by Pavils Jurjans; there are several other
implementations under the name \i{NetWalk}. The computer prepares a
network by connecting up the centres of squares in a grid, and then
shuffles the network by rotating every tile randomly. Your job is to
rotate it all back into place. The successful solution will be an
entirely connected network, with no closed loops. \#{The latter
clause means that there are no closed paths within the network.
Could this be clearer? "No closed paths"?} As a visual aid,
all tiles which are connected to the one in the middle are
highlighted. 

\B{FreeNet} \W{http://www.jurjans.lv/stuff/net/FreeNet.htm}\cw{http://www.jurjans.lv/stuff/net/FreeNet.htm}

\H{net-controls} \i{Net controls}

\IM{Net controls} controls, for Net
\IM{Net controls} keys, for Net
\IM{Net controls} shortcuts (keyboard), for Net

This game can be played with either the keyboard or the mouse. The
controls are:

\dt \e{Select tile}: mouse pointer, arrow keys

\dt \e{Rotate tile anticlockwise}: left mouse button, \q{A} key

\dt \e{Rotate tile clockwise}: right mouse button, \q{D} key

\dt \e{Rotate tile by 180 degrees}: \q{F} key

\dt \e{Lock (or unlock) tile}: middle mouse button, shift-click, \q{S} key

\dd You can lock a tile once you're sure of its orientation. You can
also unlock it again, but while it's locked you can't accidentally
turn it.

The following controls are not necessary to complete the game, but may
be useful:

\dt \e{Shift grid}: Shift + arrow keys

\dd On grids that wrap, you can move the origin of the grid, so that
tiles that were on opposite sides of the grid can be seen together.

\dt \e{Move centre}: Ctrl + arrow keys

\dd You can change which tile is used as the source of highlighting.
(It doesn't ultimately matter which tile this is, as every tile will
be connected to every other tile in a correct solution, but it may be
helpful in the intermediate stages of solving the puzzle.)

\dt \e{Jumble tiles}: \q{J} key

\dd This key turns all tiles that are not locked to random
orientations.

(All the actions described in \k{common-actions} are also available.)

\H{net-params} \I{parameters, for Net}Net parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in tiles.

\dt \e{Walls wrap around}

\dd If checked, flow can pass from the left edge to the right edge,
and from top to bottom, and vice versa.

\dt \e{Barrier probability}

\dd A number between 0.0 and 1.0 controlling whether an immovable
barrier is placed between two tiles to prevent flow between them (a
higher number gives more barriers). Since barriers are immovable, they
act as constraints on the solution (i.e., hints).

\lcont{

The grid generation in Net has been carefully arranged so that the
barriers are independent of the rest of the grid. This means that if
you note down the random seed used to generate the current puzzle
(see \k{common-id}), change the \e{Barrier probability} parameter,
and then re-enter the same random seed, you should see exactly the
same starting grid, with the only change being the number of
barriers. So if you're stuck on a particular grid and need a hint,
you could start up another instance of Net, set up the same
parameters but a higher barrier probability, and enter the game seed
from the original Net window.

}

\dt \e{Ensure unique solution}

\dd Normally, Net will make sure that the puzzles it presents have
only one solution. Puzzles with ambiguous sections can be more
difficult and more subtle, so if you like you can turn off this
feature and risk having ambiguous puzzles. (Also, finding \e{all}
the possible solutions can be an additional challenge for an
advanced player.)


\C{cube} \i{Cube}

\cfg{winhelp-topic}{games.cube}

This is another one I originally saw as a web game. This one was a
Java game \k{cube-java-game}, by Paul Scott. You have a grid of 16
squares, six of which are blue; on one square rests a cube. Your move
is to use the arrow keys to roll the cube through 90 degrees so that
it moves to an adjacent square. If you roll the cube on to a blue
square, the blue square is picked up on one face of the cube; if you
roll a blue face of the cube on to a non-blue square, the blueness is
put down again. (In general, whenever you roll the cube, the two faces
that come into contact swap colours.) Your job is to get all six blue
squares on to the six faces of the cube at the same time. Count your
moves and try to do it in as few as possible. 

Unlike the original Java game, my version has an additional feature:
once you've mastered the game with a cube rolling on a square grid,
you can change to a triangular grid and roll any of a tetrahedron, an
octahedron or an icosahedron. 

\B{cube-java-game} \W{http://www3.sympatico.ca/paulscott/cube/cube.htm}\cw{http://www3.sympatico.ca/paulscott/cube/cube.htm}

\H{cube-controls} \i{Cube controls}

\IM{Cube controls} controls, for Cube
\IM{Cube controls} keys, for Cube
\IM{Cube controls} shortcuts (keyboard), for Cube

This game can be played with either the keyboard or the mouse.

Left-clicking anywhere on the window will move the cube (or other
solid) towards the mouse pointer.

The arrow keys can also used to roll the cube on its square grid in
the four cardinal directions.
On the triangular grids, the mapping of arrow keys to directions is
more approximate. Vertical movement is disallowed where it doesn't
make sense. The four keys surrounding the arrow keys on the numeric
keypad (\q{7}, \q{9}, \q{1}, \q{3}) can be used for diagonal movement.

(All the actions described in \k{common-actions} are also available.)

\H{cube-params} \I{parameters, for Cube}Cube parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Type of solid}

\dd Selects the solid to roll (and hence the shape of the grid):
tetrahedron, cube, octahedron, or icosahedron.

\dt \e{Width / top}, \e{Height / bottom}

\dd On a square grid, horizontal and vertical dimensions. On a
triangular grid, the number of triangles on the top and bottom rows
respectively.


\C{fifteen} \i{Fifteen}

\cfg{winhelp-topic}{games.fifteen}

The old ones are the best: this is the good old \q{\i{15-puzzle}}
with sliding tiles. You have a 4\by\.4 square grid; 15 squares
contain numbered tiles, and the sixteenth is empty. Your move is to
choose a tile next to the empty space, and slide it into the space.
The aim is to end up with the tiles in numerical order, with the
space in the bottom right (so that the top row reads 1,2,3,4 and the
bottom row reads 13,14,15,\e{space}).

\H{fifteen-controls} \i{Fifteen controls}

\IM{Fifteen controls} controls, for Fifteen
\IM{Fifteen controls} keys, for Fifteen
\IM{Fifteen controls} shortcuts (keyboard), for Fifteen

This game can be controlled with the mouse or the keyboard.

A left-click with the mouse in the row or column containing the empty
space will move as many tiles as necessary to move the space to the
mouse pointer.

The arrow keys will move a tile adjacent to the space in the direction
indicated (moving the space in the \e{opposite} direction).

Pressing \q{h} will make a suggested move.  Pressing \q{h} enough
times will solve the game, but it may scramble your progress while
doing so.

(All the actions described in \k{common-actions} are also available.)

\H{fifteen-params} \I{parameters, for Fifteen}Fifteen parameters

The only options available from the \q{Custom...} option on the \q{Type}
menu are \e{Width} and \e{Height}, which are self-explanatory. (Once
you've changed these, it's not a \q{15-puzzle} any more, of course!)


\C{sixteen} \i{Sixteen}

\cfg{winhelp-topic}{games.sixteen}

Another sliding tile puzzle, visually similar to Fifteen (see
\k{fifteen}) but with a different type of move. This time, there is no
hole: all 16 squares on the grid contain numbered squares. Your move
is to shift an entire row left or right, or shift an entire column up
or down; every time you do that, the tile you shift off the grid
re-appears at the other end of the same row, in the space you just
vacated. To win, arrange the tiles into numerical order (1,2,3,4 on
the top row, 13,14,15,16 on the bottom). When you've done that, try
playing on different sizes of grid. 

I \e{might} have invented this game myself, though only by accident if
so (and I'm sure other people have independently invented it). I
thought I was imitating a screensaver I'd seen, but I have a feeling
that the screensaver might actually have been a Fifteen-type puzzle
rather than this slightly different kind. So this might be the one
thing in my puzzle collection which represents creativity on my part
rather than just engineering.

\H{sixteen-controls} \I{controls, for Sixteen}Sixteen controls

Left-clicking on an arrow will move the appropriate row or column in
the direction indicated.  Right-clicking will move it in the opposite
direction.

Alternatively, use the cursor keys to move the position indicator
around the edge of the grid, and use the return key to move the
row/column in the direction indicated. 

You can also move the tiles directly.  Move the cursor onto a tile,
hold Control and press an arrow key to move the tile under the
cursor and move the cursor along with the tile.  Or, hold Shift to
move only the tile.  Pressing Enter simulates holding down Control
(press Enter again to release), while pressing Space simulates
holding down shift.

(All the actions described in \k{common-actions} are also available.)

\H{sixteen-params} \I{parameters, for Sixteen}Sixteen parameters

The parameters available from the \q{Custom...} option on the
\q{Type} menu are:

\b \e{Width} and \e{Height}, which are self-explanatory.

\b You can ask for a limited shuffling operation to be performed on
the grid. By default, Sixteen will shuffle the grid in such a way
that any arrangement is about as probable as any other. You can
override this by requesting a precise number of shuffling moves to
be performed. Typically your aim is then to determine the precise
set of shuffling moves and invert them exactly, so that you answer
(say) a four-move shuffle with a four-move solution. Note that the
more moves you ask for, the more likely it is that solutions shorter
than the target length will turn out to be possible.


\C{twiddle} \i{Twiddle}

\cfg{winhelp-topic}{games.twiddle}

Twiddle is a tile-rearrangement puzzle, visually similar to Sixteen
(see \k{sixteen}): you are given a grid of square tiles, each
containing a number, and your aim is to arrange the numbers into
ascending order.

In basic Twiddle, your move is to rotate a square group of four
tiles about their common centre. (Orientation is not significant in
the basic puzzle, although you can select it.) On more advanced
settings, you can rotate a larger square group of tiles.

I first saw this type of puzzle in the GameCube game \q{Metroid
Prime 2}. In the Main Gyro Chamber in that game, there is a puzzle
you solve to unlock a door, which is a special case of Twiddle. I
developed this game as a generalisation of that puzzle.

\H{twiddle-controls} \I{controls, for Twiddle}Twiddle controls

To play Twiddle, click the mouse in the centre of the square group
you wish to rotate. In the basic mode, you rotate a 2\by\.2 square,
which means you have to click at a corner point where four tiles
meet.

In more advanced modes you might be rotating 3\by\.3 or even more at
a time; if the size of the square is odd then you simply click in
the centre tile of the square you want to rotate.

Clicking with the left mouse button rotates the group anticlockwise.
Clicking with the right button rotates it clockwise.

You can also move an outline square around the grid with the cursor
keys; the square is the size above (2\by\.2 by default, or larger).
Pressing the return key or space bar will rotate the current square
anticlockwise or clockwise respectively.

(All the actions described in \k{common-actions} are also available.)

\H{twiddle-parameters} \I{parameters, for Twiddle}Twiddle parameters

Twiddle provides several configuration options via the \q{Custom}
option on the \q{Type} menu:

\b You can configure the width and height of the puzzle grid.

\b You can configure the size of square block that rotates at a time.

\b You can ask for every square in the grid to be distinguishable
(the default), or you can ask for a simplified puzzle in which there
are groups of identical numbers. In the simplified puzzle your aim
is just to arrange all the 1s into the first row, all the 2s into
the second row, and so on.

\b You can configure whether the orientation of tiles matters. If
you ask for an orientable puzzle, each tile will have a triangle
drawn in it. All the triangles must be pointing upwards to complete
the puzzle.

\b You can ask for a limited shuffling operation to be performed on
the grid. By default, Twiddle will shuffle the grid so much that any
arrangement is about as probable as any other. You can override this
by requesting a precise number of shuffling moves to be performed.
Typically your aim is then to determine the precise set of shuffling
moves and invert them exactly, so that you answer (say) a four-move
shuffle with a four-move solution. Note that the more moves you ask
for, the more likely it is that solutions shorter than the target
length will turn out to be possible.


\C{rect} \i{Rectangles}

\cfg{winhelp-topic}{games.rectangles}

You have a grid of squares, with numbers written in some (but not all)
of the squares. Your task is to subdivide the grid into rectangles of
various sizes, such that (a) every rectangle contains exactly one
numbered square, and (b) the area of each rectangle is equal to the
number written in its numbered square.

Credit for this game goes to the Japanese puzzle magazine \i{Nikoli}
\k{nikoli-rect}; I've also seen a Palm implementation at \i{Puzzle
Palace} \k{puzzle-palace-rect}. Unlike Puzzle Palace's
implementation, my version automatically generates random grids of
any size you like. The quality of puzzle design is therefore not
quite as good as hand-crafted puzzles would be, but on the plus side
you get an inexhaustible supply of puzzles tailored to your own
specification.

\B{nikoli-rect} \W{http://www.nikoli.co.jp/en/puzzles/shikaku.html}\cw{http://www.nikoli.co.jp/en/puzzles/shikaku.html}
(beware of Flash)

\B{puzzle-palace-rect} \W{https://web.archive.org/web/20041024001459/http://www.puzzle.gr.jp/puzzle/sikaku/palm/index.html.en}\cw{https://web.archive.org/web/20041024001459/http://www.puzzle.gr.jp/puzzle/sikaku/palm/index.html.en}

\H{rectangles-controls} \I{controls, for Rectangles}Rectangles controls

This game is played with the mouse or cursor keys.

Left-click any edge to toggle it on or off, or left-click and drag to draw
an entire rectangle (or line) on the grid in one go (removing any existing
edges within that rectangle). Right-clicking and dragging will allow you
to erase the contents of a rectangle without affecting its edges. 

Alternatively, use the cursor keys to move the position indicator
around the board. Pressing the return key then allows you to use the
cursor keys to drag a rectangle out from that position, and pressing
the return key again completes the rectangle. Using the space bar
instead of the return key allows you to erase the contents of a
rectangle without affecting its edges, as above. Pressing escape
cancels a drag.

When a rectangle of the correct size is completed, it will be shaded.

(All the actions described in \k{common-actions} are also available.)

\H{rectangles-params} \I{parameters, for Rectangles}Rectangles parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid, in squares.

\dt \e{Expansion factor}

\dd This is a mechanism for changing the type of grids generated by
the program. Some people prefer a grid containing a few large
rectangles to one containing many small ones. So you can ask
Rectangles to essentially generate a \e{smaller} grid than the size
you specified, and then to expand it by adding rows and columns.

\lcont{

The default expansion factor of zero means that Rectangles will
simply generate a grid of the size you ask for, and do nothing
further. If you set an expansion factor of (say) 0.5, it means that
each dimension of the grid will be expanded to half again as big
after generation. In other words, the initial grid will be 2/3 the
size in each dimension, and will be expanded to its full size
without adding any more rectangles.

Setting an expansion factor of around 0.5 tends to make the game
more difficult, and also (in my experience) rewards a less deductive
and more intuitive playing style. If you set it \e{too} high,
though, the game simply cannot generate more than a few rectangles
to cover the entire grid, and the game becomes trivial.

}

\dt \e{Ensure unique solution}

\dd Normally, Rectangles will make sure that the puzzles it presents
have only one solution. Puzzles with ambiguous sections can be more
difficult and more subtle, so if you like you can turn off this
feature and risk having ambiguous puzzles. Also, finding \e{all} the
possible solutions can be an additional challenge for an advanced
player. Turning off this option can also speed up puzzle generation.


\C{netslide} \i{Netslide}

\cfg{winhelp-topic}{games.netslide}

This game combines the grid generation of Net (see \k{net}) with the
movement of Sixteen (see \k{sixteen}): you have a Net grid, but
instead of rotating tiles back into place you have to slide them
into place by moving a whole row at a time. 

As in Sixteen, \I{controls, for Netslide}control is with the mouse or
cursor keys. See \k{sixteen-controls}.

\I{parameters, for Netslide}The available game parameters have similar
meanings to those in Net (see \k{net-params}) and Sixteen (see
\k{sixteen-params}).

Netslide was contributed to this collection by Richard Boulton.


\C{pattern} \i{Pattern}

\cfg{winhelp-topic}{games.pattern}

You have a grid of squares, which must all be filled in either black
or white. Beside each row of the grid are listed the lengths of the
runs of black squares on that row; above each column are listed the
lengths of the runs of black squares in that column. Your aim is to
fill in the entire grid black or white.

I first saw this puzzle form around 1995, under the name
\q{\i{nonograms}}. I've seen it in various places since then, under
different names.

Normally, puzzles of this type turn out to be a meaningful picture
of something once you've solved them. However, since this version
generates the puzzles automatically, they will just look like random
groupings of squares. (One user has suggested that this is actually
a \e{good} thing, since it prevents you from guessing the colour of
squares based on the picture, and forces you to use logic instead.)
The advantage, though, is that you never run out of them.

\H{pattern-controls} \I{controls, for Pattern}Pattern controls

This game is played with the mouse.

Left-click in a square to colour it black. Right-click to colour it
white. If you make a mistake, you can middle-click, or hold down
Shift while clicking with any button, to colour the square in the
default grey (meaning \q{undecided}) again.

You can click and drag with the left or right mouse button to colour
a vertical or horizontal line of squares black or white at a time
(respectively). If you click and drag with the middle button, or
with Shift held down, you can colour a whole rectangle of squares
grey.

You can also move around the grid with the cursor keys. Pressing the
return key will cycle the current cell through empty, then black, then
white, then empty, and the space bar does the same cycle in reverse.

Moving the cursor while holding Control will colour the moved-over
squares black.  Holding Shift will colour the moved-over squares
white, and holding both will colour them grey.

(All the actions described in \k{common-actions} are also available.)

\H{pattern-parameters} \I{parameters, for Pattern}Pattern parameters

The only options available from the \q{Custom...} option on the \q{Type}
menu are \e{Width} and \e{Height}, which are self-explanatory.


\C{solo} \i{Solo}

\cfg{winhelp-topic}{games.solo}

You have a square grid, which is divided into as many equally sized
sub-blocks as the grid has rows. Each square must be filled in with
a digit from 1 to the size of the grid, in such a way that

\b every row contains only one occurrence of each digit

\b every column contains only one occurrence of each digit

\b every block contains only one occurrence of each digit.

\b (optionally, by default off) each of the square's two main
diagonals contains only one occurrence of each digit.

You are given some of the numbers as clues; your aim is to place the
rest of the numbers correctly.

Under the default settings, the sub-blocks are square or
rectangular. The default puzzle size is 3\by\.3 (a 9\by\.9 actual
grid, divided into nine 3\by\.3 blocks). You can also select sizes
with rectangular blocks instead of square ones, such as 2\by\.3 (a
6\by\.6 grid divided into six 3\by\.2 blocks). Alternatively, you
can select \q{jigsaw} mode, in which the sub-blocks are arbitrary
shapes which differ between individual puzzles.

Another available mode is \q{killer}. In this mode, clues are not
given in the form of filled-in squares; instead, the grid is divided
into \q{cages} by coloured lines, and for each cage the game tells
you what the sum of all the digits in that cage should be. Also, no
digit may appear more than once within a cage, even if the cage
crosses the boundaries of existing regions.

If you select a puzzle size which requires more than 9 digits, the
additional digits will be letters of the alphabet. For example, if
you select 3\by\.4 then the digits which go in your grid will be 1
to 9, plus \cq{a}, \cq{b} and \cq{c}. This cannot be selected for
killer puzzles.

I first saw this puzzle in \i{Nikoli} \k{nikoli-solo}, although it's
also been popularised by various newspapers under the name
\q{Sudoku} or \q{Su Doku}.  Howard Garns is considered the inventor
of the modern form of the puzzle, and it was first published in
\e{Dell Pencil Puzzles and Word Games}.  A more elaborate treatment
of the history of the puzzle can be found on Wikipedia
\k{wikipedia-solo}.

\B{nikoli-solo} \W{http://www.nikoli.co.jp/en/puzzles/sudoku.html}\cw{http://www.nikoli.co.jp/en/puzzles/sudoku.html}
(beware of Flash)

\B{wikipedia-solo} \W{http://en.wikipedia.org/wiki/Sudoku}\cw{http://en.wikipedia.org/wiki/Sudoku}

\H{solo-controls} \I{controls, for Solo}Solo controls

To play Solo, simply click the mouse in any empty square and then
type a digit or letter on the keyboard to fill that square. If you
make a mistake, click the mouse in the incorrect square and press
Space to clear it again (or use the Undo feature).

If you \e{right}-click in a square and then type a number, that
number will be entered in the square as a \q{pencil mark}. You can
have pencil marks for multiple numbers in the same square. Squares
containing filled-in numbers cannot also contain pencil marks.

The game pays no attention to pencil marks, so exactly what you use
them for is up to you: you can use them as reminders that a
particular square needs to be re-examined once you know more about a
particular number, or you can use them as lists of the possible
numbers in a given square, or anything else you feel like.

To erase a single pencil mark, right-click in the square and type
the same number again.

All pencil marks in a square are erased when you left-click and type
a number, or when you left-click and press space. Right-clicking and
pressing space will also erase pencil marks.

Alternatively, use the cursor keys to move the mark around the grid.
Pressing the return key toggles the mark (from a normal mark to a
pencil mark), and typing a number in is entered in the square in the
appropriate way; typing in a 0 or using the space bar will clear a
filled square. 

(All the actions described in \k{common-actions} are also available.)

\H{solo-parameters} \I{parameters, for Solo}Solo parameters

Solo allows you to configure two separate dimensions of the puzzle
grid on the \q{Type} menu: the number of columns, and the number of
rows, into which the main grid is divided. (The size of a block is
the inverse of this: for example, if you select 2 columns and 3 rows,
each actual block will have 3 columns and 2 rows.)

If you tick the \q{X} checkbox, Solo will apply the optional extra
constraint that the two main diagonals of the grid also contain one
of every digit. (This is sometimes known as \q{Sudoku-X} in
newspapers.) In this mode, the squares on the two main diagonals
will be shaded slightly so that you know it's enabled.

If you tick the \q{Jigsaw} checkbox, Solo will generate randomly
shaped sub-blocks. In this mode, the actual grid size will be taken
to be the product of the numbers entered in the \q{Columns} and
\q{Rows} boxes. There is no reason why you have to enter a number
greater than 1 in both boxes; Jigsaw mode has no constraint on the
grid size, and it can even be a prime number if you feel like it.

If you tick the \q{Killer} checkbox, Solo will generate a set of
of cages, which are randomly shaped and drawn in an outline of a
different colour.  Each of these regions contains a smaller clue
which shows the digit sum of all the squares in this region.

You can also configure the type of symmetry shown in the generated
puzzles. More symmetry makes the puzzles look prettier but may also
make them easier, since the symmetry constraints can force more
clues than necessary to be present. Completely asymmetric puzzles
have the freedom to contain as few clues as possible.

Finally, you can configure the difficulty of the generated puzzles.
Difficulty levels are judged by the complexity of the techniques of
deduction required to solve the puzzle: each level requires a mode
of reasoning which was not necessary in the previous one. In
particular, on difficulty levels \q{Trivial} and \q{Basic} there
will be a square you can fill in with a single number at all times,
whereas at \q{Intermediate} level and beyond you will have to make
partial deductions about the \e{set} of squares a number could be in
(or the set of numbers that could be in a square).
\#{Advanced, Extreme?}
At \q{Unreasonable} level, even this is not enough, and you will
eventually have to make a guess, and then backtrack if it turns out
to be wrong.

Generating difficult puzzles is itself difficult: if you select one
of the higher difficulty levels, Solo may have to make many attempts
at generating a puzzle before it finds one hard enough for you. Be
prepared to wait, especially if you have also configured a large
puzzle size.


\C{mines} \i{Mines}

\cfg{winhelp-topic}{games.mines}

You have a grid of covered squares, some of which contain mines, but
you don't know which. Your job is to uncover every square which does
\e{not} contain a mine. If you uncover a square containing a mine,
you lose. If you uncover a square which does not contain a mine, you
are told how many mines are contained within the eight surrounding
squares.

This game needs no introduction; popularised by Windows, it is
perhaps the single best known desktop puzzle game in existence.

This version of it has an unusual property. By default, it will
generate its mine positions in such a way as to ensure that you
never need to \e{guess} where a mine is: you will always be able to
deduce it somehow. So you will never, as can happen in other
versions, get to the last four squares and discover that there are
two mines left but you have no way of knowing for sure where they
are.

\H{mines-controls} \I{controls, for Mines}Mines controls

This game is played with the mouse.

If you left-click in a covered square, it will be uncovered.

If you right-click in a covered square, it will place a flag which
indicates that the square is believed to be a mine. Left-clicking in
a marked square will not uncover it, for safety. You can right-click
again to remove a mark placed in error.

If you left-click in an \e{uncovered} square, it will \q{clear
around} the square. This means: if the square has exactly as many
flags surrounding it as it should have mines, then all the covered
squares next to it which are \e{not} flagged will be uncovered. So
once you think you know the location of all the mines around a
square, you can use this function as a shortcut to avoid having to
click on each of the remaining squares one by one.

If you uncover a square which has \e{no} mines in the surrounding
eight squares, then it is obviously safe to uncover those squares in
turn, and so on if any of them also has no surrounding mines. This
will be done for you automatically; so sometimes when you uncover a
square, a whole new area will open up to be explored.

You can also use the cursor keys to move around the minefield.
Pressing the return key in a covered square uncovers it, and in an
uncovered square will clear around it (so it acts as the left button),
pressing the space bar in a covered square will place a flag
(similarly, it acts as the right button).

All the actions described in \k{common-actions} are also available.

Even Undo is available, although you might consider it cheating to
use it. If you step on a mine, the program will only reveal the mine
in question (unlike most other implementations, which reveal all of
them). You can then Undo your fatal move and continue playing if you
like. The program will track the number of times you died (and Undo
will not reduce that counter), so when you get to the end of the
game you know whether or not you did it without making any errors.

(If you really want to know the full layout of the grid, which other
implementations will show you after you die, you can always use the
Solve menu option.)

\H{mines-parameters} \I{parameters, for Mines}Mines parameters

The options available from the \q{Custom...} option on the \q{Type}
menu are:

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Mines}

\dd Number of mines in the grid. You can enter this as an absolute
mine count, or alternatively you can put a \cw{%} sign on the end in
which case the game will arrange for that proportion of the squares
in the grid to be mines.

\lcont{

Beware of setting the mine count too high. At very high densities,
the program may spend forever searching for a solvable grid.

}

\dt \e{Ensure solubility}

\dd When this option is enabled (as it is by default), Mines will
ensure that the entire grid can be fully deduced starting from the
initial open space. If you prefer the riskier grids generated by
other implementations, you can switch off this option.


\C{samegame} \i{Same Game}

\cfg{winhelp-topic}{games.samegame}

You have a grid of coloured squares, which you have to clear by 
highlighting contiguous regions of more than one coloured square;
the larger the region you highlight, the more points you get (and
the faster you clear the arena).

If you clear the grid you win. If you end up with nothing but 
single squares (i.e., there are no more clickable regions left) you
lose.

Removing a region causes the rest of the grid to shuffle up:
blocks that are suspended will fall down (first), and then empty
columns are filled from the right. 

Same Game was contributed to this collection by James Harvey.

\H{samegame-controls} \i{Same Game controls}

\IM{Same Game controls} controls, for Same Game
\IM{Same Game controls} keys, for Same Game
\IM{Same Game controls} shortcuts (keyboard), for Same Game

This game can be played with either the keyboard or the mouse.

If you left-click an unselected region, it becomes selected (possibly
clearing the current selection). 

If you left-click the selected region, it will be removed (and the
rest of the grid shuffled immediately).

If you right-click the selected region, it will be unselected. 

The cursor keys move a cursor around the grid. Pressing the Space or
Enter keys while the cursor is in an unselected region selects it;
pressing Space or Enter again removes it as above.

(All the actions described in \k{common-actions} are also available.)

\H{samegame-parameters} \I{parameters, for Same Game}Same Game parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{No. of colours}

\dd Number of different colours used to fill the grid; the more colours,
the fewer large regions of colour and thus the more difficult it is to
successfully clear the grid.

\dt \e{Scoring system}

\dd Controls the precise mechanism used for scoring. With the default
system, \q{(n-2)^2}, only regions of three squares or more will score
any points at all. With the alternative \q{(n-1)^2} system, regions of
two squares score a point each, and larger regions score relatively
more points.

\dt \e{Ensure solubility}

\dd If this option is ticked (the default state), generated grids
will be guaranteed to have at least one solution.

\lcont{

If you turn it off, the game generator will not try to guarantee
soluble grids; it will, however, still ensure that there are at
least 2 squares of each colour on the grid at the start (since a
grid with exactly one square of a given colour is \e{definitely}
insoluble). Grids generated with this option disabled may contain
more large areas of contiguous colour, leading to opportunities for
higher scores; they can also take less time to generate.

}


\C{flip} \i{Flip}

\cfg{winhelp-topic}{games.flip}

You have a grid of squares, some light and some dark. Your aim is to
light all the squares up at the same time. You can choose any square
and flip its state from light to dark or dark to light, but when you
do so, other squares around it change state as well.

Each square contains a small diagram showing which other squares
change when you flip it.

\H{flip-controls} \i{Flip controls}

\IM{Flip controls} controls, for Flip
\IM{Flip controls} keys, for Flip
\IM{Flip controls} shortcuts (keyboard), for Flip

This game can be played with either the keyboard or the mouse.

Left-click in a square to flip it and its associated squares, or
use the cursor keys to choose a square and the space bar or Enter
key to flip.

If you use the \q{Solve} function on this game, it will mark some of
the squares in red. If you click once in every square with a red
mark, the game should be solved. (If you click in a square
\e{without} a red mark, a red mark will appear in it to indicate
that you will need to reverse that operation to reach the solution.)

(All the actions described in \k{common-actions} are also available.)

\H{flip-parameters} \I{parameters, for flip}Flip parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Shape type}

\dd This control determines the shape of the region which is flipped
by clicking in any given square. The default setting, \q{Crosses},
causes every square to flip itself and its four immediate neighbours
(or three or two if it's at an edge or corner). The other setting,
\q{Random}, causes a random shape to be chosen for every square, so
the game is different every time.


\C{guess} \i{Guess}

\cfg{winhelp-topic}{games.guess}

You have a set of coloured pegs, and have to reproduce a
predetermined sequence of them (chosen by the computer) within a
certain number of guesses. 

Each guess gets marked with the number of correctly-coloured pegs
in the correct places (in black), and also the number of
correctly-coloured pegs in the wrong places (in white). 

This game is also known (and marketed, by Hasbro, mainly) as
a board game \q{\i{Mastermind}}, with 6 colours, 4 pegs per row,
and 10 guesses. However, this version allows custom settings of number
of colours (up to 10), number of pegs per row, and number of guesses. 

Guess was contributed to this collection by James Harvey.

\H{guess-controls} \i{Guess controls}

\IM{Guess controls} controls, for Guess
\IM{Guess controls} keys, for Guess
\IM{Guess controls} shortcuts (keyboard), for Guess

This game can be played with either the keyboard or the mouse.

With the mouse, drag a coloured peg from the tray on the left-hand
side to its required position in the current guess; pegs may also be
dragged from current and past guesses to copy them elsewhere. To
remove a peg, drag it off its current position to somewhere invalid.

Right-clicking in the current guess adds a \q{hold} marker; pegs
that have hold markers will be automatically added to the next guess
after marking.

Alternatively, with the keyboard, the up and down cursor keys can be
used to select a peg colour, the left and right keys to select a
peg position, and the space bar or Enter key to place a peg of the
selected colour in the chosen position. \q{D} or Backspace removes a
peg, and Space adds a hold marker.

Pressing \q{h} or \q{?} will fill the current guess with a suggested
guess.  Using this is not recommended for 10 or more pegs as it is
slow.

When the guess is complete, the smaller feedback pegs will be highlighted;
clicking on these (or moving the peg cursor to them with the arrow keys
and pressing the space bar or Enter key) will mark the current guess,
copy any held pegs to the next guess, and move the \q{current guess}
marker.

If you correctly position all the pegs the solution will be displayed
below; if you run out of guesses (or select \q{Solve...}) the solution
will also be revealed.

(All the actions described in \k{common-actions} are also available.)

\H{guess-parameters} \I{parameters, for Guess}Guess parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu. The default game matches the parameters for the 
board game \q{Mastermind}. 

\dt \e{Colours}

\dd Number of colours the solution is chosen from; from 2 to 10
(more is harder).

\dt \e{Pegs per guess}

\dd Number of pegs per guess (more is harder).

\dt \e{Guesses}

\dd Number of guesses you have to find the solution in (fewer is harder).

\dt \e{Allow blanks}

\dd Allows blank pegs to be given as part of a guess (makes it easier, because
you know that those will never be counted as part of the solution). This
is turned off by default. 

\lcont{

Note that this doesn't allow blank pegs in the solution; if you really wanted
that, use one extra colour.

}

\dt \e{Allow duplicates}

\dd Allows the solution (and the guesses) to contain colours more than once;
this increases the search space (making things harder), and is turned on by
default.


\C{pegs} \i{Pegs}

\cfg{winhelp-topic}{games.pegs}

A number of pegs are placed in holes on a board. You can remove a
peg by jumping an adjacent peg over it (horizontally or vertically)
to a vacant hole on the other side. Your aim is to remove all but one
of the pegs initially present.

This game, best known as \I{Solitaire, Peg}\q{Peg Solitaire}, is
possibly one of the oldest puzzle games still commonly known.

\H{pegs-controls} \i{Pegs controls}

\IM{Pegs controls} controls, for Pegs

To move a peg, drag it with the mouse from its current position to
its final position. If the final position is exactly two holes away
from the initial position, is currently unoccupied by a peg, and
there is a peg in the intervening square, the move will be permitted
and the intervening peg will be removed.

Vacant spaces which you can move a peg into are marked with holes. A
space with no peg and no hole is not available for moving at all: it
is an obstacle which you must work around.

You can also use the cursor keys to move a position indicator around
the board. Pressing the return key while over a peg, followed by a
cursor key, will jump the peg in that direction (if that is a legal
move).

(All the actions described in \k{common-actions} are also available.)

\H{pegs-parameters} \I{parameters, for Pegs}Pegs parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in holes.

\dt \e{Board type}

\dd Controls whether you are given a board of a standard shape or a
randomly generated shape. The two standard shapes currently
supported are \q{Cross} and \q{Octagon} (also commonly known as the
English and European traditional board layouts respectively).
Selecting \q{Random} will give you a different board shape every
time (but always one that is known to have a solution).


\C{dominosa} \i{Dominosa}

\cfg{winhelp-topic}{games.dominosa}

A normal set of dominoes \dash that is, one instance of every
(unordered) pair of numbers from 0 to 6 \dash has been arranged
irregularly into a rectangle; then the number in each square has
been written down and the dominoes themselves removed. Your task is
to reconstruct the pattern by arranging the set of dominoes to match
the provided array of numbers.

This puzzle is widely credited to O. S. Adler, and takes part of its
name from those initials.

\H{dominosa-controls} \i{Dominosa controls}

\IM{Dominosa controls} controls, for Dominosa

Left-clicking between any two adjacent numbers places a domino
covering them, or removes one if it is already present. Trying to
place a domino which overlaps existing dominoes will remove the ones
it overlaps.

Right-clicking between two adjacent numbers draws a line between
them, which you can use to remind yourself that you know those two
numbers are \e{not} covered by a single domino. Right-clicking again
removes the line.

You can also use the cursor keys to move a cursor around the grid.
When the cursor is half way between two adjacent numbers, pressing
the return key will place a domino covering those numbers, or
pressing the space bar will lay a line between the two squares.
Repeating either action removes the domino or line.

Pressing a number key will highlight all occurrences of that
number. Pressing that number again will clear the highlighting. Up to two
different numbers can be highlighted at any given time.

(All the actions described in \k{common-actions} are also available.)

\H{dominosa-parameters} \I{parameters, for Dominosa}Dominosa parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Maximum number on dominoes}

\dd Controls the size of the puzzle, by controlling the size of the
set of dominoes used to make it. Dominoes with numbers going up to N
will give rise to an (N+2) \by (N+1) rectangle; so, in particular,
the default value of 6 gives an 8\by\.7 grid.

\dt \e{Ensure unique solution}

\dd Normally, Dominosa will make sure that the puzzles it presents
have only one solution. Puzzles with ambiguous sections can be more
difficult and sometimes more subtle, so if you like you can turn off
this feature. Also, finding \e{all} the possible solutions can be an
additional challenge for an advanced player. Turning off this option
can also speed up puzzle generation.


\C{untangle} \i{Untangle}

\cfg{winhelp-topic}{games.untangle}

You are given a number of points, some of which have lines drawn
between them. You can move the points about arbitrarily; your aim is
to position the points so that no line crosses another.

I originally saw this in the form of a Flash game called \i{Planarity}
\k{Planarity}, written by John Tantalo.

\B{Planarity} \W{http://planarity.net}\cw{http://planarity.net}

\H{untangle-controls} \i{Untangle controls}

\IM{Untangle controls} controls, for Untangle

To move a point, click on it with the left mouse button and drag it
into a new position.

(All the actions described in \k{common-actions} are also available.)

\H{untangle-parameters} \I{parameters, for Untangle}Untangle parameters

There is only one parameter available from the \q{Custom...} option
on the \q{Type} menu:

\dt \e{Number of points}

\dd Controls the size of the puzzle, by specifying the number of
points in the generated graph.


\C{blackbox} \i{Black Box}

\cfg{winhelp-topic}{games.blackbox}

A number of balls are hidden in a rectangular arena. You have to
deduce the positions of the balls by firing lasers positioned at
the edges of the arena and observing how their beams are deflected. 

Beams will travel straight from their origin until they hit the
opposite side of the arena (at which point they emerge), unless
affected by balls in one of the following ways:

\b A beam that hits a ball head-on is absorbed and will never
   re-emerge. This includes beams that meet a ball on the first rank
   of the arena.

\b A beam with a ball in its front-left square and no ball ahead of it
   gets deflected 90 degrees to the right.

\b A beam with a ball in its front-right square and no ball ahead of
   it gets similarly deflected to the left.

\b A beam that would re-emerge from its entry location is considered to be
   \q{reflected}. 

\b A beam which would get deflected before entering the arena by a
   ball to the front-left or front-right of its entry point is also
   considered to be \q{reflected}.

Beams that are reflected appear as a \q{R}; beams that hit balls
head-on appear as \q{H}. Otherwise, a number appears at the firing
point and the location where the beam emerges (this number is unique
to that shot).

You can place guesses as to the location of the balls, based on the
entry and exit patterns of the beams; once you have placed enough
balls a button appears enabling you to have your guesses checked. 

Here is a diagram showing how the positions of balls can create each
of the beam behaviours shown above:

\c  1RHR---- 
\c |..O.O...|
\c 2........3
\c |........|
\c |........|
\c 3........|
\c |......O.|
\c H........|
\c |.....O..|
\c  12-RR---

As shown, it is possible for a beam to receive multiple reflections
before re-emerging (see turn 3). Similarly, a beam may be reflected
(possibly more than once) before receiving a hit (the \q{H} on the
left side of the example).

Note that any layout with more than 4 balls may have a non-unique
solution.  The following diagram illustrates this; if you know the
board contains 5 balls, it is impossible to determine where the fifth
ball is (possible positions marked with an \cw{x}):

\c  -------- 
\c |........|
\c |........|
\c |..O..O..|
\c |...xx...|
\c |...xx...|
\c |..O..O..|
\c |........|
\c |........|
\c  --------

For this reason, when you have your guesses checked, the game will
check that your solution \e{produces the same results} as the
computer's, rather than that your solution is identical to the
computer's. So in the above example, you could put the fifth ball at
\e{any} of the locations marked with an \cw{x}, and you would still
win.

Black Box was contributed to this collection by James Harvey.

\H{blackbox-controls} \i{Black Box controls}

\IM{Black Box controls} controls, for Black Box
\IM{Black Box controls} keys, for Black Box
\IM{Black Box controls} shortcuts (keyboard), for Black Box

To fire a laser beam, left-click in a square around the edge of the
arena. The results will be displayed immediately. Clicking or holding
the left button on one of these squares will highlight the current go
(or a previous go) to confirm the exit point for that laser, if
applicable.

To guess the location of a ball, left-click within the arena and a
black circle will appear marking the guess; click again to remove the
guessed ball.

Locations in the arena may be locked against modification by
right-clicking; whole rows and columns may be similarly locked by
right-clicking in the laser square above/below that column, or to the
left/right of that row.

The cursor keys may also be used to move around the grid. Pressing the
Enter key will fire a laser or add a new ball-location guess, and
pressing Space will lock a cell, row, or column.

When an appropriate number of balls have been guessed, a button will
appear at the top-left corner of the grid; clicking that (with mouse
or cursor) will check your guesses. 

If you click the \q{check} button and your guesses are not correct,
the game will show you the minimum information necessary to
demonstrate this to you, so you can try again. If your ball
positions are not consistent with the beam paths you already know
about, one beam path will be circled to indicate that it proves you
wrong. If your positions match all the existing beam paths but are
still wrong, one new beam path will be revealed (written in red)
which is not consistent with your current guesses.

If you decide to give up completely, you can select Solve to reveal
the actual ball positions. At this point, correctly-placed balls
will be displayed as filled black circles, incorrectly-placed balls
as filled black circles with red crosses, and missing balls as filled
red circles. In addition, a red circle marks any laser you had already
fired which is not consistent with your ball layout (just as when you
press the \q{check} button), and red text marks any laser you
\e{could} have fired in order to distinguish your ball layout from the
correct one.

(All the actions described in \k{common-actions} are also available.)

\H{blackbox-parameters} \I{parameters, for Black Box}Black Box parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares. There are 2 \by \e{Width} \by \e{Height} lasers 
per grid, two per row and two per column. 

\dt \e{No. of balls}

\dd Number of balls to place in the grid. This can be a single number,
or a range (separated with a hyphen, like \q{2-6}), and determines the
number of balls to place on the grid. The \q{reveal} button is only
enabled if you have guessed an appropriate number of balls; a guess
using a different number to the original solution is still acceptable,
if all the beam inputs and outputs match.


\C{slant} \i{Slant}

\cfg{winhelp-topic}{games.slant}

You have a grid of squares. Your aim is to draw a diagonal line
through each square, and choose which way each line slants so that
the following conditions are met:

\b The diagonal lines never form a loop.

\b Any point with a circled number has precisely that many lines
meeting at it. (Thus, a 4 is the centre of a cross shape, whereas a
zero is the centre of a diamond shape \dash or rather, a partial
diamond shape, because a zero can never appear in the middle of the
grid because that would immediately cause a loop.)

Credit for this puzzle goes to \i{Nikoli} \k{nikoli-slant}.

\B{nikoli-slant}
\W{http://www.nikoli.co.jp/ja/puzzles/gokigen_naname}\cw{http://www.nikoli.co.jp/ja/puzzles/gokigen_naname}
(in Japanese)

\H{slant-controls} \i{Slant controls}

\IM{Slant controls} controls, for Slant

Left-clicking in a blank square will place a \cw{\\} in it (a line
leaning to the left, i.e. running from the top left of the square to
the bottom right). Right-clicking in a blank square will place a
\cw{/} in it (leaning to the right, running from top right to bottom
left).

Continuing to click either button will cycle between the three
possible square contents. Thus, if you left-click repeatedly in a
blank square it will change from blank to \cw{\\} to \cw{/} back to
blank, and if you right-click repeatedly the square will change from
blank to \cw{/} to \cw{\\} back to blank. (Therefore, you can play
the game entirely with one button if you need to.)

You can also use the cursor keys to move around the grid. Pressing the
return or space keys will place a \cw{\\} or a \cw{/}, respectively,
and will then cycle them as above.  You can also press \cw{/} or
\cw{\\} to place a \cw{/} or \cw{\\}, respectively, independent of
what is already in the cursor square.  Backspace removes any line from
the cursor square.

(All the actions described in \k{common-actions} are also available.)

\H{slant-parameters} \I{parameters, for Slant}Slant parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. At Hard level,
you are required to do deductions based on knowledge of
\e{relationships} between squares rather than always being able to
deduce the exact contents of one square at a time. (For example, you
might know that two squares slant in the same direction, even if you
don't yet know what that direction is, and this might enable you to
deduce something about still other squares.) Even at Hard level,
guesswork and backtracking should never be necessary.


\C{lightup} \i{Light Up}

\cfg{winhelp-topic}{games.lightup}

You have a grid of squares. Some are filled in black; some of the
black squares are numbered. Your aim is to \q{light up} all the
empty squares by placing light bulbs in some of them.

Each light bulb illuminates the square it is on, plus all squares in
line with it horizontally or vertically unless a black square is
blocking the way.

To win the game, you must satisfy the following conditions:

\b All non-black squares are lit.

\b No light is lit by another light.

\b All numbered black squares have exactly that number of lights adjacent to
   them (in the four squares above, below, and to the side).

Non-numbered black squares may have any number of lights adjacent to them. 

Credit for this puzzle goes to \i{Nikoli} \k{nikoli-lightup}.

Light Up was contributed to this collection by James Harvey.

\B{nikoli-lightup}
\W{http://www.nikoli.co.jp/en/puzzles/akari.html}\cw{http://www.nikoli.co.jp/en/puzzles/akari.html}
(beware of Flash)

\H{lightup-controls} \i{Light Up controls}

\IM{Light Up controls} controls, for Light Up

Left-clicking in a non-black square will toggle the presence of a light
in that square. Right-clicking in a non-black square toggles a mark there to aid
solving; it can be used to highlight squares that cannot be lit, for example. 

You may not place a light in a marked square, nor place a mark in a lit square.

The game will highlight obvious errors in red. Lights lit by other
lights are highlighted in this way, as are numbered squares which
do not (or cannot) have the right number of lights next to them.
  
Thus, the grid is solved when all non-black squares have yellow
highlights and there are no red lights.

(All the actions described in \k{common-actions} are also available.)

\H{lightup-parameters} \I{parameters, for Light Up}Light Up parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{%age of black squares}

\dd Rough percentage of black squares in the grid.

\lcont{

This is a hint rather than an instruction. If the grid generator is
unable to generate a puzzle to this precise specification, it will
increase the proportion of black squares until it can.

}

\dt \e{Symmetry}

\dd Allows you to specify the required symmetry of the black squares
in the grid. (This does not affect the difficulty of the puzzles
noticeably.)

\dt \e{Difficulty}

\dd \q{Easy} means that the puzzles should be soluble without
backtracking or guessing, \q{Hard} means that some guesses will
probably be necessary.


\C{map} \i{Map}

\cfg{winhelp-topic}{games.map}

You are given a map consisting of a number of regions. Your task is
to colour each region with one of four colours, in such a way that
no two regions sharing a boundary have the same colour. You are
provided with some regions already coloured, sufficient to make the
remainder of the solution unique.

Only regions which share a length of border are required to be
different colours. Two regions which meet at only one \e{point}
(i.e. are diagonally separated) may be the same colour.

I believe this puzzle is original; I've never seen an implementation
of it anywhere else. The concept of a \i{four-colouring} puzzle was
suggested by Owen Dunn; credit must also go to Nikoli and to Verity
Allan for inspiring the train of thought that led to me realising
Owen's suggestion was a viable puzzle. Thanks also to Gareth Taylor
for many detailed suggestions.

\H{map-controls} \i{Map controls}

\IM{Map controls} controls, for Map

To colour a region, click the left mouse button on an existing
region of the desired colour and drag that colour into the new
region.

(The program will always ensure the starting puzzle has at least one
region of each colour, so that this is always possible!)

If you need to clear a region, you can drag from an empty region, or
from the puzzle boundary if there are no empty regions left.

Dragging a colour using the \e{right} mouse button will stipple the
region in that colour, which you can use as a note to yourself that
you think the region \e{might} be that colour. A region can contain
stipples in multiple colours at once. (This is often useful at the
harder difficulty levels.)

You can also use the cursor keys to move around the map: the colour of
the cursor indicates the position of the colour you would drag (which
is not obvious if you're on a region's boundary, since it depends on the
direction from which you approached the boundary). Pressing the return
key starts a drag of that colour, as above, which you control with the
cursor keys; pressing the return key again finishes the drag. The
space bar can be used similarly to create a stippled region. 
Double-pressing the return key (without moving the cursor) will clear
the region, as a drag from an empty region does: this is useful with
the cursor mode if you have filled the entire map in but need to 
correct the layout.

If you press L during play, the game will toggle display of a number
in each region of the map. This is useful if you want to discuss a
particular puzzle instance with a friend \dash having an unambiguous
name for each region is much easier than trying to refer to them all
by names such as \q{the one down and right of the brown one on the
top border}.

(All the actions described in \k{common-actions} are also available.)

\H{map-parameters} \I{parameters, for Map}Map parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Regions}

\dd Number of regions in the generated map.

\dt \e{Difficulty}

\dd In \q{Easy} mode, there should always be at least one region
whose colour can be determined trivially. In \q{Normal} and \q{Hard}
modes, you will have to use increasingly complex logic to deduce the
colour of some regions. However, it will always be possible without
having to guess or backtrack.

\lcont{

In \q{Unreasonable} mode, the program will feel free to generate
puzzles which are as hard as it can possibly make them: the only
constraint is that they should still have a unique solution. Solving
Unreasonable puzzles may require guessing and backtracking.

}


\C{loopy} \i{Loopy}

\cfg{winhelp-topic}{games.loopy}

You are given a grid of dots, marked with yellow lines to indicate
which dots you are allowed to connect directly together. Your aim is
to use some subset of those yellow lines to draw a single unbroken
loop from dot to dot within the grid.

Some of the spaces between the lines contain numbers. These numbers
indicate how many of the lines around that space form part of the
loop. The loop you draw must correctly satisfy all of these clues to
be considered a correct solution.

In the default mode, the dots are arranged in a grid of squares;
however, you can also play on triangular or hexagonal grids, or even
more exotic ones.

Credit for the basic puzzle idea goes to \i{Nikoli}
\k{nikoli-loopy}.

Loopy was originally contributed to this collection by Mike Pinna,
and subsequently enhanced to handle various types of non-square grid
by Lambros Lambrou.

\B{nikoli-loopy}
\W{http://www.nikoli.co.jp/en/puzzles/slitherlink.html}\cw{http://www.nikoli.co.jp/en/puzzles/slitherlink.html}
(beware of Flash)

\H{loopy-controls} \i{Loopy controls}

\IM{Loopy controls} controls, for Loopy

Click the left mouse button on a yellow line to turn it black,
indicating that you think it is part of the loop. Click again to
turn the line yellow again (meaning you aren't sure yet).

If you are sure that a particular line segment is \e{not} part of
the loop, you can click the right mouse button to remove it
completely. Again, clicking a second time will turn the line back to
yellow.

(All the actions described in \k{common-actions} are also available.)

\H{loopy-parameters} \I{parameters, for Loopy}Loopy parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid, measured in number of regions across and down. For
square grids, it's clear how this is counted; for other types of
grid you may have to think a bit to see how the dimensions are
measured.

\dt \e{Grid type}

\dd Allows you to choose between a selection of types of tiling.
Some have all the faces the same but may have multiple different
types of vertex (e.g. the \e{Cairo} or \e{Kites} mode); others have
all the vertices the same but may have different types of face (e.g.
the \e{Great Hexagonal}). The square, triangular and honeycomb grids
are fully regular, and have all their vertices \e{and} faces the
same; this makes them the least confusing to play.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle.
\#{FIXME: what distinguishes Easy, Medium, and Hard? In particular,
when are backtracking/guesswork required, if ever?}


\C{inertia} \i{Inertia}

\cfg{winhelp-topic}{games.inertia}

You are a small green ball sitting in a grid full of obstacles. Your
aim is to collect all the gems without running into any mines.

You can move the ball in any orthogonal \e{or diagonal} direction.
Once the ball starts moving, it will continue until something stops
it. A wall directly in its path will stop it (but if it is moving
diagonally, it will move through a diagonal gap between two other
walls without stopping). Also, some of the squares are \q{stops};
when the ball moves on to a stop, it will stop moving no matter what
direction it was going in. Gems do \e{not} stop the ball; it picks
them up and keeps on going.

Running into a mine is fatal. Even if you picked up the last gem in
the same move which then hit a mine, the game will count you as dead
rather than victorious.

This game was originally implemented for Windows by Ben Olmstead
\k{bem}, who was kind enough to release his source code on request
so that it could be re-implemented for this collection.

\B{bem} \W{http://xn13.com/}\cw{http://xn13.com/}

\H{inertia-controls} \i{Inertia controls}

\IM{Inertia controls} controls, for Inertia
\IM{Inertia controls} keys, for Inertia
\IM{Inertia controls} shortcuts (keyboard), for Inertia

You can move the ball in any of the eight directions using the
numeric keypad. Alternatively, if you click the left mouse button on
the grid, the ball will begin a move in the general direction of
where you clicked.

If you use the \q{Solve} function on this game, the program will
compute a path through the grid which collects all the remaining
gems and returns to the current position. A hint arrow will appear
on the ball indicating the direction in which you should move to
begin on this path. If you then move in that direction, the arrow
will update to indicate the next direction on the path. You can also
press Space to automatically move in the direction of the hint
arrow. If you move in a different direction from the one shown by
the arrow, arrows will be shown only if the puzzle is still solvable.

All the actions described in \k{common-actions} are also available.
In particular, if you do run into a mine and die, you can use the
Undo function and resume playing from before the fatal move. The
game will keep track of the number of times you have done this.

\H{inertia-parameters} \I{parameters, for Inertia}Inertia parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.


\C{tents} \i{Tents}

\cfg{winhelp-topic}{games.tents}

You have a grid of squares, some of which contain trees. Your aim is
to place tents in some of the remaining squares, in such a way that
the following conditions are met:

\b There are exactly as many tents as trees.

\b The tents and trees can be matched up in such a way that each
tent is directly adjacent (horizontally or vertically, but not
diagonally) to its own tree. However, a tent may be adjacent to
other trees as well as its own.

\b No two tents are adjacent horizontally, vertically \e{or
diagonally}.

\b The number of tents in each row, and in each column, matches the
numbers given round the sides of the grid.

This puzzle can be found in several places on the Internet, and was
brought to my attention by e-mail. I don't know who I should credit
for inventing it.

\H{tents-controls} \i{Tents controls}

\IM{Tents controls} controls, for Tents

Left-clicking in a blank square will place a tent in it.
Right-clicking in a blank square will colour it green, indicating
that you are sure it \e{isn't} a tent. Clicking either button in an
occupied square will clear it.

If you \e{drag} with the right button along a row or column, every
blank square in the region you cover will be turned green, and no
other squares will be affected. (This is useful for clearing the
remainder of a row once you have placed all its tents.)

You can also use the cursor keys to move around the grid. Pressing the
return key over an empty square will place a tent, and pressing the
space bar over an empty square will colour it green; either key will
clear an occupied square.  Holding Shift and pressing the cursor keys
will colour empty squares green.  Holding Control and pressing the
cursor keys will colour green both empty squares and squares with tents.

(All the actions described in \k{common-actions} are also available.)

\H{tents-parameters} \I{parameters, for Tents}Tents parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. More difficult
puzzles require more complex deductions, but at present none of the
available difficulty levels requires guesswork or backtracking.


\C{bridges} \i{Bridges}

\cfg{winhelp-topic}{games.bridges}

You have a set of islands distributed across the playing area. Each
island contains a number. Your aim is to connect the islands
together with bridges, in such a way that:

\b Bridges run horizontally or vertically.

\b The number of bridges terminating at any island is equal to the
number written in that island.

\b Two bridges may run in parallel between the same two islands, but
no more than two may do so.

\b No bridge crosses another bridge.

\b All the islands are connected together.

There are some configurable alternative modes, which involve
changing the parallel-bridge limit to something other than 2, and
introducing the additional constraint that no sequence of bridges
may form a loop from one island back to the same island. The rules
stated above are the default ones.

Credit for this puzzle goes to \i{Nikoli} \k{nikoli-bridges}.

Bridges was contributed to this collection by James Harvey.

\B{nikoli-bridges}
\W{http://www.nikoli.co.jp/en/puzzles/hashiwokakero.html}\cw{http://www.nikoli.co.jp/en/puzzles/hashiwokakero.html}
(beware of Flash)

\H{bridges-controls} \i{Bridges controls}

\IM{Bridges controls} controls, for Bridges

To place a bridge between two islands, click the mouse down on one
island and drag it towards the other. You do not need to drag all
the way to the other island; you only need to move the mouse far
enough for the intended bridge direction to be unambiguous. (So you
can keep the mouse near the starting island and conveniently throw
bridges out from it in many directions.)

Doing this again when a bridge is already present will add another
parallel bridge. If there are already as many bridges between the
two islands as permitted by the current game rules (i.e. two by
default), the same dragging action will remove all of them.

If you want to remind yourself that two islands definitely \e{do
not} have a bridge between them, you can right-drag between them in
the same way to draw a \q{non-bridge} marker.

If you think you have finished with an island (i.e. you have placed
all its bridges and are confident that they are in the right
places), you can mark the island as finished by left-clicking on it.
This will highlight it and all the bridges connected to it, and you
will be prevented from accidentally modifying any of those bridges
in future. Left-clicking again on a highlighted island will unmark
it and restore your ability to modify it.

You can also use the cursor keys to move around the grid: if possible
the cursor will always move orthogonally, otherwise it will move
towards the nearest island to the indicated direction. Holding Control
and pressing a cursor key will lay a bridge in that direction (if
available); Shift and a cursor key will lay a \q{non-bridge} marker.
Pressing the return key followed by a cursor key will also lay a
bridge in that direction.

You can mark an island as finished by pressing the space bar or by
pressing the return key twice.

By pressing a number key, you can jump to the nearest island with that
number.  Letters \q{a}, ..., \q{f} count as 10, ..., 15 and \q{0} as
16.

Violations of the puzzle rules will be marked in red:

\b An island with too many bridges will be highlighted in red.

\b An island with too few bridges will be highlighted in red if it
is definitely an error (as opposed to merely not being finished
yet): if adding enough bridges would involve having to cross another
bridge or remove a non-bridge marker, or if the island has been
highlighted as complete.

\b A group of islands and bridges may be highlighted in red if it is
a closed subset of the puzzle with no way to connect it to the rest
of the islands. For example, if you directly connect two 1s together
with a bridge and they are not the only two islands on the grid,
they will light up red to indicate that such a group cannot be
contained in any valid solution.

\b If you have selected the (non-default) option to disallow loops
in the solution, a group of bridges which forms a loop will be
highlighted.

(All the actions described in \k{common-actions} are also available.)

\H{bridges-parameters} \I{parameters, for Bridges}Bridges parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Difficulty}

\dd Difficulty level of puzzle.

\dt \e{Allow loops}

\dd This is set by default. If cleared, puzzles will be generated in
such a way that they are always soluble without creating a loop, and
solutions which do involve a loop will be disallowed.

\dt \e{Max. bridges per direction}

\dd Maximum number of bridges in any particular direction. The
default is 2, but you can change it to 1, 3 or 4. In general, fewer
is easier.

\dt \e{%age of island squares}

\dd Gives a rough percentage of islands the generator will try and
lay before finishing the puzzle. Certain layouts will not manage to
lay enough islands; this is an upper bound.

\dt \e{Expansion factor (%age)}

\dd The grid generator works by picking an existing island at random
(after first creating an initial island somewhere). It then decides
on a direction (at random), and then works out how far it could
extend before creating another island. This parameter determines how
likely it is to extend as far as it can, rather than choosing
somewhere closer.

\lcont{

High expansion factors usually mean easier puzzles with fewer
possible islands; low expansion factors can create lots of
tightly-packed islands.

}


\C{unequal} \i{Unequal}

\cfg{winhelp-topic}{games.unequal}

You have a square grid; each square may contain a digit from 1 to
the size of the grid, and some squares have clue signs between
them. Your aim is to fully populate the grid with numbers such that:

\b Each row contains only one occurrence of each digit

\b Each column contains only one occurrence of each digit

\b All the clue signs are satisfied. 

There are two modes for this game, \q{Unequal} and \q{Adjacent}.

In \q{Unequal} mode, the clue signs are greater-than symbols indicating one
square's value is greater than its neighbour's. In this mode not all clues
may be visible, particularly at higher difficulty levels. 

In \q{Adjacent} mode, the clue signs are bars indicating
one square's value is numerically adjacent (i.e. one higher or one lower)
than its neighbour. In this mode all clues are always visible: absence of
a bar thus means that a square's value is definitely not numerically adjacent
to that neighbour's.  

In \q{Trivial} difficulty level (available via the \q{Custom} game type
selector), there are no greater-than signs in \q{Unequal} mode; the puzzle is
to solve the \i{Latin square} only.

At the time of writing, the \q{Unequal} mode of this puzzle is appearing in the
Guardian weekly under the name \q{\i{Futoshiki}}.

Unequal was contributed to this collection by James Harvey.

\H{unequal-controls} \i{Unequal controls}

\IM{Unequal controls} controls, for Unequal

Unequal shares much of its control system with Solo.

To play Unequal, simply click the mouse in any empty square and then
type a digit or letter on the keyboard to fill that square. If you
make a mistake, click the mouse in the incorrect square and press
Space to clear it again (or use the Undo feature).

If you \e{right}-click in a square and then type a number, that
number will be entered in the square as a \q{pencil mark}. You can
have pencil marks for multiple numbers in the same square. Squares
containing filled-in numbers cannot also contain pencil marks.

The game pays no attention to pencil marks, so exactly what you use
them for is up to you: you can use them as reminders that a
particular square needs to be re-examined once you know more about a
particular number, or you can use them as lists of the possible
numbers in a given square, or anything else you feel like.

To erase a single pencil mark, right-click in the square and type
the same number again.

All pencil marks in a square are erased when you left-click and type
a number, or when you left-click and press space. Right-clicking and
pressing space will also erase pencil marks.

As for Solo, the cursor keys can be used in conjunction with the digit
keys to set numbers or pencil marks. You can also use the \q{M} key to
auto-fill every numeric hint, ready for removal as required, or the \q{H}
key to do the same but also to remove all obvious hints. 

Alternatively, use the cursor keys to move the mark around the grid.
Pressing the return key toggles the mark (from a normal mark to a
pencil mark), and typing a number in is entered in the square in the
appropriate way; typing in a 0 or using the space bar will clear a
filled square. 

Left-clicking a clue will mark it as done (grey it out), or unmark it
if it is already marked.  Holding Control or Shift and pressing an
arrow key likewise marks any clue adjacent to the cursor in the given
direction.

(All the actions described in \k{common-actions} are also available.)

\H{unequal-parameters} \I{parameters, for Unequal}Unequal parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Mode}

\dd Mode of the puzzle (\q{Unequal} or \q{Adjacent})

\dt \e{Size (s*s)}

\dd Size of grid.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. At Trivial
level, there are no greater-than signs; the puzzle is to solve the
Latin square only. At Recursive level (only available via the
\q{Custom} game type selector) backtracking will be required, but
the solution should still be unique. The levels in between require
increasingly complex reasoning to avoid having to backtrack.



\C{galaxies} \i{Galaxies}

\cfg{winhelp-topic}{games.galaxies}

You have a rectangular grid containing a number of dots. Your aim is
to draw edges along the grid lines which divide the rectangle into
regions in such a way that every region is 180\u00b0{-degree}
rotationally symmetric, and contains exactly one dot which is
located at its centre of symmetry.

This puzzle was invented by \i{Nikoli} \k{nikoli-galaxies}, under
the name \q{Tentai Show}; its name is commonly translated into
English as \q{Spiral Galaxies}.

Galaxies was contributed to this collection by James Harvey.

\B{nikoli-galaxies} \W{http://www.nikoli.co.jp/en/puzzles/astronomical_show.html}\cw{http://www.nikoli.co.jp/en/puzzles/astronomical_show.html}

\H{galaxies-controls} \i{Galaxies controls}

\IM{Galaxies controls} controls, for Galaxies

Left-click on any grid line to draw an edge if there isn't one
already, or to remove one if there is. When you create a valid
region (one which is closed, contains exactly one dot, is
180\u00b0{-degree} symmetric about that dot, and contains no
extraneous edges inside it) it will be highlighted automatically; so
your aim is to have the whole grid highlighted in that way.

During solving, you might know that a particular grid square belongs
to a specific dot, but not be sure of where the edges go and which
other squares are connected to the dot. In order to mark this so you
don't forget, you can right-click on the dot and drag, which will
create an arrow marker pointing at the dot. Drop that in a square of
your choice and it will remind you which dot it's associated with.
You can also right-click on existing arrows to pick them up and move
them, or destroy them by dropping them off the edge of the grid.
(Also, if you're not sure which dot an arrow is pointing at, you can
pick it up and move it around to make it clearer. It will swivel
constantly as you drag it, to stay pointed at its parent dot.)

You can also use the cursor keys to move around the grid squares and
lines.  Pressing the return key when over a grid line will draw or
clear its edge, as above. Pressing the return key when over a dot will
pick up an arrow, to be dropped the next time the return key is
pressed; this can also be used to move existing arrows around, removing
them by dropping them on a dot or another arrow.

(All the actions described in \k{common-actions} are also available.)

\H{galaxies-parameters} \I{parameters, for Galaxies}Galaxies parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. More difficult
puzzles require more complex deductions, and the \q{Unreasonable}
difficulty level may require backtracking.



\C{filling} \i{Filling}

\cfg{winhelp-topic}{games.filling}

You have a grid of squares, some of which contain digits, and the
rest of which are empty. Your job is to fill in digits in the empty
squares, in such a way that each connected region of squares all
containing the same digit has an area equal to that digit.

(\q{Connected region}, for the purposes of this game, does not count
diagonally separated squares as adjacent.)

For example, it follows that no square can contain a zero, and that
two adjacent squares can not both contain a one.  No region has an
area greater than 9 (because then its area would not be a single
digit).

Credit for this puzzle goes to \i{Nikoli} \k{nikoli-fillomino}.

Filling was contributed to this collection by Jonas K\u00F6{oe}lker.

\B{nikoli-fillomino}
\W{http://www.nikoli.co.jp/en/puzzles/fillomino.html}\cw{http://www.nikoli.co.jp/en/puzzles/fillomino.html}

\H{filling-controls} \I{controls, for Filling}Filling controls

To play Filling, simply click the mouse in any empty square and then
type a digit on the keyboard to fill that square. By dragging the
mouse, you can select multiple squares to fill with a single keypress.
If you make a mistake, click the mouse in the incorrect square and
press 0, Space, Backspace or Enter to clear it again (or use the Undo
feature).

You can also move around the grid with the cursor keys; typing a digit will
fill the square containing the cursor with that number; typing 0 will clear
it.  You can also select multiple squares for numbering or clearing with the
return and arrow keys, before typing a digit to fill or clear the highlighted
squares (as above).  The space bar adds and removes single squares to and from
the selection.  Backspace and escape remove all squares from the selection.

(All the actions described in \k{common-actions} are also available.)

\H{filling-parameters} \I{parameters, for Filling}Filling parameters

Filling allows you to configure the number of rows and columns of the
grid, through the \q{Type} menu.


\C{keen} \i{Keen}

\cfg{winhelp-topic}{games.keen}

You have a square grid; each square may contain a digit from 1 to
the size of the grid. The grid is divided into blocks of varying
shape and size, with arithmetic clues written in them. Your aim is
to fully populate the grid with digits such that:

\b Each row contains only one occurrence of each digit

\b Each column contains only one occurrence of each digit

\b The digits in each block can be combined to form the number
stated in the clue, using the arithmetic operation given in the
clue. That is:

\lcont{

\b An addition clue means that the sum of the digits in the block
must be the given number. For example, \q{15+} means the contents of
the block adds up to fifteen.

\b A multiplication clue (e.g. \q{60\times}), similarly, means that
the product of the digits in the block must be the given number.

\b A subtraction clue will always be written in a block of size two,
and it means that one of the digits in the block is greater than the
other by the given amount. For example, \q{2\minus} means that one
of the digits in the block is 2 more than the other, or equivalently
that one digit minus the other one is 2. The two digits could be
either way round, though.

\b A division clue (e.g. \q{3\divide}), similarly, is always in a
block of size two and means that one digit divided by the other is
equal to the given amount.

Note that a block may contain the same digit more than once
(provided the identical ones are not in the same row and column).
This rule is precisely the opposite of the rule in Solo's \q{Killer}
mode (see \k{solo}).

}

This puzzle appears in the Times under the name \q{\i{KenKen}}.


\H{keen-controls} \i{Keen controls}

\IM{Keen controls} controls, for Keen

Keen shares much of its control system with Solo (and Unequal).

To play Keen, simply click the mouse in any empty square and then
type a digit on the keyboard to fill that square. If you make a
mistake, click the mouse in the incorrect square and press Space to
clear it again (or use the Undo feature).

If you \e{right}-click in a square and then type a number, that
number will be entered in the square as a \q{pencil mark}. You can
have pencil marks for multiple numbers in the same square. Squares
containing filled-in numbers cannot also contain pencil marks.

The game pays no attention to pencil marks, so exactly what you use
them for is up to you: you can use them as reminders that a
particular square needs to be re-examined once you know more about a
particular number, or you can use them as lists of the possible
numbers in a given square, or anything else you feel like.

To erase a single pencil mark, right-click in the square and type
the same number again.

All pencil marks in a square are erased when you left-click and type
a number, or when you left-click and press space. Right-clicking and
pressing space will also erase pencil marks.

As for Solo, the cursor keys can be used in conjunction with the
digit keys to set numbers or pencil marks. Use the cursor keys to
move a highlight around the grid, and type a digit to enter it in
the highlighted square. Pressing return toggles the highlight into a
mode in which you can enter or remove pencil marks.

Pressing M will fill in a full set of pencil marks in every square
that does not have a main digit in it.

(All the actions described in \k{common-actions} are also available.)

\H{keen-parameters} \I{parameters, for Keen}Keen parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Grid size}

\dd Specifies the size of the grid. Lower limit is 3; upper limit is
9 (because the user interface would become more difficult with
\q{digits} bigger than 9!).

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. At Unreasonable
level, some backtracking will be required, but the solution should
still be unique. The remaining levels require increasingly complex
reasoning to avoid having to backtrack.

\dt \e{Multiplication only}

\dd If this is enabled, all boxes will be multiplication boxes.
With this rule, the puzzle is known as \q{Inshi No Heya}.

\C{towers} \i{Towers}

\cfg{winhelp-topic}{games.towers}

You have a square grid. On each square of the grid you can build a
tower, with its height ranging from 1 to the size of the grid.
Around the edge of the grid are some numeric clues.

Your task is to build a tower on every square, in such a way that:

\b Each row contains every possible height of tower once

\b Each column contains every possible height of tower once

\b Each numeric clue describes the number of towers that can be seen
if you look into the square from that direction, assuming that
shorter towers are hidden behind taller ones. For example, in a
5\by\.5 grid, a clue marked \q{5} indicates that the five tower
heights must appear in increasing order (otherwise you would not be
able to see all five towers), whereas a clue marked \q{1} indicates
that the tallest tower (the one marked 5) must come first.

In harder or larger puzzles, some towers will be specified for you
as well as the clues round the edge, and some edge clues may be
missing.

This puzzle appears on the web under various names, particularly
\q{\i{Skyscrapers}}, but I don't know who first invented it.


\H{towers-controls} \i{Towers controls}

\IM{Towers controls} controls, for Towers

Towers shares much of its control system with Solo, Unequal and Keen.

To play Towers, simply click the mouse in any empty square and then
type a digit on the keyboard to fill that square with a tower of the
given height. If you make a mistake, click the mouse in the
incorrect square and press Space to clear it again (or use the Undo
feature).

If you \e{right}-click in a square and then type a number, that
number will be entered in the square as a \q{pencil mark}. You can
have pencil marks for multiple numbers in the same square. A square
containing a tower cannot also contain pencil marks.

The game pays no attention to pencil marks, so exactly what you use
them for is up to you: you can use them as reminders that a
particular square needs to be re-examined once you know more about a
particular number, or you can use them as lists of the possible
numbers in a given square, or anything else you feel like.

To erase a single pencil mark, right-click in the square and type
the same number again.

All pencil marks in a square are erased when you left-click and type
a number, or when you left-click and press space. Right-clicking and
pressing space will also erase pencil marks.

As for Solo, the cursor keys can be used in conjunction with the
digit keys to set numbers or pencil marks. Use the cursor keys to
move a highlight around the grid, and type a digit to enter it in
the highlighted square. Pressing return toggles the highlight into a
mode in which you can enter or remove pencil marks.

Pressing M will fill in a full set of pencil marks in every square
that does not have a main digit in it.

Left-clicking a clue will mark it as done (grey it out), or unmark it
if it is already marked.  Holding Control or Shift and pressing an
arrow key likewise marks any clue in the given direction.

(All the actions described in \k{common-actions} are also available.)

\H{towers-parameters} \I{parameters, for Towers}Towers parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Grid size}

\dd Specifies the size of the grid. Lower limit is 3; upper limit is
9 (because the user interface would become more difficult with
\q{digits} bigger than 9!).

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. At Unreasonable
level, some backtracking will be required, but the solution should
still be unique. The remaining levels require increasingly complex
reasoning to avoid having to backtrack.


\C{singles} \i{Singles}

\cfg{winhelp-topic}{games.singles}

You have a grid of white squares, all of which contain numbers. Your task
is to colour some of the squares black (removing the number) so as to satisfy
all of the following conditions:

\b No number occurs more than once in any row or column.

\b No black square is horizontally or vertically adjacent to any other black
square.

\b The remaining white squares must all form one contiguous region
(connected by edges, not just touching at corners).

Credit for this puzzle goes to \i{Nikoli} \k{nikoli-hitori} who call it
\i{Hitori}. 

Singles was contributed to this collection by James Harvey.

\B{nikoli-hitori}
\W{http://www.nikoli.com/en/puzzles/hitori.html}\cw{http://www.nikoli.com/en/puzzles/hitori.html}
(beware of Flash)

\H{singles-controls} \i{Singles controls}

\IM{Singles controls} controls, for Singles

Left-clicking on an empty square will colour it black; left-clicking again 
will restore the number. Right-clicking will add a circle (useful for 
indicating that a cell is definitely not black). 

You can also use the cursor keys to move around the grid. Pressing the
return or space keys will turn a square black or add a circle respectively,
and pressing the key again will restore the number or remove the circle.

(All the actions described in \k{common-actions} are also available.)

\H{singles-parameters} \I{parameters, for Singles}Singles parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle.


\C{magnets} \i{Magnets}

\cfg{winhelp-topic}{games.magnets}

A rectangular grid has been filled with a mixture of magnets (that is,
dominoes with one positive end and one negative end) and blank dominoes
(that is, dominoes with two neutral poles).
These dominoes are initially only seen in silhouette. Around the grid
are placed a number of clues indicating the number of positive and
negative poles contained in certain columns and rows.

Your aim is to correctly place the magnets and blank dominoes such that
all the clues are satisfied, with the additional constraint that no two
similar magnetic poles may be orthogonally adjacent (since they repel).
Neutral poles do not repel, and can be adjacent to any other pole. 

Credit for this puzzle goes to \i{Janko} \k{janko-magnets}.

Magnets was contributed to this collection by James Harvey.

\B{janko-magnets}
\W{http://www.janko.at/Raetsel/Magnete/index.htm}\cw{http://www.janko.at/Raetsel/Magnete/index.htm}

\H{magnets-controls} \i{Magnets controls}

\IM{Magnets controls} controls, for Magnets

Left-clicking on an empty square places a magnet at that position with
the positive pole on the square and the negative pole on the other half
of the magnet; left-clicking again reverses the polarity, and a third
click removes the magnet.

Right-clicking on an empty square places a blank domino there.
Right-clicking again places two question marks on the domino, signifying
\q{this cannot be blank} (which can be useful to note deductions while
solving), and right-clicking again empties the domino. 

Left-clicking a clue will mark it as done (grey it out), or unmark it if
it is already marked.

You can also use the cursor keys to move a cursor around the grid. 
Pressing the return key will lay a domino with a positive pole at that
position; pressing again reverses the polarity and then removes the
domino, as with left-clicking. Using the space bar allows placement
of blank dominoes and cannot-be-blank hints, as for right-clicking. 

(All the actions described in \k{common-actions} are also available.)

\H{magnets-parameters} \I{parameters, for Magnets}Magnets parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares. There will be half \e{Width} \by \e{Height}
dominoes in the grid: if this number is odd then one square will be blank.

\lcont{

(Grids with at least one odd dimension tend to be easier to solve.)

}

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle. At Tricky level,
you are required to make more deductions about empty dominoes and
row/column counts. 

\dt \e{Strip clues}

\dd If true, some of the clues around the grid are removed at generation
time, making the puzzle more difficult.


\C{signpost} \i{Signpost}

\cfg{winhelp-topic}{games.signpost}

You have a grid of squares; each square (except the last one)
contains an arrow, and some squares also contain numbers. Your job
is to connect the squares to form a continuous list of numbers
starting at 1 and linked in the direction of the arrows \dash so the
arrow inside the square with the number 1 will point to the square
containing the number 2, which will point to the square containing
the number 3, etc. Each square can be any distance away from the
previous one, as long as it is somewhere in the direction of the
arrow.

By convention the first and last numbers are shown; one or more
interim numbers may also appear at the beginning. 

Credit for this puzzle goes to \i{Janko} \k{janko-arrowpath}, who call it
\q{Pfeilpfad} (\q{arrow path}).

Signpost was contributed to this collection by James Harvey.

\B{janko-arrowpath}
\W{http://janko.at/Raetsel/Pfeilpfad/index.htm}\cw{http://janko.at/Raetsel/Pfeilpfad/index.htm}

\H{signpost-controls} \I{controls, for Signpost}Signpost controls

To play Signpost, you connect squares together by dragging from one
square to another, indicating that they are adjacent in the
sequence. Drag with the left button from a square to its successor,
or with the right button from a square to its predecessor.

If you connect together two squares in this way and one of them has
a number in it, the appropriate number will appear in the other
square. If you connect two non-numbered squares, they will be
assigned temporary algebraic labels: on the first occasion, they
will be labelled \cq{a} and \cq{a+1}, and then \cq{b} and \cq{b+1},
and so on. Connecting more squares on to the ends of such a chain
will cause them all to be labelled with the same letter.

When you left-click or right-click in a square, the legal squares to
connect it to will be shown.

The arrow in each square starts off black, and goes grey once you
connect the square to its successor. Also, each square which needs a
predecessor has a small dot in the bottom left corner, which
vanishes once you link a square to it. So your aim is always to
connect a square with a black arrow to a square with a dot.

To remove any links for a particular square (both incoming and
outgoing), left-drag it off the grid. To remove a whole chain,
right-drag any square in the chain off the grid.

You can also use the cursor keys to move around the grid squares and
lines. Pressing the return key when over a square starts a link
operation, and pressing the return key again over a square will
finish the link, if allowable. Pressing the space bar over a square
will show the other squares pointing to it, and allow you to form a
backward link, and pressing the space bar again cancels this.

(All the actions described in \k{common-actions} are also available.)

\H{signpost-parameters} \I{parameters, for Signpost}Signpost parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Force start/end to corners}

\dd If true, the start and end squares are always placed in opposite corners
(the start at the top left, and the end at the bottom right). If false the start
and end squares are placed randomly (although always both shown). 

\C{range} \i{Range}

\cfg{winhelp-topic}{games.range}

You have a grid of squares; some squares contain numbers.  Your job is
to colour some of the squares black, such that several criteria are
satisfied:

\b no square with a number is coloured black.

\b no two black squares are adjacent (horizontally or vertically).

\b for any two white squares, there is a path between them using only
white squares.

\b for each square with a number, that number denotes the total number
of white squares reachable from that square going in a straight line
in any horizontal or vertical direction until hitting a wall or a
black square; the square with the number is included in the total
(once).

For instance, a square containing the number one must have four black
squares as its neighbours by the last criterion; but then it's
impossible for it to be connected to any outside white square, which
violates the second to last criterion.  So no square will contain the
number one.

Credit for this puzzle goes to \i{Nikoli}, who have variously called
it \q{Kurodoko}, \q{Kuromasu} or \q{Where is Black Cells}.
\k{nikoli-range}.

Range was contributed to this collection by Jonas K\u00F6{oe}lker.

\B{nikoli-range}
\W{http://www.nikoli.co.jp/en/puzzles/where_is_black_cells.html}\cw{http://www.nikoli.co.jp/en/puzzles/where_is_black_cells.html}

\H{range-controls} \I{controls, for Range}Range controls

Click with the left button to paint a square black, or with the right
button to mark a square with a dot to indicate that you are sure it
should \e{not} be painted black. Repeated clicking with either button
will cycle the square through the three possible states (filled,
dotted or empty) in opposite directions.

You can also use the cursor keys to move around the grid squares.
Pressing Return does the same as clicking with the left button, while
pressing Space does the same as a right button click.  Moving with the
cursor keys while holding Shift will place dots in all squares that
are moved through.


(All the actions described in \k{common-actions} are also available.)

\H{range-parameters} \I{parameters, for Range}Range parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\C{pearl} \i{Pearl}

\cfg{winhelp-topic}{games.pearl}

You have a grid of squares. Your job is to draw lines between the
centres of horizontally or vertically adjacent squares, so that the
lines form a single closed loop. In the resulting grid, some of the
squares that the loop passes through will contain corners, and some
will be straight horizontal or vertical lines. (And some squares can
be completely empty \dash the loop doesn't have to pass through every
square.)

Some of the squares contain black and white circles, which are clues
that the loop must satisfy.

A black circle in a square indicates that that square is a corner, but
neither of the squares adjacent to it in the loop is also a corner.

A white circle indicates that the square is a straight edge, but \e{at
least one} of the squares adjacent to it in the loop is a corner.

(In both cases, the clue only constrains the two squares adjacent
\e{in the loop}, that is, the squares that the loop passes into after
leaving the clue square. The squares that are only adjacent \e{in the
grid} are not constrained.)

Credit for this puzzle goes to \i{Nikoli}, who call it \q{Masyu}.
\k{nikoli-pearl}

Thanks to James Harvey for assistance with the implementation.

\B{nikoli-pearl}
\W{http://www.nikoli.co.jp/en/puzzles/masyu.html}\cw{http://www.nikoli.co.jp/en/puzzles/masyu.html}
(beware of Flash)

\H{pearl-controls} \I{controls, for Pearl}Pearl controls

Click with the left button on a grid edge to draw a segment of the
loop through that edge, or to remove a segment once it is drawn.

Drag with the left button through a series of squares to draw more
than one segment of the loop in one go. Alternatively, drag over an
existing part of the loop to undraw it, or to undraw part of it and
then go in a different direction.

Click with the right button on a grid edge to mark it with a cross,
indicating that you are sure the loop does not go through that edge.
(For instance, if you have decided which of the squares adjacent to a
white clue has to be a corner, but don't yet know which way the corner
turns, you might mark the one way it \e{can't} go with a cross.)

Alternatively, use the cursor keys to move the cursor.  Use the Enter
key to begin and end keyboard \q{drag} operations.  Use the Space,
Escape or Backspace keys to cancel the drag.  Or, hold Control while
dragging with the cursor keys to toggle segments as you move between
squares.

Pressing Control-Shift-arrowkey or Shift-arrowkey simulates a left or
right click, respectively, on the edge in the direction of the key.

(All the actions described in \k{common-actions} are also available.)

\H{pearl-parameters} \I{parameters, for Pearl}Pearl parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\C{undead} \i{Undead}

\cfg{winhelp-topic}{games.undead}

You are given a grid of squares, some of which contain diagonal
mirrors. Every square which is not a mirror must be filled with one of
three types of undead monster: a ghost, a vampire, or a zombie.

Vampires can be seen directly, but are invisible when reflected in
mirrors. Ghosts are the opposite way round: they can be seen in
mirrors, but are invisible when looked at directly. Zombies are
visible by any means.

You are also told the total number of each type of monster in the
grid. Also around the edge of the grid are written numbers, which
indicate how many monsters can be seen if you look into the grid along
a row or column starting from that position. (The diagonal mirrors are
reflective on both sides. If your reflected line of sight crosses the
same monster more than once, the number will count it each time it is
visible, not just once.)

This puzzle type was invented by David Millar, under the name
\q{Haunted Mirror Maze}. See \k{janko-undead} for more details.

Undead was contributed to this collection by Steffen Bauer.

\B{janko-undead}
\W{http://www.janko.at/Raetsel/Spukschloss/index.htm}\cw{http://www.janko.at/Raetsel/Spukschloss/index.htm}

\H{undead-controls} \I{controls, for Undead}Undead controls

Undead has a similar control system to Solo, Unequal and Keen.

To play Undead, click the mouse in any empty square and then type a
letter on the keyboard indicating the type of monster: \q{G} for a
ghost, \q{V} for a vampire, or \q{Z} for a zombie. If you make a
mistake, click the mouse in the incorrect square and press Space to
clear it again (or use the Undo feature).

If you \e{right}-click in a square and then type a letter, the
corresponding monster will be shown in reduced size in that square, as
a \q{pencil mark}. You can have pencil marks for multiple monsters in
the same square. A square containing a full-size monster cannot also
contain pencil marks.

The game pays no attention to pencil marks, so exactly what you use
them for is up to you: you can use them as reminders that a particular
square needs to be re-examined once you know more about a particular
monster, or you can use them as lists of the possible monster in a
given square, or anything else you feel like.

To erase a single pencil mark, right-click in the square and type
the same letter again.

All pencil marks in a square are erased when you left-click and type a
monster letter, or when you left-click and press Space. Right-clicking
and pressing space will also erase pencil marks.

As for Solo, the cursor keys can be used in conjunction with the letter
keys to place monsters or pencil marks. Use the cursor keys to move a
highlight around the grid, and type a monster letter to enter it in
the highlighted square. Pressing return toggles the highlight into a
mode in which you can enter or remove pencil marks.

If you prefer plain letters of the alphabet to cute monster pictures,
you can press \q{A} to toggle between showing the monsters as monsters or
showing them as letters.

Left-clicking a clue will mark it as done (grey it out), or unmark it
if it is already marked.

(All the actions described in \k{common-actions} are also available.)

\H{undead-parameters} \I{parameters, for Undead}Undead parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle.

\C{unruly} \i{Unruly}

\cfg{winhelp-topic}{games.unruly}

You are given a grid of squares, which you must colour either black or
white. Some squares are provided as clues; the rest are left for you
to fill in. Each row and column must contain the same number of black
and white squares, and no row or column may contain three consecutive
squares of the same colour.

This puzzle type was invented by Adolfo Zanellati, under the name
\q{Tohu wa Vohu}. See \k{janko-unruly} for more details.

Unruly was contributed to this collection by Lennard Sprong.

\B{janko-unruly}
\W{http://www.janko.at/Raetsel/Tohu-Wa-Vohu/index.htm}\cw{http://www.janko.at/Raetsel/Tohu-Wa-Vohu/index.htm}

\H{unruly-controls} \I{controls, for Unruly}Unruly controls

To play Unruly, click the mouse in a square to change its colour.
Left-clicking an empty square will turn it black, and right-clicking
will turn it white. Keep clicking the same button to cycle through the
three possible states for the square. If you middle-click in a square
it will be reset to empty.

You can also use the cursor keys to move around the grid. Pressing the
return or space keys will turn an empty square black or white
respectively (and then cycle the colours in the same way as the mouse
buttons), and pressing Backspace will reset a square to empty.

(All the actions described in \k{common-actions} are also available.)

\H{unruly-parameters} \I{parameters, for Unruly}Unruly parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares. (Note that the rules of the game require
both the width and height to be even numbers.)

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle.

\dt \e{Unique rows and columns}

\dd If enabled, no two rows are permitted to have exactly the same
pattern, and likewise columns. (A row and a column can match, though.)

\C{flood} \i{Flood}

\cfg{winhelp-topic}{games.flood}

You are given a grid of squares, coloured at random in multiple
colours. In each move, you can flood-fill the top left square in a
colour of your choice (i.e. every square reachable from the starting
square by an orthogonally connected path of squares all the same
colour will be filled in the new colour). As you do this, more and
more of the grid becomes connected to the starting square.

Your aim is to make the whole grid the same colour, in as few moves as
possible. The game will set a limit on the number of moves, based on
running its own internal solver. You win if you can make the whole
grid the same colour in that many moves or fewer.

I saw this game (with a fixed grid size, fixed number of colours, and
fixed move limit) at http://floodit.appspot.com (no longer accessible).

\H{flood-controls} \I{controls, for Flood}Flood controls

To play Flood, click the mouse in a square. The top left corner and
everything connected to it will be flood-filled with the colour of the
square you clicked. Clicking a square the same colour as the top left
corner has no effect, and therefore does not count as a move.

You can also use the cursor keys to move a cursor (outline black
square) around the grid. Pressing the return key will fill the top
left corner in the colour of the square under the cursor.

(All the actions described in \k{common-actions} are also available.)

\H{flood-parameters} \I{parameters, for Flood}Flood parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of the grid, in squares.

\dt \e{Colours}

\dd Number of colours used to fill the grid. Must be at least 3 (with
two colours there would only be one legal move at any stage, hence no
choice to make at all), and at most 10.

\dt \e{Extra moves permitted}

\dd Controls the difficulty of the puzzle, by increasing the move
limit. In each new grid, Flood will run an internal solver to generate
its own solution, and then the value in this field will be added to
the length of Flood's solution to generate the game's move limit. So a
value of 0 requires you to be just as efficient as Flood's automated
solver, and a larger value makes it easier.

\lcont{

(Note that Flood's internal solver will not necessarily find the
shortest possible solution, though I believe it's pretty close. For a
real challenge, set this value to 0 and then try to solve a grid in
\e{strictly fewer} moves than the limit you're given!)

}

\C{tracks} \i{Tracks}

\cfg{winhelp-topic}{games.tracks}

You are given a grid of squares, some of which are filled with train
tracks. You need to complete the track from A to B so that the rows and
columns contain the same number of track segments as are indicated in the
clues to the top and right of the grid.

There are only straight and 90 degree curved rails, and the track may not
cross itself.

Tracks was contributed to this collection by James Harvey.

\H{tracks-controls} \I{controls, for Tracks}Tracks controls

Left-clicking on an edge between two squares adds a track segment between
the two squares. Right-clicking on an edge adds a cross on the edge,
indicating no track is possible there.

Left-clicking in a square adds a colour indicator showing that you know the
square must contain a track, even if you don't know which edges it crosses
yet. Right-clicking in a square adds a cross indicating it contains no
track segment.

Left- or right-dragging between squares allows you to lay a straight line
of is-track or is-not-track indicators, useful for filling in rows or
columns to match the clue.

(All the actions described in \k{common-actions} are also available.)

\H{tracks-parameters} \I{parameters, for Tracks}Tracks parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of the grid, in squares.

\dt \e{Difficulty}

\dd Controls the difficulty of the generated puzzle: at Tricky level,
you are required to make more deductions regarding disregarding moves
that would lead to impossible crossings later.

\dt \e{Disallow consecutive 1 clues}

\dd Controls whether the Tracks game generation permits two adjacent
rows or columns to have a 1 clue, or permits the row or column of the
track's endpoint to have a 1 clue. By default this is not permitted,
to avoid long straight boring segments of track and make the games
more twiddly and interesting. If you want to restore the possibility,
turn this option off.


\C{palisade} \i{Palisade}

\cfg{winhelp-topic}{games.palisade}

You're given a grid of squares, some of which contain numbers.  Your
goal is to subdivide the grid into contiguous regions, all of the same
(given) size, such that each square containing a number is adjacent to
exactly that many edges (including those between the inside and the
outside of the grid).

Credit for this puzzle goes to \i{Nikoli}, who call it \q{Five Cells}.
\k{nikoli-palisade}.

Palisade was contributed to this collection by Jonas K\u00F6{oe}lker.

\B{nikoli-palisade}
\W{http://nikoli.co.jp/en/puzzles/five_cells.html}\cw{http://nikoli.co.jp/en/puzzles/five_cells.html}

\H{palisade-controls} \I{controls, for Palisade}Palisade controls

Left-click to place an edge.  Right-click to indicate \q{no edge}.
Alternatively, the arrow keys will move a keyboard cursor.  Holding
Control while pressing an arrow key will place an edge.  Press
Shift-arrowkey to switch off an edge.  Repeat an action to perform
its inverse.

(All the actions described in \k{common-actions} are also available.)

\H{Palisade-parameters} \I{parameters, for Palisade}Palisade parameters

These parameters are available from the \q{Custom...} option on the
\q{Type} menu.

\dt \e{Width}, \e{Height}

\dd Size of grid in squares.

\dt \e{Region size}

\dd The size of the regions into which the grid must be subdivided.

\A{licence} \I{MIT licence}\ii{Licence}

This software is \i{copyright} 2004-2014 Simon Tatham.

Portions copyright Richard Boulton, James Harvey, Mike Pinna, Jonas
K\u00F6{oe}lker, Dariusz Olszewski, Michael Schierl, Lambros Lambrou,
Bernd Schmidt, Steffen Bauer, Lennard Sprong and Rogier Goossens.

Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation files
(the \q{Software}), to deal in the Software without restriction,
including without limitation the rights to use, copy, modify, merge,
publish, distribute, sublicense, and/or sell copies of the Software,
and to permit persons to whom the Software is furnished to do so,
subject to the following conditions:

The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED \q{AS IS}, WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

\IM{command-line}{command line} command line

\IM{default parameters, specifying} default parameters, specifying
\IM{default parameters, specifying} preferences, specifying default

\IM{Unix} Unix
\IM{Unix} Linux

\IM{generating game IDs} generating game IDs
\IM{generating game IDs} game ID, generating

\IM{specific} \q{Specific}, menu option
\IM{custom} \q{Custom}, menu option

\IM{game ID} game ID
\IM{game ID} ID, game
\IM{ID format} ID format
\IM{ID format} format, ID
\IM{ID format} game ID, format

\IM{keys} keys
\IM{keys} shortcuts (keyboard)

\IM{initial state} initial state
\IM{initial state} state, initial

\IM{MIT licence} MIT licence
\IM{MIT licence} licence, MIT

\versionid Simon Tatham's Portable Puzzle Collection, version 20170606.272beef