background colour for covered and uncovered squares in Mines, since
otherwise you have to distinguish them by the edge highlights alone.
So here one is; it's not _very_ different (it just looked odd if it
was any darker than this), but anyone who wants a bigger difference
can reconfigure it using the MINES_COLOUR_1 environment variable.
[originally from svn r5876]
were already making sure that no shuffle move was the precise
inverse of the previous one, or contributed to repeating the
previous one so many times as to turn it into effectively fewer
moves (doing the same rotation three times in Twiddle, or shifting a
row by more than half its length in Sixteen). However, they were
only checking against the _last_ move, which meant that in any
situation where there were completely disjoint move spaces (4x4n2
Twiddle, or any Sixteen at all) it was still possible to have A then
B then inv(A) occurring in the shuffle, leading to an unnecessarily
easy game.
Now both shuffle routines keep separate track of all
_non-overlapping_ recent moves, and will avoid inverting any move
which hasn't had another move overlap it since it was made. This
should reduce the incidence of too-easy limited shuffle games,
although it can't be prevented _entirely_ (since, if nothing else,
it's always possible to increase the shuffle limit past the maximum
group radius).
[originally from svn r5875]
puzzle of a different size to be redrawn before the pixmap is
resized, and since backends never redraw already-drawn stuff this is
a problem. Was biting me when I entered a Mines game ID of a
different size than the current settings into the Specific box.
[originally from svn r5872]
between on the one hand generating indeterminate game descriptions
awaiting the initial click, and on the other hand generating
concrete ones which have had their initial click. This makes `mines
--generate' do something useful.
[originally from svn r5869]
indicates whether a particular game state should have the timer
going (for Mines the initial indeterminate state does not have this
property, and neither does a dead or won state); a midend function
that optionally (on request from the game) prepends a timer to the
front of the status bar text; some complicated midend timing code.
It's not great. It's ugly; it's probably slightly inaccurate; it's
got no provision for anyone but the game author decreeing whether a
game is timed or not. But Mines can't be taken seriously without a
timer, so it's a start.
[originally from svn r5866]
blank grid until you make the first click; to ensure solubility, it
does not generate the mine layout until that click, and then ensures
it is solvable starting from that position.
This has involved three infrastructure changes:
- random.c now offers functions to encode and decode an entire
random_state as a string
- each puzzle's new_game() is now passed a pointer to the midend
itself, which most of them ignore
- there's a function in the midend which a game can call back to
_rewrite_ its current game description.
So Mines now has two entirely separate forms of game ID. One
contains the generation-time parameters (n and unique) plus an
encoding of a random_state; the other actually encodes the grid once
it's been generated, and also contains the initial click position.
When called with the latter, new_game() does plausibly normal stuff.
When called with the former, it notes down all the details and waits
until the first square is opened, and _then_ does the grid
generation and updates the game description in the midend. So if,
_after_ your first click, you decide you want to share this
particular puzzle with someone else, you can do that fine.
Also in this checkin, the mine layout is no longer _copied_ between
all the game_states on the undo chain. Instead, it's in a separate
structure and all game_states share a pointer to it - and the
structure is reference-counted to ensure deallocation.
[originally from svn r5862]
both get passed a pointer to the game_ui. This means that if they
need to note down information for the redraw function about what
_type_ of flash or animation is required, they now have somewhere to
do so.
[originally from svn r5858]
puzzles, rather than just doing its nondeterministic number
placement thing. This enables the use of the `Solve' menu option on
externally entered game IDs, provided of course that they aren't
_too_ difficult.
[originally from svn r5852]
Jacob wasn't able to find a satisfactory fix: the whole area was a
horrid mess. Fortunately, the reason it was a horrid mess was
because the Net drawing routines predated the introduction of clip()
in the frontend interface, and it turns out that clip() makes it
possible to do all this more easily and better. So, a complete
rearchitecting of barrier corners: the corner flags in the
`barriers' array are now gone (and good riddance), and corner
information is computed on the fly so as to take into account the
moving grid edges. In the process I've also updated the corner
mechanism so that a barrier `corner' (really endpoint) is drawn at
the end of _every_ barrier, not just where two meet. This has
changed the appearance of a single isolated barrier, to what I would
have wanted it to look like in the first place but achieving it
without clip() was just too fiddly.
[originally from svn r5846]
[r5844 == 865e8ad6ca3d83ad2a585ceeb1809e9f68c18a20]
(Adding modifier+cursors handling has had minor knock-on effects on the other
puzzles, so that they can continue to ignore modifiers.)
(An unfortunate side effect of this is some artifacts in exterior barrier
drawing; notably, a disconnected corner can now appear at the corner of the
grid under some circumstances. I haven't found a satisfactory way round
this yet.)
[originally from svn r5844]
number. Many thanks to Chris Thomas, for helping with the detailed
UI design by means of testing an endless series of prototypes.
[originally from svn r5842]
unique solution. This, it turns out, is because there is literally
no such thing. Protective constraint added to validate_params(),
with a proof in a comment alongside.
If you really want a 2xn or nx2 wrapping puzzle, you can still have
one if you turn uniqueness off.
[originally from svn r5835]
The previous checkin stopped it choking on them, but it didn't
actually manage to _deduce_ that all the edges bordering them had to
be closed. Now it does better.
[originally from svn r5829]
over on a grid containing a 0 (completely blank) tile. This can't
happen in self-generated grids, but can happen if you type in a grid
from another Net implementation. Previously, the solver would notice
(technically correctly!) that a completely blank tile connects to no
other tiles and thus forms an isolated subgraph, and would therefore
complain that no orientation of that tile could possibly yield a
valid solution...
[originally from svn r5828]
- fix documentation of Net's unique solution option (should have
tested before last checkin)
- make unique solutions optional in Rectangles too (same reasons)
- tidy up various issues in parameter encoding in both games.
[originally from svn r5818]
enabled by default), since ambiguous sections in grids can present
additional interesting challenges. I think uniqueness is a better
default, though.
[originally from svn r5816]
puzzles. I generate the grid of rectangles as normal, but before I
place the numbers I run it through a non-deterministic solver
algorithm which tries to do as much as it can with as little
information about where the numbers are going to be. The solver
itself narrows down the number placement when it runs out of steam,
but does so as little as possible. Once it reaches a state where it
has ensured solubility, and then the generation algorithm chooses
random number placement from whatever's left.
Occasionally it paints itself into a corner and can't ensure a
unique solution no matter what happens; in that situation we just
have to give up, generate a fresh grid, and try again.
[originally from svn r5809]
from the currently configured ones, and then bring the Random Seed
box back up, the wrong parameters get shown, and the resulting
random seed is incorrect for the current game.
At least, it was, until this checkin.
[originally from svn r5808]
cross-platform, and rename the environment variables so that they
follow the puzzle name. Should allow a static environment
configuration for each puzzle. Also introduced a <game>_PRESETS
variable for people whose favourite configuration isn't on the Type
menu by default.
[originally from svn r5801]
particularly useful keypress, particularly given how easy it is to
confuse it with `Redo'. So both r and ^R are now Redo, and Restart
is relegated to being a menu-only option.
[originally from svn r5796]
Gareth and I have been using to analyse the game, and also the
number pad. They don't work sensibly for all sizes, but they'll be
handy for the most common ones.
[originally from svn r5793]
of the undo list rather than destroying it. Partly this is because
accidental restarts are a real pain, and partly because it allows
you to compare the initial to the current state by restart-then-undo
which is handy in some puzzles.
In order to do this, I've introduced an additional per-entry field
in the undo list in the midend, which tracks which states were
created by `unusual' operations (Solve and Restart). The midend
takes care of suppressing animation and completion flashes during
transitions between a `special' state and its predecessor, relieving
the game backends of having to do it individually.
(This probably means I could remove some complexity in the
flash_time() functions in most backends, but I haven't done that in
this checkin.)
[originally from svn r5791]
