Simon Tatham 6285c44610 Group: hard-mode identity deduction.
This fills in the deduction feature I mentioned in commit 7acc554805,
of determining the identity by elimination, having ruled out all other
candidates.

In fact, it goes further: as soon as we know that an element can't be
the group identity, we rule out every possible entry in its row and
column which would involve it acting as a left- or right-identity for
any individual element.

This noticeably increases the number of puzzles that can be solved at
Hard mode without resorting to Unreasonable-level recursion. In a test
of 100 Hard puzzles generated with this change, 80 of them are
reported as Unreasonable by the previous solver.

(One of those puzzles is 12i:m12b9a1zd9i6d10c3y2l11q4r , the example
case that exposed the latin.c validation bug described by the previous
two commits. That was reported as ambiguous with the validation bug,
as Unreasonable with the validation bug fixed, and now it's merely
Hard, because this identity-based deduction eliminates the need for
recursion.)
2020-05-23 09:08:25 +01:00
2018-01-21 19:03:38 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2019-04-14 21:24:19 +01:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2019-12-30 08:10:34 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-06-01 06:50:00 +01:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2015-10-18 17:53:28 +01:00
2019-12-30 08:10:34 +00:00
2017-09-20 18:03:44 +01:00
2019-12-30 08:10:34 +00:00
2018-11-13 21:48:24 +00:00
2017-05-07 16:25:56 +01:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2018-11-13 21:48:24 +00:00
2020-02-26 06:32:35 +00:00
2018-11-13 21:48:24 +00:00
2019-09-01 22:26:22 +01:00
2018-11-13 21:48:24 +00:00
2017-05-07 16:25:56 +01:00
2017-05-07 16:25:56 +01:00

This is the README accompanying the source code to Simon Tatham's
puzzle collection. The collection's web site is at
<https://www.chiark.greenend.org.uk/~sgtatham/puzzles/>.

If you've obtained the source code by downloading a .tar.gz archive
from the Puzzles web site, you should find several Makefiles in the
source code. However, if you've checked the source code out from the
Puzzles git repository, you won't find the Makefiles: they're
automatically generated by `mkfiles.pl', so run that to create them.

The Makefiles include:

 - `Makefile.am', together with the static `configure.ac', is intended
   as input to automake. Run `mkauto.sh' to turn these into a
   configure script and Makefile.in, after which you can then run
   `./configure' to create an actual Unix Makefile.

 - `Makefile.vc' should work under MS Visual C++ on Windows. Run
   'nmake /f Makefile.vc' in a Visual Studio command prompt.

 - `Makefile.cyg' should work under Cygwin / MinGW. With appropriate
   tweaks and setting of TOOLPATH, it should work for both compiling
   on Windows and cross-compiling on Unix.

 - `Makefile.osx' should work under Mac OS X, provided the Xcode
   tools are installed. It builds a single monolithic OS X
   application capable of running any of the puzzles, or even more
   than one of them at a time.

 - `Makefile.wce' should work under MS eMbedded Visual C++ on
   Windows and the Pocket PC SDK; it builds Pocket PC binaries.

Many of these Makefiles build a program called `nullgame' in
addition to the actual game binaries. This program doesn't do
anything; it's just a template for people to start from when adding
a new game to the collection, and it's compiled every time to ensure
that it _does_ compile and link successfully (because otherwise it
wouldn't be much use as a template). Once it's built, you can run it
if you really want to (but it's very boring), and then you should
ignore it.

DO NOT EDIT THE MAKEFILES DIRECTLY, if you plan to send any changes
back to the maintainer. The makefiles are generated automatically by
the Perl script `mkfiles.pl' from the file `Recipe' and the various
.R files. If you need to change the makefiles as part of a patch,
you should change Recipe, *.R, and/or mkfiles.pl.

The manual is provided in Windows Help format for the Windows build;
in text format for anyone who needs it; and in HTML for the Mac OS X
application and for the web site. It is generated from a Halibut
source file (puzzles.but), which is the preferred form for
modification. To generate the manual in other formats, rebuild it,
or learn about Halibut, visit the Halibut website at
<https://www.chiark.greenend.org.uk/~sgtatham/halibut/>.
Description
No description provided
Readme 26 MiB
Languages
C 93.3%
JavaScript 1.4%
Objective-C 1.1%
CMake 1.1%
HTML 0.8%
Other 2.2%