This is the new (as of November 2011) version of Fermat for Windows. It uses exactly the same C code as the Linux and Mac versions. For
the old, now obsolete, Windows version see http://home.bway.net/lewis/owind.html. That old version is distinctly inferior in many ways.
This software was created with Cygwin. See Cygwin. Thank you Cygwin!
Download this Windows version of Fermat, 3.9.9x, here. Expand. To run, just double-click on ferl.exe.
The download includes the important file cygwin1.dll. That file is freely available on the Cygwin website too. Unless you already have it, leave it in the Fermat folder.
This version runs on every Windows system that I have tried, including Windows 7.
Since it is exactly the same as the Linux and Mac version, the manual is the same. Revised PDF and HTML Manuals.
This is 32 bit code. It runs fine on Windows XP. A user reported that it works on Windows 7. Since it is new, it should be considered a beta version.
The only difference I have seen is that some matrix normalization routines (determinant, Pseudet) run more slowly than on Linux unless the pivot strategy &u = 5 is
selected. I don't know why this should be.
Windows does not seem to manage RAM (heap, stack) quite so well as Linux or Mac. Researchers with demanding problems would probably be better using Linux or Mac.
Fermat is freeware. Institutions or those with grant money should contribute $US 60 per machine.
There was a bug in a speedup technique involving multiplying large two variable polynomials. This would manifest as a "Fermat error". Also,
refined heuristics for multivariable polynomial gcd.
Version 3.9.9x, November 6, 2012. Mac OSX Intel, Windows, Linux:
There was a bug in Pseudet when the matrix was not of maximal rank and pivot strategy 5 was chosen. It would manifest unpredictably, but
inevitably Pseudet would finish too soon with the wrong answer.
Version 3.9.9x, June 12, 2012. Mac OSX Intel, Windows, Linux:
There was a bug in a speedup technique for multiplying large two variable polynomials. It could fail if the degrees were large enough. This would manifest as a "Fermat error".
Version 3.9.9x, January 2, 2012. Mac OSX Intel, Windows, Linux:
New feature:
- $, which means greatest integer in Fermat, will now similarly truncate a quolynomial. This is most useful in the new float versions. For example, with floats, using Gaussian elimination,
the determinant of a matrix containing polynomials might be returned as a rational function ("quolynomial") due to round-off errors. Applying $ to the answer should produce a very
good approximation.
Bug fixes:
- Polynomial GCD when polymodding over Z had a rare bug.
Please email me about any problems that you encounter.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNERS OR CONTRIBUTORS
BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
Please note that although the author plans to release enhanced versions of this software, the author cannot guarantee indefinite software support.
since February 5, 1999
(Thanks WebCounter!)