This is William Hart's GPL'd highly optimized multi-polynomial
quadratic sieve for integer factorization:

       http://www.friedspace.com/QS/

See also http://www.maths.warwick.ac.uk/~masfaw/preprint.html
See also the svn server:
 https://svn.sourceforge.net/svnroot/fastlibnt/trunk/QS/

To get a new version:

svn export https://svn.sourceforge.net/svnroot/fastlibnt/trunk/QS/ fastlibnt

== Changelog ==

=== flintqs-20070817.p8 (Jeroen Demeyer, 21 June 2012) ===
 * Trac #9544: Add 'stdint.patch' to fix FreeBSD build.

=== flintqs-20070817.p7 (Jeroen Demeyer, 25 May 2012) ===
 * Trac #12855: Only call log() on float or double types.  This works
   around a problem with Solaris' libc where log() in C++ is overloaded
   for float, double and long double but not integer types.
 * Use `patch` for patching, use standard template for spkg-install.
 * Don't retry the build upon failure on x86_64, which seems pointless 
   after #11351.

=== flintqs-20070817.p6 (Mariah Lenox, 19 May 2011) ===
 * Trac #11351: make x86_64 systems build with -m64 rather than
   -march=opteron.

=== flintqs-20070817.p5 ===
 * no record

=== flintqs-20070817.p4 (David Kirkby, 24th May 2010) ===
 * #9034 Use the 'makefile.osx64' for any 64-bit system, not just 
   OS X, as it is quite generic. This allows flitqs to build 
   64-bit on OpenSolaris, and will probably build 64-bit on any half
   reasonable operating system if SAGE64 is set to "yes"
 * Remove the 'dist' subdirectory - see #5903

=== flintqs-20070817.p4 (Michael Abshoff, April 14th, 2008) ===
 * Fix OSX 64 bit build

=== flintqs-20070817.p3 (Michael Abshoff, April 14th, 2008) ===
 * apply gcc 4.3 build patch. This has to go upstream