Skip to content

steenpass/Singular

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

How to compile SINGULAR from sources
************************************

SINGULAR is a Computer Algebra system for polynomial computations with
special emphasize on the needs of commutative algebra, algebraic
geometry, singularity theory and polynomial system solving. For a more
detailed overview of SINGULAR, see
     `http://www.singular.uni-kl.de/Overview/'

So far, there are only versions for Linux, HPUX, SunOS, IRIX, AIX,
FreeBSD 4.x/5.x, Mac OS X, and Windows 95/98/NT/2K/XP.

Version 2-0-2 was the last version for MacOS 9.

The Windows version also runs on Windows Vista, provided
that "rebase all" was done after installing Cygwin.

Note for Singular-3-0-x:
Currently, there are only only "complete" archives: Singular-3-0-x-x.tar.gz
which contain all needed files. 

Compiling SINGULAR for a Unix platform
======================================

Make sure that you have approximately 120 MByte of free disk space and
follow these steps.
(the "optional" for bison applies to version 3-0-3-1 and later)

  1. Make sure that you have the following software (the list may be
     incomplete):
       1.   general unix utilities: grep, test, sed, cp, mv, gzip,
          uudecode, ...

       2.   autoconf

       3.   GNU make

       4.   gcc, g++ (version 2.95.3 preferred,   but gcc version 2.7.2
          to 2.95.4, egcs 1.0.3 and 1.1.2 should work,   gcc 2.96 does
          not work)
          gcc 3.0 does not work, gcc 3.1.x, gcc 3.2.x, gcc 3.3.x/3.4.x should
          gcc 4.x: beginnig with Singular-3-0-0-4, for some architectures
                   beginning with Singular-3-0-2: for all architectures

       5.   optional: if you want to change Singular/grammar.y:
          bison (version >= 1.2.2)
            (bison (GNU Bison) 1.75 seems not to work)
            (version 1.75 is supported in Singular 2-0-4a and newer)
            
            Remark: 
            bison 1.875 on solaris produces code which gcc does not accept

       6.   optional: if you want to change Singular/scanner.l
          or modules/modgen/scanner.l:
          flex (version >=2.4, <=2.5.4)

       7.   perl (version >=5)

       8.   library gmp (version >=3.1.1)
          (version 4.x.x is supported in Singular 2-0-4 and newer)

       9.   library libncurses or libtermcap or libcurses

      10.   (optional) library libreadline
          (required on linux glibc systems)

      11.   (optional, required to build the documentation:)
          latex2html, tex, texi2html, texinfo
          (your latex2html should produce gif files, not png files)

      12.   (optional, required to build gfanlib and its interface,
          required for handling convex objects in Singular in general)
          cddlib with GMP support. you can download a patched cddlib 
          based on the latest Gentoo patch, supporting shared libraries
          and building the GMP version by default, from:
          http://www.mathematik.uni-kl.de/~ren/cddlib-patched.tbz2
          install it by running "./configure" and "make install".

      13.   (optional, required to build the polymake interface)
          polymake version 2.12 or higher

  2. (optional, required for MP:)
     make sure that you can log in to your computer via rsh/remsh or
     ssh without a password (check .rhosts/.shosts);
     (in versions before 3-0-0: the
     configure script stops while checking this: if you cannot fix it,
     simply kill the rsh/ssh-processes - the configure will then
     disable this particulare feature)
     The following should work:
     rsh localhost uname
     ssh localhost uname
     rsh `hostname` uname
     ssh `hostname` uname

  3. You need to download all of the .tar.gz-files
     (since 3-0-0 we have only one large source .tar.gz-file)

  4. create a new working directory

  5. unpack all packages into that working directory

  6. (optional, for porting to a unknown machine:) fix ./singuname.sh
     not to report something unknown, but to recognize the machine

  7. run ./configure
     or run "./configure --enable-gfanlib" in order to build Singular with 
     Gfanlib, i.e. Singular with support for convex objects.
     or run "./configure --enable-polymake" in order to build Singular with 
     Polymake interface. Gfanlib will be automatically installed as well.
     (try ./configure -help, ./omalloc/configure -help, ./MP/configure
     -help, ./factory/configure -help, ./Singular/configure -help for a
     list of all options of configure)

     BEWARE: some of the option are for testing only,
     they will usually not work in general:
     --with-Plural,--with-plural (you need a Singular version > 2.1.2 for this)
       (standard since  version 3-0-0)
     --with-sgroup (you need the sgroup programs for this)
     --with-smallgmp (not maintained), removed in 3-0-0
     --with-namespaces (you need a Singular version > 2.1.x for this)
       (standard since  version 3-0-0)
     

  8. make install (resp. gmake install - the make program must be
     GNU Make) the make process will (hopefully) build:
       1.   omalloc: the memory management routines (required)

       2.   MP: Multi-Protocoll (optional)

       3.   NTL (optional, but strongly recommended)

       4.   factory and libfac 
            (optional, but strongly recommended:
	    some functionality will be lost if they are missing)

       5.   kernel, Singular (Kernel, required)

       6.   only for share-dist: doc
            (only possible, if all optional packages are included)

       7.   only for share-dist: emacs (only possible, if doc was build)

       8.   IntegerProgramming

  9. if you want to make any change in the directory Singular, run make
     depend there

 10. to install the resulting files at any other place in the system,
     run
     make bindist sharedist
     (sharedist can only be build if doc was build sucessfully) and
     install the resulting .tar.gz files as explained for the binary
     distribution
     The sharedist-result does not depend on your machine:
     so you can also download our version of it

 11. if you compiled Singular for a non-standard architecture and you
     are willing to share your porting effort, upload the changed
     sources and the resulting binaries to
     `ftp://www.mathematik.uni-kl.de/pub/Incoming' and send us an email
     about it ([email protected])

 12. bug-fixes or other improvements to the source code are welcome

Notes for specific platforms
============================

   - Windows (Cygwin):
     ./configure -with-malloc=system
     Windows Vista:
     run         
       rebaseall
     before trying anything else

   - Mac OS X:
     you need a file values.h where MP can find it (we copied one from
     Linux)
     ./configure -with-malloc=system

   - FreeBSD 4.x/5.x:
     you need a file values.h where MP can find it (we copied one from
     Linux)
     ./configure -with-malloc=system

   - Solaris with CC/cc
     - you need to set CONFIG_SHELL to /bin/bash
      (/bin/sh misses some substitutions in the generated Makefiles)
     - set CC = "cc -xarch=v9"
          CXX = "CC -xarch=v9"
          CFLAGS          = -O
          CXXFLAGS        = -O2
     - configure with
      ./configure  --without-dynamic-kernel --without-MP
      

Compiling SINGULAR for a Non-Unix platform
==========================================

This is a non-trivial task. Please contact us directly.
(`mailto:[email protected]')

GOOD LUCK and ENJOY!  Your SINGULAR team.


Compiling SINGULAR with cmake
=============================

compiling SINGULAR with cmake is currently experimental and 
under development.
Use it at your own risk! Parents are liable for their children! 

About

sources for the Singular computer algebra system

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 40.5%
  • Scilab 34.0%
  • C 20.0%
  • Java 1.9%
  • TeX 1.1%
  • Emacs Lisp 1.1%
  • Other 1.4%