[Gcl-devel] maxima for windows with modern gcc/gcl

2012/1/19 Donald Winiecki <dwiniecki at boisestate.edu>

> Stefano (or anyone!),
>
> The attached 7zip file contains documentation and other files to guide the
> creation of a Win32 installer for GCL.  The scripts work in `Inno Setup',
> available from http://www.jrsoftware.org/isinfo.php
>
> Let me know if you run into any bumps.
>
> Best,
>
> _don
>
>

Thank you for that, I will try to follow your suggestions as soon as I can
(now I have no access to a win32 machine).

Same thing for a new install.win32, as soon as I have time I will write
what I have learned.

Stefano



>
> On Thu, Jan 19, 2012 at 2:13 AM, Stefano Ferri <ferriste at gmail.com> wrote:
>
>> Thanks David and Camm. I've succesfully built gcl with the latest
>> Mingw/Msys without pain. A simple checkout and a configure worked fine.
>> Yes, I've noticed too that the option to disable xdr was broken, I didn't
>> mention this fact in my previous email.
>> Now I will try to finish the build process and the creation of the
>> installer (just to learn how to do it).
>>
>> I think we should update the information given by install.win32, they are
>> obsolete in some points and now the procedure is easier, at least for the
>> gcl-related part.
>>
>> Stefano
>>
>>
>>
>>
>> 2012/1/17 Camm Maguire <camm at maguirefamily.org>
>>
>>> Greetings!  Great to hear of your progress.  BTW, I build frequently
>>> under latest wine/mingw packages in Debian, together with
>>> maxima/acl2/axiom, without issue.  Had a native windows box too recently
>>> and got things working.
>>>
>>> axiom needs xdr.  Latest glibc moves the functions, and my configure fix
>>> was too stringent.  How about this:
>>>
>>> --- configure.in        16 Jan 2012 20:08:03 -0000
>>>  1.112.4.1.2.2.2.47.2.3.2.1.4.2.4.2.4.70
>>> +++ configure.in        17 Jan 2012 17:21:56 -0000
>>> @@ -66,7 +66,7 @@
>>>  [try_japi=$enableval],[try_japi="no"])
>>>
>>>  AC_ARG_ENABLE(xdr,[ --enable-xdr=yes will compile in support for XDR],
>>> -[try_xdr=$enableval],[try_xdr="no"])
>>> +[enable_xdr=$enableval],[enable_xdr="yes"])
>>>
>>>  AC_ARG_ENABLE(xgcl,[ --enable-xgcl=yes will compile in support for
>>> XGCL],
>>>  [enable_xgcl=$enableval],[enable_xgcl="yes"])
>>> @@ -1046,8 +1046,12 @@
>>>  fi
>>>
>>>
>>> -AC_CHECK_FUNC(xdr_double,,
>>> -          AC_CHECK_LIB(tirpc,xdr_double,TLIBS="$TLIBS
>>> -ltirpc",AC_MSG_ERROR([Need xdr_double])))
>>> +if test "$enable_xdr" = "yes" ; then
>>> +   AC_CHECK_FUNC(xdr_double,AC_DEFINE(HAVE_XDR),
>>> +          AC_CHECK_LIB(tirpc,xdr_double,AC_DEFINE(HAVE_XDR)
>>> TLIBS="$TLIBS -ltirpc",
>>> +          AC_CHECK_LIB(rpc,xdr_double,AC_DEFINE(HAVE_XDR) TLIBS="$TLIBS
>>> -lrpc",
>>> +          AC_CHECK_LIB(oncrpc,xdr_double,AC_DEFINE(HAVE_XDR)
>>> TLIBS="$TLIBS -loncrpc"))))
>>> +fi
>>>
>>>  AC_MSG_CHECKING(__builtin___clear_cache)
>>>  AC_TRY_COMPILE([],
>>> @@ -1373,17 +1377,17 @@
>>>                                 EXTRA_LOBJS="${EXTRA_LOBJS} gcl_japi.o"
>>>                                 LIBS="${LIBS} -ljapi -lwsock32"] )
>>>  fi
>>> -if test "$use" = "mingw" ; then
>>> - if test "$try_xdr" = "yes" ; then
>>> -   AC_CHECK_HEADERS(rpc/rpc.h,[AC_DEFINE(HAVE_XDR)
>>> -                                LIBS="${LIBS} -loncrpc"] )
>>> - fi
>>> -else
>>> - if test "$try_xdr" = "yes" ; then
>>> -   AC_CHECK_HEADERS(rpc/rpc.h,[AC_DEFINE(HAVE_XDR)
>>> -                                LIBS="${LIBS} -lrpc"] )
>>> - fi
>>> -fi
>>> +dnl if test "$use" = "mingw" ; then
>>> +dnl  if test "$try_xdr" = "yes" ; then
>>> +dnl    AC_CHECK_HEADERS(rpc/rpc.h,[AC_DEFINE(HAVE_XDR)
>>> +dnl                             LIBS="${LIBS} -loncrpc"] )
>>> +dnl  fi
>>> +dnl else
>>> +dnl  if test "$try_xdr" = "yes" ; then
>>> +dnl    AC_CHECK_HEADERS(rpc/rpc.h,[AC_DEFINE(HAVE_XDR)
>>> +dnl                             LIBS="${LIBS} -lrpc"] )
>>> +dnl  fi
>>> +dnl fi
>>>
>>>  # Should really find a way to check for prototypes, but this
>>>  # basically works for now.  CM
>>>
>>> David Billinghurst <dbmaxima at gmail.com> writes:
>>>
>>> > A couple of days ago I posted a windows installer for maxima-5.26.0 at
>>> >
>>> > http://billinghurst.customer.netspace.net.au/maxima-5.26.0.exe
>>> >
>>> > See http://www.math.utexas.edu/pipermail/maxima/2012/027451.html
>>> >
>>> > Having set up the build system, I decided to try it on a current
>>> mingw/msys installation. As some of you know, it is a tricky build as gcl on
>>> > windows is touchy.  The last version of gcc that could be used to
>>> build gcl was gcc-3.3.1 from August 2003, and you need to have an ancient
>>> mingw/
>>> > msys install.  See the INSTALL.win32 file in maxima cvs for the
>>> details.  There has been some progress with gcl.  One can now build gcl
>>> with the
>>> > latest mingw32/msys using gcc-4.6.2 and successfully build a working
>>> maxima.  While the build works on 32-bit WinXP, I haven't managed to
>>> reproduce
>>> > it (yet) on a 64-bit Windows 7 machine.
>>> >
>>> > Anyway, I have uploaded a test installer of maxima cvs 2012-01-16
>>> using gcc-4.6.2 to
>>> >
>>> > http://billinghurst.customer.netspace.net.au/maxima-5.26.0-test.exe
>>> >
>>> > This is still work in progress, and some documentation is required,
>>> but I'd welcome feedback.
>>> >
>>> >         David
>>> >
>>> > _______________________________________________
>>> > Maxima mailing list
>>> > Maxima at math.utexas.edu
>>> > http://www.math.utexas.edu/mailman/listinfo/maxima
>>>
>>> --
>>> Camm Maguire
>>> camm at maguirefamily.org
>>>
>>> ==========================================================================
>>> "The earth is but one country, and mankind its citizens."  --
>>>  Baha'u'llah
>>> _______________________________________________
>>> Maxima mailing list
>>> Maxima at math.utexas.edu
>>> http://www.math.utexas.edu/mailman/listinfo/maxima
>>>
>>
>>
>> _______________________________________________
>> Gcl-devel mailing list
>> Gcl-devel at gnu.org
>> https://lists.gnu.org/mailman/listinfo/gcl-devel
>>
>>
>