#7021: [with patch; needs work] Update prereq from 0.3 to 0.4
-------------------------+--------------------------------------------------
Reporter: drkirkby | Owner: tbd
Type: enhancement | Status: new
Priority: blocker | Milestone: sage-4.1.2
Component: build | Keywords:
Reviewer: | Author: David Kirkby, Minh Nguyen
Merged: |
-------------------------+--------------------------------------------------
Comment(by drkirkby):
Thank you for all your comments. In light of these, and some thoughts I
had, I've made a few changes.
* I removed the configure.ac.bak file, which was an oversight on my part.
* I left the directory ''autom4te.cache'', as that is something created
by ''autoreconf'' and is not simply a left-over from me typing
''configure''. However, I did remove the file ''autom4te.cache/requests''
as it says at the top of the file that it can be safely removed. There is
nothing to indicate to me that it is safe to remove the other files in the
directory ''autom4te.cache'', so I think they are best left.
* I removed Williams's name, but suggested sage-support was emailed, not
sage-devel, as the failure to find the files checked for (make, ar, perl,
m4, ranlib, tar and gcc) are really support issues, not developer issues.
I provided a link to the Google group sage-support.
* I added 'g++' to that list of files too, just in case someone has GCC
built for only compiling C, and not C++.
* I do not believe there should be any parentheses after AC_PROG_CPP see
[http://www.gnu.org/software/autoconf/manual/html_node/C-Compiler.html the
autoconf manual] Hence I left that unchanged.
* I changed the license as suggested, though I'm not convinced that was
really necessary, as the GPL 2 specifically says you can use a later
version.
* There were two places CYGWIN were checked. Both said it was not
supported, and both caused it to exit. I removed the second check, but
left the first, which will now exits unless SAGE_PORT is exported. If
someone really wants to build on cygwin, we should not stop them. The code
now has:
{{{
if [ "$SAGE_PORT" = "" ]; then
if [ `uname | sed -e 's/WIN.\+/WIN/'` = "CYGWIN" ]; then
echo "Unfortunately, building SAGE on Cygwin is not currently
supported,"
echo "though we are actively working on supporting it. If you
would like"
echo "to help with the porting effort, please post to"
echo ""
echo " http://groups.google.com/group/sage-windows";
echo ""
echo "Also, see http://trac.sagemath.org/sage_trac/ticket/6743";
echo "In the meantime, to run Sage on Windows, please use"
echo "a virtualization solution instead, such as VMware."
echo "To get past this message, export the variable 'SAGE_PORT'
to"
echo "something non-empty."
exit 1
elif [ `uname` = "SunOS" -a "`uname -p`" != "sparc" ]; then
}}}
* I changed one line of code
{{{
elif [ `uname` != "SunOS" -a `uname` != "Darwin" -a `uname` != "Linux" ];
then
}}}
to
{{{
elif [ `uname` != "SunOS" ] && [ `uname` != "Darwin" ] && [ `uname` !=
"Linux" ]; then
}}}
as I understand the former is not portable if there is more than one '-a'.
Since this code is run by /bin/sh, and not bash, we can't assume that more
that one '-a' will work.
The changes can be found on the web at:
http://sage.math.washington.edu/home/kirkby/Solaris-fixes/prereq-0.4-3rd-
try/
but if you download the files, the execute permissions on
prereq-0.4-install will be lost, so they will need to be added back
manually.
If you are happy with those changes, could you please check them in my
name. If you take the files from sage.math from the directory
{{{
/home/kirkby/Solaris-fixes/prereq-0.4-3rd-try/
}}}
then the permissions should be ok.
There is an issue which I noticed, which might want addressing later, but
I have left, as it probably needs some discussion.
If one wished to build Sage on a platform without GCC, it would fail due
to the checks for gcc and g++, '''even if one's aim was to use another
compiler'''. When I try to build Sage with Sun Studio, it would actually
be advantages to '''not''' have gcc in my path, as then I would spot the
bits of code that rely on gcc. Perhaps it might be sensible to not check
gcc and g++ if SAGE_PORT was defined to something non-empty. Anyway, for
now at least, having gcc and g++ is a requirement, even if you wished to
improve sage so it does not rely on gcc, but used better compilers when
available.
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/7021#comment:13>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/sage-trac?hl=en
-~----------~----~----~----~------~----~------~--~---
