#8263: Document ALL environment variables used by Sage
-----------------------------+----------------------------------------------
Reporter: drkirkby | Owner: drkirkby
Type: defect | Status: needs_work
Priority: major | Milestone: sage-4.5
Component: documentation | Keywords:
Author: | Upstream: N/A
Reviewer: | Merged:
Work_issues: |
-----------------------------+----------------------------------------------
Comment(by jhpalmieri):
Here's a new patch which deals with some of these comments. This is still
"needs work"; for example, I don't suggest we keep the text surrounding
"SAGE_FAT_BINARY".
Replying to [comment:13 drkirkby]:
> * I would say that "SAGE_PARALLEL_SPKG_BUILD" is experimental
Okay.
> * I think to say setting CC and CXX '''may''' be unreliable is an
understatement.
Okay.
> * CFLAGS/CXXFLAGS - I'd make the point that these do not work for all
packages, so are of limited use and setting them may cause build problems.
Okay.
> * The meaning of the SAGE64 variable is not correctly described.
Sorry, I got this straight out of the sage-env script.
SAGE64 does not work on only on Solaris. The original aim of SAGE64 was to
force a 64-bit build on '''OS X''' systems where
> Some versions of OS X default to 64-bit binaries
This is either OS X 10.6, or maybe just 64-bit machines on OS X 10.6.
Must be the latter. But we should check with William or sage-devel.
> * SAGE_USE_OLD_GCC. I would add the fact Sage will not build with
versions of gcc older than 4.0.1, so this is only of use if you intend
changing the Sage source code to allow it to build with older versions of
GCC. I think we need to stop people wasting their time thinking it might
work, because it most certainly will not.
I've changed this part a bit, but maybe it should be more strongly worded.
> * CFLAG64. I would add that it is unnecessary to set this on any
mainstream operating system, and the only use would be if attempting to
port Sage to a platform like AIX or HP-UX using a non-GNU compiler.
Okay.
> * INCLUDE_MPFR_PATCH. There is a space missing between at this point
"sun4v machines.If"
Okay
> * SAGE_MATPLOTLIB_GUI. There is a 't' missing from the word 'attempt'.
Okay
> * SAGE_PORT. I would add that any attempt to build Sage with a compiler
other than GCC will need this set, and furthermore than Sage has never
been successfully built with any compiler other than GCC. (This variable
is sort of self-documenting, as it explains exactly what to do, so perhaps
we don't need to spell it out too much.)
Okay.
> * SAGE_TIMEOUT & SAGE_TIMEOUT_LONG are not described.
They deal with doctesting, not the build process, so I'm not sure they
belong here. I've added them now.
> * SAGE_CHECK. I would add that only a small subset of Sage packages
support this, but this is being expanded. A reasonably upto date list of
packages supporting this may be found at
http://trac.sagemath.org/sage_trac/ticket/9281
I don't know if we should mention this or just work hard at #9281. This
particular piece of documentation doesn't seem to get updated very much --
see the statements at the top of the page I'm editing about Sage not
working on Solaris, for example -- so I don't want to post information
which is too time-sensitive.
> * Should these variables be listed alphabetically? In categories? At
the moment, the order seems a bit random. I would think put them in
categories, but then sort them alphabetically in there.
I've put them in categories, and then within each category I've started
the ones that seem most useful.
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/8263#comment:14>
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.
