Did you open a trac ticket to fix the sage-bdist script?
On Monday, September 1, 2014 2:17:57 PM UTC+1, Daniel Friedan wrote:
>
> I believe I've now succeeded in building a Mac OS X 10.6 .app version
> of Sage 6.3 (with help recorded in a separate thread).
>
> I had to fix a mistake in the file src/bin/sage-bdist (a mistake
> that affects the official OS X .app builds), and I also had to
> make a few other changes to that file. (Details at the end of
> this posting.)
>
> I'd be happy to upload
> sage-6.3-x86_64-Darwin-app.dmg
> if somebody will tell me how to do it.
>
> I'd also be happy to volunteer to produce future Mac OS X 10.6
> .app versions.
>
> I'm a bit unsure that I've done enough to validate the build:
> (1) 'make' finishes with no error message at the end
> (2) './sage -bdist' finishes with "** BUILD SUCCEEDED **"
> (3) the 'file' commands
> $ file Sage-6.3.app/Contents/MacOS/Sage
> $ file Sage-6.3.app/Contents/Resources/sage/local/bin/*
> $ file Sage-6.3.app/Contents/Resources/sage/local/lib/*
> show all architectures are 64-bit.
> (4) Sage-6.3.app runs under OS X 10.6.8
> (5) a sample workbook executes properly
>
> Is there more that one should do to validate the build?
>
> Is there any way to tell from inside Sage that it is running in
> 64-bit mode?
>
> The output from '$ ./sage -bdist' showed some strange looking command line
> options:
> -mmacosx-version-min=10.4
> and
> setenv MACOSX_DEPLOYMENT_TARGET 10.4
> and gave a warning:
> <built-in>:0: warning: Mac OS X version 10.5 or later is needed for
> use of the new objc abi
>
> Why do these command line options refer to '10.4'?
>
> ==============================================
> Procedure for building a Mac OS X 10.6 .app version of Sage 6.3
>
> (1) The file src/bin/sage-bdist contains a mistake:
> 'ARCHES' should be 'ARCHS'.
> The effect of the mistake is that the OS X application is
> always built as 32-bit ('i386'). This can be seen by running
> $ file Sage-6.3.app/Contents/MacOS/Sage
> in the OS X 10.9 .app version currently being distributed
> officially.
>
> (2) sage-bdist uses `uname -m` to determine the target
> architecture. Some of the older 64-bit Macintoshes can only boot into
> a 32-bit system. These machines run 64-bit Sage perfectly well.
> They can make and build 64-bit Sage perfectly well. The problem is that
> `uname -m` returns 'i386'. I added an environment variable to
> override `uname -m` (SAGE_APP_TARGET_ARCH==x86_64).
>
> I also added some code to sage-bdist to append '-app' to the name
> of the .dmg file produced.
>
> I also added an environment variable (SAGE_APP_GZ=no) to prevent
> the final compression stage, to save time during debugging
> sage-bdist.
>
> The procedure that worked for making and building the Mac OS X 10.6 .app
> version of Sage 6.3 was:
>
> $ git clone git://github.com/sagemath/sage.git
> $ cd sage
> $ export MAKE="make -j2"
> $ $MAKE
> $ cp ../sage-bdist-TARGET_ARCH-app src/bin/sage-bdist
> $ git remote add trac git://trac.sagemath.org/sage.git -t master
> $ git fetch trac u/iandrus/trac-16796
> $ git checkout FETCH_HEAD
> $ export SAGE_APP_TARGET_ARCH=x86_64
> $ export SAGE_APP_DMG=yes
> $ export SAGE_APP_BUNDLE=yes
> $ ./sage -bdist
>
> It produced a somewhat smaller version of Sage 6.3 than the
> officially distributed OS X 10.6 version. It looks like the size
> difference is due to git stuff. Starting with a fresh clone
> seems worthwhile (cloning is a small fraction of the 'make' time.)
>
> Daniel
>
>
--
You received this message because you are subscribed to the Google Groups
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.
