Geoff,
Thanks for testing this.
I left a set -x on line 371 which causes the double output you're
seeing on -h. I've removed it.
Also, I've changes the svn
for plib to http://svn.code.sf.net/p/plib/code/trunk/.
OSG and PLIB are slated for different treatment in a future version.
If the chosen version of OSG, or the 1.8.5 version of PLIB is
installed or previously built, there's no reason to reconfigure
or rebuild them.
In general, I'd like to be able to detect when git does not add a file
to the sources. Then I could condition the configure step on whether
its actually needed or not. If I find a way to do that quickly, it
could be applied to any component that uses git, not just OSG.
I'll be using your approach for plib at least.
The latest is still 1.9.12:
https://www.gitorious.org/fg/fg-download-and-compile-fgmeta/raw/master:download_and_compile.sh
-Pat
On Thu,
10 Oct 2013 12:50:15 +0200 Geoff McLane ubu...@geoffair.info wrote:
Hi Pat,
Finally got a chance to try download_and_compile.sh
1.9.12 from the ...-fgmeta repo...
I think you need to adjust the svn repo for plib! It
has permanently changed I think
About ln 649
svn $PLIB_STABLE_REVISION_ co
http://plib.svn.sourceforge.net/svnroot/plib/trunk/ plib
to the new repo
svn $PLIB_STABLE_REVISION_ co
http://svn.code.sf.net/p/plib/code/trunk plib
Also, as a suggestion, why not first search the system for an
installed plib, and use that rather than a new download
and compile? These days I have add plib as a 'system'
install...
In my makefg, which as you know was based on one of
Brisa's original versions, circa Jan 2009, I use
script like -
# if NO plib installed path from user, and none from
# previous configuration file, then
if [ -z $INSTALL_DIR_PLIB ]; then
# try the 'system' places
PLIB_PATH=/usr
if [ -f $PLIB_PATH/include/plib/ul.h ]; then
show_plib_version $PLIB_PATH
PLIB_INSTALL_OK=1
elif [ -f $PLIB_PATH/local/include/plib/ul.h ]; then
show_plib_version $PLIB_PATH/local
PLIB_INSTALL_OK=1
else
PLIB_PATH=
fi
fi
The function 'show_plib_version' just reads the version
from the ul.h header...
Anyway, just a suggestion...
After the above svn correction, the script ran to
the end, building fgcom, fgfs, fgrun, openrti,
OpenSceneGraph, plib, simgear... great stuff...
I only had a chance to try it with the ALL option ;=))
But found out later that is not really 'ALL'. I could
add FGX, FGO?, FGCOMGUI, ATLAS and DATA, but I have
already done the DATA separately overnight...
I will try to get around to trying the others...
presently doing -p FGX, but it insists on rebuilding
OSG which is slow...
One little thing. When I say run it with the -h help
option for example my bash shell echos the echo
lines, like the script has say -
echo Switches:
--
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register
http://pubads.g.doubleclick.net/gampad/clk?id=60134071iu=/4140/ostg.clktrk
___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel