gdal needs to rebuild for openssl soname bump. I tried disabling the
grass-devel build requires, but that doesn't seem to have worked. Not only that
gdal tried to bring gdal-devel in (perhaps for grass-devel?)
Can you please work on this to get a build of gdal (without grass) built, so
that we can build grass against it?
Update. I forcefully disabled grass in the spec and now the build gets to
start, but fails due to a compile error:
g++ -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector
--param=ssp-buffer-size=4 -m32 -I/builddir/build/BUILD/gdal-1.4.2-fedora/port
-I/builddir/build/BUILD/gdal-1.4.2-fedora/ogr -Ihdf-eos -DOGR_ENABLED
-I/usr/include/ogdi -I/usr/include/netcdf-3 -I/usr/include/hdf
-I/usr/include/cfitsio -I/usr/include/cfitsio/include -c hdf4dataset.cpp -fPIC
-DPIC -o ../o/.libs/hdf4dataset.o
hdf4dataset.cpp: In member function 'CPLErr
hdf4dataset.cpp:576: error: 'MAX_NC_NAME' was not declared in this scope
hdf4dataset.cpp:589: error: 'szAttrName' was not declared in this scope
hdf4dataset.cpp: In static member function 'static GDALDataset*
hdf4dataset.cpp:746: error: 'MAX_VAR_DIMS' was not declared in this scope
hdf4dataset.cpp:831: error: 'aiDimSizes' was not declared in this scope
hdf4dataset.cpp:931: error: 'aiDimSizes' was not declared in this scope
hdf4dataset.cpp:982: error: 'aiDimSizes' was not declared in this scope
hdf4dataset.cpp:1004: error: 'aiDimSizes' was not declared in this scope
hdf4dataset.cpp:1033: error: 'aiDimSizes' was not declared in this scope
hdf4dataset.cpp:1040: error: 'aiDimSizes' was not declared in this scope
make: *** [../o/hdf4dataset.o] Error 1
make: Leaving directory `/builddir/build/BUILD/gdal-1.4.2-fedora/frmts/hdf4'
make: *** [hdf4-install-obj] Error 2
make: Leaving directory `/builddir/build/BUILD/gdal-1.4.2-fedora/frmts'
make: *** [frmts-target] Error 2
error: Bad exit status from /var/tmp/rpm-tmp.16196 (%build)
Ping? Can the maintainer please respond? This has been broken in rawhide for
almost 2 weeks and it blocking the rebuild of several other packages including
grass, mapserver and others.
Hmm, a clue seems to be that since gdal was last rebuilt, the hdf package which
is a BuildRequires, has undergone a version upgrade:
and from the changelog it looks like it has done some odd business with the netcdf:
* Mon Oct 29 2007 Patrice Dumas <firstname.lastname@example.org> 4.2r2-4
- install the netcdf.h file that describes the netcdf2 hdf enabled
* Mon Oct 29 2007 Patrice Dumas <email@example.com> 4.2r2-3
- ship hdf enabled nc* utils as hnc*
- add --disable-netcdf that replaces HAVE_NETCDF
- keep include files timestamps, and have the same accross arches
- fix multiarch difference in include files (#341491)
* Wed Oct 17 2007 Patrice Dumas <firstname.lastname@example.org> 4.2r2-2
- update to 4.2r2
and I notice on the gdal wiki it mentions some compatibility issues with netcdf:
Investigating this it appears that some changes in hdf causes this (see bug #189337)
Bumping priority and severity as this continues to block 4 or 5 other packages
and causes many broken deps in rawhide.
As the reviewer of gdal, I tried to check and with some fixes
However, it seems that Balint does not allow cvsgroup access...
Created attachment 290803 [details]
Patch against devel/gdal.spec
This is the patch against the spec file extracted from the SRPM mentioned in
comment #5 above. Please apply ASAP.
(In reply to comment #5)
> As the reviewer of gdal, I tried to check and with some fixes
> gdal builds.
> However, it seems that Balint does not allow cvsgroup access...
Hello Tasaka, Jesse
Would you like please to grant for my Balint Cristian <email@example.com> aceess:
My account looks like this (i pass CLA and so on):
cvsextras rezso user unapproved None
fedorabugs rezso user unapproved Unneeded
I guess i need new sponsorshisp since i cannot access firstname.lastname@example.org
anymore, so i have to use non RH account for future.
Package Change Request
Package Name: gdal
Owner needs to be a fas account name not an email address.
Package Change Request
Package Name: gdal
This is now fixed, so closing bug. But the new rebuild introduced some bogus
Perl deps which cause more broken deps (I've opened up a separate bug: bug #428541)