NeoOffice 2.0 (or the beginnings of work on it.)

[pluby]

Mox,

I have removed the --with-system-python argument from the neojava/makefile so that when you checkout the code, it will have a reasonably good chance of building all of the way through.

Even though not using the system Python makes that build bigger and longer, my main goal is to get the build past the OOo build so that I can continue work on the NeoOffice custom code and, more importantly, reboot back into Panther so that I can continue to work on Neo 1.2 bugs.

Patrick

[Mox]

[Mox]

[jjmckenzie51]

[Mox]

I have no reason to believe that the source tar.gz that is available next to the 2.0.1 binary releases is somehow different than OO 2_0_1 -tag. That would be insane.

Your continuous tries to use OOo cvs puts unnecessary strain on the CVS. And if you are paranoid, you can always do cvs -z3 co -r ... in the build directory where you unpacked the OO 2.0.1 from the tarfile (just be sure to cd into OOA680_m1 or move the files.). That way, only the (unlikely) changes get transferred across.

I think that the errors that you are seeing are either timeout errors because of slow connection or other errors because of slow OOo cvs. There isn't much you can do about them.

Mox

[jjmckenzie51]

[Mox]

[jjmckenzie51]

[jjmckenzie51]

Patrick/Mox:

I got as far as building berkeleydb and the build failed with the following error:

/bin/sh ./libtool --mode=link /sw/bin/ccache /usr/bin/gcc -avoid-version -rpath /usr/local/BerkeleyDB.4.2/lib -o libdb-4.2.la mut_tas.lo db185.lo bt_compare.lo bt_conv.lo bt_curadj.lo bt_cursor.lo bt_delete.lo bt_method.lo bt_open.lo bt_put.lo bt_rec.lo bt_reclaim.lo bt_recno.lo bt_rsearch.lo bt_search.lo bt_split.lo bt_stat.lo bt_upgrade.lo btree_auto.lo hash.lo hash_auto.lo hash_conv.lo hash_dup.lo hash_meta.lo hash_method.lo hash_open.lo hash_page.lo hash_rec.lo hash_reclaim.lo hash_stat.lo hash_upgrade.lo hash_verify.lo qam.lo qam_auto.lo qam_conv.lo qam_files.lo qam_method.lo qam_open.lo qam_rec.lo qam_stat.lo qam_upgrade.lo qam_verify.lo rep_method.lo rep_record.lo rep_region.lo rep_util.lo db_ovfl_vrfy.lo db_vrfy.lo db_vrfyutil.lo bt_verify.lo crypto_stub.lo crdel_auto.lo crdel_rec.lo db.lo db_am.lo db_auto.lo db_byteorder.lo db_cam.lo db_conv.lo db_dispatch.lo db_dup.lo db_err.lo db_getlong.lo db_idspace.lo db_iface.lo db_join.lo db_log2.lo db_meta.lo db_method.lo db_open.lo db_overflow.lo db_pr.lo db_rec.lo db_reclaim.lo db_rename.lo db_remove.lo db_ret.lo db_salloc.lo db_shash.lo db_truncate.lo db_upg.lo db_upg_opd.lo dbm.lo dbreg.lo dbreg_auto.lo dbreg_rec.lo dbreg_util.lo env_file.lo env_method.lo env_open.lo env_recover.lo env_region.lo fileops_auto.lo fop_basic.lo fop_rec.lo fop_util.lo hash_func.lo hmac.lo hsearch.lo lock.lo lock_deadlock.lo lock_method.lo lock_region.lo lock_stat.lo lock_util.lo log.lo log_archive.lo log_compare.lo log_get.lo log_method.lo log_put.lo mp_alloc.lo mp_bh.lo mp_fget.lo mp_fopen.lo mp_fput.lo mp_fset.lo mp_method.lo mp_region.lo mp_register.lo mp_stat.lo mp_sync.lo mp_trickle.lo mutex.lo os_abs.lo os_alloc.lo os_clock.lo os_config.lo os_dir.lo os_errno.lo os_fid.lo os_fsync.lo os_handle.lo os_id.lo os_map.lo os_method.lo os_oflags.lo os_open.lo os_region.lo os_rename.lo os_root.lo os_rpath.lo os_rw.lo os_seek.lo os_sleep.lo os_spin.lo os_stat.lo os_tmpdir.lo os_unlink.lo sha1.lo txn.lo txn_auto.lo txn_method.lo txn_rec.lo txn_recover.lo txn_region.lo txn_stat.lo txn_util.lo xa.lo xa_db.lo xa_map.lo
libtool: link: `mut_tas.lo' is not a valid libtool object
make[1]: *** [libdb-4.2.la] Error 1
dmake: Error code 2, while making './unxmacxp.pro/misc/build/so_built_so_berkeleydb'
'---* tg_merge.mk *---'

ERROR: Error 65280 occurred while making /Users/jamesmckenzie/NeoOfficeJ/NeoOffice2.0_test/neojava/build/berkeleydb
make: *** [build.oo_all] Error 1

I'm guessing that ccache did not finish building this object before libtool tried to check it. I am using the following line to invoke make:

~/NeoOfficeJ/NeoOffice2.0_test/neojava jamesmckenzie$ make GNUCP=/sw/bin/cp PKG_CONFIG=/sw/bin/pkg-config CC="/sw/bin/ccache /usr/bin/gcc" CXX="/sw/bin/ccache /usr/bin/g++" >& log_build0_gcc33_2005_1230_1019

The redirect sends both stdout and stderr to the logging file (tcsh will not allow separate logging.) I also had to enclose the $(CC) and $(CXX) variables in double quotes in order to use the makefile (CC="$(CC)").

Anyway I will rerun the build without using ccache and see what happens...

James

[pluby]

[jjmckenzie51]

[Mox]

[jjmckenzie51]

[Mox]

[jjmckenzie51]