Technical notes on version 2.1
General:
-19Jun07
+28Jui07
+ebl Fix a director segfault during job setup #907
+kes Correct seg fault in postgresql when first operation returns no
+ rows. Fixes Dirks File browser problem seg fault with bat.
+25Jul07
+kes Apply patch from Felix Schwarz <Felix.Schwarz@web.de> that allows
+ building wxconsole on Fedora 7 with wxWidgets 2.8 (it works with
+ 2.6 as well).
+24Jul07
+kes Implement PopUp YES/NO dialog for bat.
+23Jul07
+ebl Fix a compilation bug when using #define BACL_WANT_NUMERIC_IDS
+22Jul07
+kes Correct RunScript to terminate job in FD on failure.
+kes Change abort_on_error and AbortJobOnError to fail_on_error
+ and FailJobOnError in RunScripts.
+kes Minor tweaks of code formating in RunScripts, principally to
+ avoid depassing 80 columns.
+21Jul07
+kes Eliminate one #ifdef Win32 in bsmtp
+kes Implement grow tool to grow a file for testing very large databases.
+kes Add big-vol-test to test Bacula seeking with > 5GB.
+Release 2.1.28 beta
+19Jul07
+kes Modify find_next_volume() to return all values in Media record.
+kes Ensure that we keep Recycle Pool when moving Scratch volumes.
+18Jul07
+kes Update Release notes
+17Jul07
+ebl fix a bug in get_scratch_volume() report by Andreas Helmcke
+ Warning, in previous beta version, all media get from scratch keep their
+ old retention.
+16Jul07
+kes Tweaked Win32 installer to remove temp conf files at uninstall time.
+ This should fix bug #842. Made a note to look at removing the temp
+ files at the end of the install.
+kes Add ability to mount and unmount a file device providing it is
+ marked as being removable and has the appropriate mount and unmount
+ commands defined in the Device resource. This fixes bug #830.
+kes When applying a storage override, release all previous storage
+ definitions rather than just prepending the new storage.
+ This should prevent a good deal of confusion.
+14Jul07
+kes Implement ./configure search for qwt libraries.
+13Jul07
+kes Disable posting the WM_CLOSE message in KillRunningCopy of the
+ Win32 FD. This should fix bug #893.
+kes Remove an unnecessary beep in Win32 init code.
+kes Update copyright dates in Win32 executable about boxes.
+kes Replace the FSF copyrighted getopt.c and getopt.h in the Win32
+ code with a BSD getopt.c and getopt.h.
+kes Remove sched.h from the Win32 compat directory -- not used.
+kes Remove the Exit menu item from the Win32 tray monitor. If you really
+ need to shut it down, do it through the services panel or the command
+ line.
+kes Remove some inappropriate Bacula FSFE copyrights.
+ebl Tweak status client=xxx to be more easy to parse (more like in 2.0.x)
+kes Fix regress config non-portable test as pointed out by
+ Florian Heigl <florian.heigl@gmail.com>
+kes Fix PSCMD for HP-UX in configure.in to have proper syntax.
+ as pointed out by Florian Heigl <florian.heigl@gmail.com>
+12Jul07
+kes Fix missing rctx. prefixes on the Darwin code in src/filed/restore.c
+ Reported by Frank Sweetser.
+
+Release 2.1.26 beta
+12Jul07
+kes Fix client-only build.
+kes Integrate client-only fixes for bsys (initgroups) and conio from
+ Sergey Svishchev <svs@ropnet.ru> for the AIX platform.
+kes Make arguments to get_scratch_volume() in same order as other
+ such calls.
+kes Rework the prune_volumes() subroutine so that after purging and
+ no volume found in desired pool, it will check for a scratch
+ volume. This prevents recycling the whole scratch pool.
+kes Add code to status storage to print boffset_t (largefile support).
+11Jul07
+kes Move qwt (Graphics libarary for Qt4) to depkgs.
+kes You now need a --with-qwt=<dir> to be able to build bat.
+kes Update src/win32/cats/bacula_cats.def to include the new
+ db_get_query_dbids() call on Win32.
+07Jul07
+kes Correct a problem when selecting a scratch volume and moving
+ it to another pool that lost some columns.
+kes Tweak some code in ua_update.c to use POOL_MEM instead of POOLMEM.
+kes Enhance Autochanger error messages to include Volume and device name.
+kes Rework prune_volumes() code to take account of InChanger flag,
+ and to handle recycling volumes going to the Scratch pool and
+ current pool because the RecyclePool directive.
+kes Implement a better and more efficient db_get_query_dbids() to
+ handle creating and passing back a list of DBIds.
+
+Release 2.1.24 beta
+03Jul07
+kes Start work on new more efficient DBId subroutine. First use
+ will be for recycling volume to Scratch inchanger.
+kes Increase number of JobIds in pruning from 1000 to 10000.
+ This to be replaced by above routine.
+kes Begin implementation of building Qt4 on Win32.
+kes Correct typo in fix I added for bad TLS shutdown.
+kes Pull 2.0.3 patches into patches directory.
+kes Update Release notes. Include qt-console in line count.
+kes Update Projects file.
+30Jun07
+kes Integrate patch from Sergey Svishchev <svs@ropnet.ru> that fixes
+ bug in migration code where a job that spanned two volumes
+ was migrated twice.
+29Jun07
+kes Implement new BST_DESPOOLING blocked state. Change from locking
+ during despooling in SD to blocking. This means that other threads
+ can work with the device structure, in particular the reservations
+ system while despooling.
+28Jun07
+kes Fix return in reservation message queue that missed clearing
+ the jcr lock (implemented 26Jun07 below).
+kes Rename a number of dev methods to make locking function names
+ a bit clearer.
+kes Document locking in lock.c. Move lock structures to new file
+ lock.h.
+26Jun07
+kes Move reservations message lock to lock jcr only this
+ fixes bug #861.
+kes Move main SD locking code into lock.c (new file).
+kes Update Win32 build to include lock.c
+
+Release 2.1.22 beta
+26Jun07
+kes Dirk committed the qwt library code for drawing graphs in bat.
+kes Tweak install of qwt so that it is integrated with ./configure.
+25Jun07
+kes Fix HAVE_BATCH_INSERT to be HAVE_BATCH_FILE_INSERT, pointed
+ out by Eric (my typo).
+ebl Fix configure.in to get batch mode enable
+kes Return correct string from authenticate.cpp in bat when connecting
+ to Dir.
+kes Apply patch suggested by Frank Sweetser to fix bug #888 --
+ spurious line drops when using TLS.
+kes Do not file reparse points when restoring a file that already
+ exists -- Win32.
+24Jun07
+kes Implement Windows reparse points -- similar to directories, but
+ we do not descend into it. This is a first cut. They seem to
+ be backed up, but restore is not yet tested.
+kes Remove restore_blocking in tls code when shutting a socket to leave
+ it in blocking mode. Hopefully this will fix the encryption bug
+ reported by Frank Sweetser.
+kes When opening a file for backup, tell the OS that we are going to
+ read it sequentially -- optimization.
+kes Change variable names in authenticate_director() to not conflict
+ with member names in bat. Caused great confusion with compiler,
+ but no warnings.
+23Jun07
+kes Queue openssl error messages in case of comm problem.
+kes Do shutdown() call if socket terminated.
+kes Remove some duplicated code in tls.c
+kes Move more of the bsock member variables into private and access
+ them with methods.
+kes Add Frank Sweetser's tls-test to the regression suite.
+22Jun07
+kes Add Swedish sv.po file
+21Jun07
+kes Don't print clock skew message in FD if less than 3 seconds diff.
+kes Add a bit of VSS info to status client.
+kes Implement a first cut of Vista VSS, using Win2003 code.
+
+Release: 2.1.20 beta
+20Jun07
+kes Fixed bug #886 (multidrive autochanger: SD doesn't use drive with
+ loaded tape but uses first drive).
+kes Fixed bug #807 Restore encrypted backup failed on Windows 2003
+ This was indeed a Win32 issue.
+19Jun07
+kes Simplify search of autochanger for in-use Volume.
+kes Fix exit condition of check for in-use volume. kes Remove
+ inappropriate my_postgresql_free_result() in db_close()
+ routine.
kes Fix bad return in FD when encryption signature not found.
Probably caused a crash.
kes Inhibit missing encryption signature error messages for
that the URLs don't change and so that the developers will all
work off the same copy. Rebuild from scrach.
kes Upgrade Win32 SQLite3 from 3.3.8 to 3.3.17.
+
Release: 2.1.18 beta
16Jun07
kes Fix seg fault in FD from incorrect digest size.