1 Technical notes on version 1.39
6 - Fix barcode test, eliminate BOOL (doesn't exist on Linux).
7 - Write first cut of README.mingw32
8 - Add back ua_label barcode fix from Rufolf Cejka.
9 - Integrate the majority of the Mingw cross-tools submission
10 made by Howard Thomson. Everything compiles, with zlib, openssl,
11 and VSS turned off. Linking needs a pthreads library ...
13 - Add 'Dir Status' button to the gnome console.
14 - More work on migration.
15 - Restore non-compatible base64 coding to permit compatibility.
16 - Display heap stats in Storage daemon without debug level.
18 - Integrate most of the MinGW changes -- more to do.
20 - Implement using pg_config for finding PostgreSQL files.
21 Fixes bug #600. Patch supplied by user.
22 - Remove -t option from mktemp in mtx_changer.in and use
23 working directory. Fixes bug #578.
24 - Update job start time after the any run before job so that
25 files created by the script are only backed up once. Fixes
27 - Strip trailing newline only from filenames entered in
28 the restore command when reading a file. This permits
29 the user to enter filenames with trailing spaces. Fixes
30 bug #549. The user supplied a patch that I modified slightly.
31 - Use the most recent time (st_mtime, st_ctime) in the dir
32 command in restore. This gives the user a better idea of what
33 the newest file really is. This fixes bug #574. The fix
34 was suggested by the user.
35 - Implement a compatible version of base64. This permits external
36 programs to duplicate Bacula's base64 algorithm using standard
37 routines. This fixes bugs #296, and 565. Patch submitted by
39 ================= Note ========================
40 Previous Signatures stored in the database are no longer
41 compatible with this. The main downside is for Verify jobs,
42 and doing an InitCatalog run will fix the problem. Also, the
43 authentication between the deamons is changed, so all daemons
44 must be simultaneously upgraded.
45 ==============================================
46 If you don't like this fix, set:
47 const bool compatible = true;
49 const bool compatible = false;
52 - Ensure that DB signature is never NULL.
53 - Ensure that DB table names are not translated.
54 - Fix scheduler to handle time skew (eg daylight savings).
55 - Fix scheduler to use lock_jobs() to avoid most problems
56 with reload. Window is now milliseconds.
58 - Apply patch from Christopher Hull
59 - Allow multiple connections to database with different
61 - Invalidate the scheduler when doing a reload. Fixes seg
62 fault, but still 60 second window.
63 - Additional info in Reschedule message.
64 - Use set_jcr_job_status() everywhere to prevent loss of
66 - Display peer IP in FD if error from connecting DIR.
67 - Don't increment file count for DIRBEGIN.
68 - Replace illegal characters in Win32 filename by _.
69 - Add SE_CREATE_PERMANENT_NAME privilege in Win32.
70 - Hash hard link filenames rather than linked list.
71 - Fix for security failure in chdir on Win32.
72 - Add CreateDirectoryA/W win32 API entry points.
73 - Add /silent option to Win32 FD for Install/Remove service.
74 - Always print Heap statistics in FD.
76 - Remove the -f option from the chown in Makefile.in for more
78 - Change setting the group in Makefile.in to use chgrp for
80 - Implement a write_store and read_store to replace
82 - Implement a Bacula read/write lock for Python rather
83 than using the Python lock to avoid recursive problems.
84 - Correct the uninstall directory names in filed/Makefile.in
85 as reported by a user.
87 - Correct some problems with database creation (new tables).
88 - Replace a bunch of old 0x%x by %p.
89 - Get first cut of Migration with Job selection working.
91 - Change store_bit() to store_bool() for the Spooling in
92 dird_conf.c -- bug reported by Robert Nelson.
94 - Correct Makefile bug found by Dan Langille.
96 - Correct Makefile for Solaris /bin/sh
97 - Correct mtx-changer.in for Solaris /bin/sh
98 - Abort if a conf resource does not have a Name =
100 - Change the name of the regex program to bregex.
101 - Add the bwild program to the tools directory. It is similar
102 to the bregex program.
103 - Implement create bregex.h and bregex.c in src/lib from the
104 Python regexp program.
105 - Use the new bregex.c to implement Regex expressions on Win32.
107 - More work on migration.
108 - Implement wild program in tools directory for testing
109 wild-cards. Almost identical to the regex program.
110 - Up port VSS single thread locking code.
111 - Replace N_(..) by NT_(...) to inhibit translation of
113 - Modify Makefile to change the permissions on Working Directory
114 to 770 if the directory is created.
115 - Do not fail the Makefile if changing the permissions or
116 owner/group on WorkingDir fails.
117 - Remove early selection of Scratch Volumes in the recycling
119 - Correct the old recycling algorithm so that Scratch Volumes
120 are selected when looking for a Volume in the changer.
121 - Correct a typo in the Verify SQL reported by Joe Park.
123 - Remove automatic case folding on Windows FDs. You must
124 explictly use the 'Ignore Case = yes' option.
125 - Remove the code added to 1.38.6 and 1.38.7 that pulls a
126 scratch volume in an Autochanger early in the 'recycling'
129 - Make weof() and clrerror() methods of class DEVICE.
131 - Tweak license to include Microsoft restrictions.
132 - Move mysql.reconnect to after real_connect(). Thanks to
133 Frank Sweetser for the patch.
134 - Disallow a backslash in a File = directive (Windows junk)
135 unless the string is quoted.
136 - Apply Eric's patch to ua_label.c so that daemon protocol
138 - Add NT_ definition for strings that should not be translated.
139 - Apply NT_() to ua_label.c
141 - Remove timed wait for VSS on Win2K3 as it is not yet
143 - Correct bacula.in script to reference bacula-ctl-xx in the
144 sysconfig directory rather than the bin directory.
146 - Fix a long standing bug in the bacula start/stop scripts.
147 - Attempt to add the new bacula-ctl-* files to the rpm.
148 - Switch to using typeof() for list traversing to avoid
149 as many FC5 g++ compiler warnings as possible.
151 - Split the bacula start/start script into four files:
152 bacula -- starts and stops calling other scripts
153 bacula-ctl-dir -- starts/stops the director
154 bacula-ctl-fd -- starts/stops the File daemon
155 bacula-ctl-sd -- starts/stops the Storage daemon
157 - Create datestyle fix for PostgreSQL. Fixes bug #574.
158 - Correct editing of JobId from int to int64 in fd_cmds.c
159 - Eliminate FileSet name race with bash_spaces() and multiple
160 threads by bashing in a local.
161 - Fix error return from 'use storage' to print a correct error
162 message rather than nothing.
163 - Correct false re-read last block error message when two jobs
164 are simultaneously writing at the end of a tape.
165 - Simplify exit conditions in the reserve.c code to avoid
166 possible non-release of reservation_lock().
167 - Suffle lock order in reserve to avoid deadlock between
168 reservation lock and device mutex.
170 - Initialize jcr mutex before first use. Thanks to Thorsten for
171 tracking this down for me !!!! as it broke the Win32 build.
173 - Integrate addition of line count limitation to bsmtp -l from
174 Sebastian Stark <stark at tuebingen.mpg.de>
176 - Implement regex test program in tools directory.
177 - Attempt to fix time problem with bsmtp with foreign langs.
178 - Add strip_trailing_newline() submitted by user.
179 - Implement regex matching in migrate.c
181 - Fix bug #537 to allow arbitrary time to mount a volume for
182 restore, if polling is turned on.
183 - If dir_user or dir-group is specified in ./configure apply it to
184 the working-dir. Fixes bug #533.
185 - If rescheduling a job cancel the previous incarnation with the SD.
186 Fixes bugs #566 and 557.
187 - Fix bug #567 do_message() definition type conflict.
190 - Add more jcr methods and make mutex and use_count private.
191 - Create lock/unlock methods for jcr.
192 - Fix PostgreSQL bug doing sql_data_seeek() by explicitly reading
193 records to get to seek position.
194 - Integrate patch from bug #561 to correct conio.c signal definitions.
195 - Fix Rescheduling failed Jobs. Ensure that SD message thread
196 terminates correctly by doing pthread_kill(). Do not destroy
197 SD cond wait variable between executions of the job. Use local
198 mutex for cond variable to avoid blocking jcr chain. Fix poor
199 use of jcr use count in jobq.c for restarted jobs.
200 - Fix obsolete usage of foreach_dlist() to use foreach_jcr() in
201 lib/jcr.c -- prevents locking the jcr chaing.
202 - Apply patch from bug #564, which corrects listing volumes with
203 multiple autochangers. Apply same fix to next volume list.
204 - Fix bug #562 where restore bootstrap file is not unique.
205 - More details in implementing David's migration syntax proposal.
206 - Save and restore dcr when swapping drives so that if we have
207 a write dcr it is not lost.
208 - Use new routine lock_reservations() to lock the reservations
209 system, and call it while looking for a volume in askdir.c.
210 This could possibly fix bug #543.
211 - Stop SD command loop if job is canceled.
214 - Remove old code from findlib/create_file.c
215 - Rename mac.c migrate.c
216 - Add user friendly display of VolBytes in job report.
217 - Rename target... to previous... to make it a bit easier to
219 - Add selection type and selection pattern to Migration (idea
220 given by David Boyes).
223 - The po files should now be current.
224 - Fix new sql_use_result() code to properly release the
225 buffers in all cases.
226 - Use the keyword ujobid to mean the unique job id; job or jobname
227 to mean the Job name given on the Name directive, and jobid to
228 be the numeric (non-unique) job id.
229 - Allow listing by any of the above.
230 - Add the user friendly job report code for reporting job elapsed time
231 and rates with suffexes from John Kodis <kodis at comcast.net>.
232 - Add Priority and JobLevel as Python settable items.
233 - Use TEMPORARY table creation where the table is created by
235 - Add new code submitted by Eric for waiting on specific jobid.
236 - Add ACL checking for the dot commands.
237 - Fix restore of writable FIFOs.
238 - Fix a bug in bpipe where the string was freed too early.
240 - Modify the Python class examples to inherit object -- new way
241 of defining classes. Patch from Felix Schwarz.
242 - Implement jobuid to replace old usage of job in keywords as
243 suggested by Eric Bollengier.
244 - Apply patch for enhancing wait from Eric Bollengier. On can now:
245 wait (wait for all jobs to stop)
247 wait jobuid=unique id
249 - Implement write variables for Python to set Priority (anytime), and
250 Job Level, only during JobInit event.
252 - Fix the block listing bug pointed out by Arno.
253 - Update the po files at Eric's request.
255 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
256 accept time qualifiers.
257 - Make catalog respect ACL.
258 - Add recycle count to Media record.
260 Changes to 1.39.6-beta1 24Feb04
262 - Add Rudolf Cejka's new rc-chio-changer.
263 - Implement pulling Volume from Scratch Pool if the
264 Volume is in the autochanger.
265 - Implement additional command arguments for update Volume.
269 - Back port changes to 1.38.5
270 - Fix recycle SQL for StorageId.
271 - Fix misplaced quote in mtx-changer script.
273 - Note. Your database must be updated, or you must create
274 a new database. I have not yet incremented the database level.
275 - Add Location table.
276 - Add LocationId, DeviceId, and MediaTypeId to Media record.
278 - Implement create/get mac record in database for adding extended
279 Migration data to the job record.
280 - Add new MAC table to update/make database scripts.
281 - Return Storage name used when getting VolumeNames for a job.
282 - Change bsr file keyword Storage to Device, which is more accurate.
283 - Ensure that Mac records are pruned/purged.
284 - Tweak SD tools to deal with changing media type.
285 - Integrate more dev.c subroutines as methods (e.g. strerror, bsr, ...)
286 - Pass pointer to dcr pointer to acquire_device_for_read() so
287 that the subroutine can switch devices, and hence dcrs.
288 - Modify the multiple MediaType read code to re-use the same
289 dcr when switching devices. This makes the code much more
291 - Integrate patch from Karl Hakimian that reads JobIds, FileIndexes
292 from a table for restore.
293 - Add Storage name to VolParams, but it really should be Device.
295 - Add disk-changer to scripts directory + configure/Makefile
296 - Eliminate PoolId from jcr -- it is in jcr->jr.PoolId
297 - Implement store_bit scanner to replace store_yesno. Mostly done.
298 - Implement new store_bool that stores in a bool.
299 - Add true/false to yes/no conf directives.
300 - Make first cut at changing appropriate store_yesno to store_bool.
301 - Complete implementation of Pool storage devices.
302 - Move starting clones to job.c
303 - Move create_restore_bootstrap_file() to job.c
304 - Make copy_storage() more general to be able to handle
306 - Cleanup a lot of migration code for manual running, including
308 - Move getting a scratch Volume into a subroutine.
309 - Make all places a Volume that is added to a pool to
311 - Fix bug in autochanger recycle code (improper edit 64 bit).
312 - Fix segfault in restore command when no value specified.
313 - Start adding code to handle multiple MediaTypes in restore.
314 - Eliminate race condition in getting Volume name for
315 two drive autochanger.
316 - More debug code in autochanger.
317 - Add storage keyword to bootstrap file, add parsing.
318 - Move slot in bsr file into Volume record as there will be
319 a different slot for each Volume.
322 - Implement Pool storage overrides.
324 - Implement first cut of Migration.
325 - Implement mysql_use_result() from patch by Karl Hakimian.
326 This reduces significantly the memory consumption during
327 the restore tree building, and hence runs faster too.
328 - Implement StorageId patch supplied by user (reggie) in bug
329 #536. This should permit Bacula to work correctly with two
331 - Implement Job listing variations suggested by a user.
332 - Move updating bootstrap code in backup.c to subroutine
333 update_bootstrap_file().
334 - Add new job status elapsed time and bytes written user
335 friendly job report output patch sent by John Kodis
336 <kodis at comcast.ne>.
337 - Implement a storage list in Pools.
338 - Separate out setup_job() code from run_job().
339 - Get migration working -- lots of changes in mac.c in both
341 - Apply patch from user (Eric Bollinger I think) that fixes a
342 DIR crash when no arguments are supplied to a dot command.
343 - Fix typo (strcpy->strcmp) in ua_prune.c as reported by Martin.
344 - Fix command arg name->volume in label command of gnome-console.
345 - Fix SD acquire.c to release correct DCR by explicitly testing
346 on the dcr address rather than trying to devine if it is a read
347 or write dcr. This failed in error conditions when the device was
351 - Apply user supplied patch for more readable rate output
353 - Continue implementing migration.
354 - Implement support for removable filesystems in SD.
355 - Ensure that btraceback scripts can be read by anyone.
356 - Replace dvd-freespace and dvd-writepart by dvd-handler.
357 - Correct bug where canceling restore before the FD contacts
358 the SD causes the drive to be left in read mode.
359 - Move ofline_or_rewind into DEVICE::close().
360 - Eliminate close_device.
361 - Convert several dev subroutines to methods (e.g. bsf,
363 - Eliminate force_close_device().
364 - Implement Device Type directive in Device resource that
365 can have values File, Tape, Fifo, DVD, or Prog.
366 - Add has_cap() method to Device.
369 - Add patch from bug #527 to allow RedHat user to specify
370 options/user/group for starting each daemon in
373 - Add two new queries to query.sql provided by Arno. One
374 list volumes known to the Storage device, and the other
375 lists volumes possibly needing replacement (error, ...).
377 - Add periodic (every 24 hours) garbage collection of memory
378 pool by releasing free buffers.
380 - Correct bug counting sized (for display only) in smartall.c
381 - Print FD mempool stats if debug > 0 rather than 5.
383 - Make db_lock() mutex error fail the job rather than abort
384 Bacula. Canceling the job caused the mutex to fail.
385 - Correct bug in alist.c that re-allocated the list if the
386 number of items goes to zero.
387 - Move the reservation system thread locking to the top level
388 so that one job at a time tries all possible drives before
390 - Implement a reservation 'fail' message queue that is built
391 and destroyed on each pass through the reservation system.
392 These messages are displayed in a 'Jobs waiting to reserve
393 a drive' list during a 'status storage='. Note, multiple
394 messages will generally print for each JobId because they
395 represent the different problems with either the same drive
396 or different drives. If this output proves too confusing
397 of voluminous, I will display it only when debug level 1
398 or greater is enabled in the SD.
400 - Add enable/disable job=<job-name>. This command prevents
401 the specified job from being scheduled. Even when disabled,
402 the job can be manually started from the console.
403 - During 'update slots' clear all InChanger flags where the
404 StorageId is zero (old Media records).
408 - Fix autochanger code to strip leading spaces from returned
409 slots number. Remove bc from chio-changer.
410 - Back port a bit of 1.39 crypto code to reduce diffs.
411 - Fix first call to autochanger that missed close()ing the
412 drive. Put close() just before each run_program(). Fixes
415 - Add PoolId to Job record when updating it at job start time.
417 - Pull in more code from 1.39 so that there are fewer file
418 differences (the new ua_dotcmds.c, base64.h, crypto.h
419 hmac.c jcr.c (dird and lib) lib.h md5.h parse_conf.c
420 util.c. Aside from ua_dotcmds.c these are mostly crypto
422 - Implement new method of walking the jcr chain. The
423 incr/dec of the use_count is done within the walking
424 routines. This should prevent a jcr from being freed
425 from under the walk routines.
430 - Start implementing Verify list output.
431 - Move the suitable_drive flag to a better place to prevent
432 premature termination of the reservation if all drives
433 are busy -- should fix Arno's diff/inc pool failures.
435 - Add mutex to single thread VSS code in Win32.
437 - Simplify code in askdir.c that waits for creating an appendable
438 volume so that it can handle multiple returns from the wait
440 - Modify the wait code to permit multiple returns.
441 - Return a zero when "autochanger drives" is called and
442 it is not an autochanger.
443 - Make rewind_dev() a method taking a DCR as an argument.
444 This permits closing and reopening the drive if the
445 rewind fails as happens if the drive was loaded while the
446 file descriptor was open. This refreshes the file descriptor.
447 - Remove the ST_OPENED flag and always rely on fd < 0 for knowing
448 if the device is open or not. This should eliminate
450 - Return error if reserve cannot find at least one suitable device.
451 - Make wait_for_sysop() return correct state information.
452 - Fix Win32 state file problem. write was not using compat
453 code. This should fix bug #500.
455 - Modify gui on command to set only GUI mode and not batch.
456 - Modify .messages command to always print messages regardless
458 - If GUI mode is on, suppress automatic printing of
460 - Delete old bnet packet code.
461 - Ignore new BNET_START_SELECT and BNET_END_SELECT signals in
463 - Modify restore command in wx-console to set gui on and to use
464 only .messages instead of messages. Hopefully this fixes bug
466 - Fix seg fault in exit of acquire when canceling a job --
467 reported by Wolfgang Denk
468 - Pull in latest reservation system changes from 1.38
469 - Make .messages command always print messages regardless
470 of the automessages flag.
472 - Fix seg fault if user labels a drive directory bug #513
473 - Remove quotes around Version as it breaks things.
475 - Merge in Aleksandar Milivojevic's mods to the spec file.
476 - Apply sparse code fix for raw drives and fifos. Bug 506
477 - Thorsten fixed Unicode cd problem with wx-console bug 505.
479 - Correct reservation system to do a last ditch try
480 for any mounted volume, then anyone anywhere.
481 - Add quotes around table Version because of
482 error in MySQL 4.1.15 -- bug report submitted.
483 - Correct some minor problems with btape in the fill
485 - Updates to ssh-tunnel from Joshua Kugler.
486 - Added a report.pl program from Jonas Bjorklund.
487 - Simplify the O_NONBLOCK open() code for tape drives,
488 and always open nonblocking.
489 - Do not wait for open() if EIO returned (shouldn't happen).
490 - Eliminate 3 argument to tape open().
491 - Correct the slot # edited in the 3995 Bad autochanger unload
493 - With -S on bscan (show progress) do not divide by zero.
495 - Make cancel pthread_cond_signal() pthread_cond_broadcast().
496 - When dcr is freed, also broadcast dev->wait_next_vol signal.
497 - Remove unused code in wait_for_device.
498 - Make wait_for_device() always return after 60 seconds of wait.
502 - Add stubs for non-crypto build.
504 - Use localhost if no network configured
506 - Eliminated duplicate MaxVolBytes in cat update -- bug 509.
507 - Remove debug print.
508 - Add bail_out in error during state file reading.
510 - Merge changes made to 1.38.3 into HEAD
511 - Add stubs for pygtk-console code
512 - Create Makefile.in for pygtk-console code
514 - Merge updates into 1.38 branch
515 - Update specs to include mysql4 define.
516 - Fix when attributes are sent, must be after binit().
517 - Stop read_record() if status not ok in second loop.
518 - Return rec->FileIndex in dcr->VolLastIndex for normal
519 and partial records in read_record(). This allows bscan
520 to get FileIndex at EOT correct.
521 - Fix butil.c to correctly set dcr -- fixes seg fault in bls.
523 - Fix Win32 built to work with new crypto code.
524 - Apply patch supplied by user (slightly modified) to fix
525 correct detection of holes in block devices and FIFOs.
527 - Apply patch supplied by user (slightly modified)
528 to fix SD hang with multiple pools and bad client
531 - Add nagios plugin to the examples directory. Submitted by
533 - Remove warning message about multiple saves of hardlinked files
534 from find_one.c as it can generate too many warning messages.
535 - Modify most restore error messages to be queued so that they
536 appear at the end of the job rather than mixted with the restore
537 listing where they could be "lost".
539 - Reset timeout values before select() per patch from
540 Frank Sweetser for problems with non-blocking sockets.
541 - Unlink the state file if either reading or writing it gets
542 errors. Hopefully this will fix Win32 exit problems.
543 - Add sanity check in append.c to ensure that dcr is not NULL.
544 This can happen if multiple drive autochanger SCSI control
545 channel and drive indicies do not correspond.
547 - Get next volume from Scratch pool before creating a volume.
548 - Set new Pool defaults in Vol when moved from Scratch Pool.
549 - Remove argument from create_bacula_database for SQLite as it
551 - Add back index code so that two drive autochangers can get
553 - Change a bunch of debug levels to aid debugging autochangers.
554 - Fix reservation so that mutexes are properly applied.
555 - Rework reservation algorithm so that two drives can be used
558 - Landon merged his data encription changes into the HEAD
559 - Apply days keyword patch from Alexander.Bergolth at wu-wien.ac.at
560 If this patch is applied, the number of days can be specified with
561 "list nextvol days=xx"
564 My use case is to be able to preview the next scheduled job (and the
565 next tape to be used) on fridays if there are no scheduled jobs during
569 - Fix font code in gnome2 console user patch. Fixes bug #501.
570 - Fix malformatted bnet error message that caused seg fault
572 - Applied user patch to improve README.vc8 in src/win32.
574 - Add Migrate, Copy, Archive Job types (some where there)
575 - Correct some more editing of JobId's (for 64 bit compatibility).
576 - Ensure that StorageId is stored in Media record when ever possible.
577 - Add Migration Job to Job.
578 - Add Migration Time, Migration High Bytes, Migration Low Bytes
579 Next Pool to Pool resource.
580 - Add more code to mac.c (migration archive copy).
581 - Change Start Storage daemon job to require read and write storage
583 - Pass read storage data to SD as well as write storage data.
584 - Remove old code from winservice.cpp
585 - Break on error in scan.
586 - Fix typo in signal.c
587 - Separate read/write DCR in SD. Add jcr->read_dcr.
588 - Cleanup how find_device() works.
589 - Add read output to Status in SD.
592 - Add red-black btree routines
594 - Remove abs() in bfile.c so that it compiles on Solaris.
597 - Fix crash in tray-monitor when daemon disconnects. Bug #479.
598 - Fix bnet-server bug found on OpenBSD. Bug #486
599 - Fix cancel failure bug. Bug #481
600 - Fix failure when Pool name has spaces. Bug #487
601 - Fix SD crash in autochanger code. Mutex failure. Bug #488
602 - Fix a couple of free()s in src/filed/acl.c
603 - Fix memory overrun in bfile.c in building OS X resource
604 fork filename. Bug #489
605 - Add Pool name to SD status output.
607 - Apply SunOS patch for ACLs submitted by David Duchscher.
608 - Make sure to set storage before trying to set drive.
609 - Add bacula_mail_summary.sh to examples directory. It makes
610 a single email summary of any number of jobs. Submitted
612 - Make sure when we do a mount to unblock the device even
613 if the drive could not be opened.
615 - Remove the USE_WIN32STREAMEXTRACTION #defines (always on)
616 and correct a few minor problems to make it build on Linux.
618 - Remove delete of CVS from all Makefiles
619 - Fix seg fault when clicking on Add button in wx-console
620 restore panel. Bug #470.
621 - Fix copyright date and URL typo -- bug #468.
622 - Change autostart install for FreeBSD to look for rc.conf
623 rather than rc.local as suggested fix for bug #466.
624 - Apply patch supplied by Eric Bollinger to fix PostgreSQL
625 grant on status. Bug #465
626 - Apply patch supplied by Eric Bollinger to fix PostgreSQL
627 update script. Bug #464
628 - Tweak #ifdefing a bit in new Win32 stream code.
629 - Fix #ifdeffing for FD_NO_SEND_TEST.
630 - Add documentation of performance #defines