From: Kern Sibbald Date: Mon, 25 Nov 2013 17:00:15 +0000 (+0100) Subject: Update news.txt X-Git-Tag: Release-7.0.0~32 X-Git-Url: https://git.sur5r.net/?a=commitdiff_plain;h=30491a02e6eadd2d5eed17bd9e1004ddc1777845;p=bacula%2Fdocs Update news.txt --- diff --git a/docs/home-page/de/news.txt b/docs/home-page/de/news.txt index 6e82c8d1..e0df4e40 100644 --- a/docs/home-page/de/news.txt +++ b/docs/home-page/de/news.txt @@ -1,3 +1,529 @@ +Kern;;;25 November 2013 at 17:53 CET +Bacula Status Report + +Hello, + +I would like to speak to you about the following points: + +1. The rumors of the death of Bacula (the Community version) +2. The Bareos fork of Bacula +3. Bacula Systems and the FSFE (Free Software Foundation Europe) +4. The future of Bacula (the Community version) + +1. The rumors of the death of Bacula (the Community version): + +I borrow words from a quote of Mark Twain : The rumors of the death +of Bacula are highly exaggerated! + +I began working on Bacula 14 years ago (in January 2000), and it has +been Open Source from the time it was publicly released in April +2002, and it will remain Open Source. I have been and am fully +devoted to Open Source, and in particular to Bacula, which is like +my “baby”. So to hear rumors that Bacula is dead or that I have +withheld commits because they are Enterprise features is shocking +and hurtful to me as well as not true. + +I did inform the Bacula Community several years ago that my personal +participation in Bacula would decrease a bit for several years to +allow me to focus more on getting Bacula Systems started. In my +opinion, that has not been a serious disadvantage for the Bacula +project since Bacula Systems over that period has contributed far +more code to Bacula than I could have alone over the same period, +and as you will see a bit later in this status report, Bacula +Systems contributions are absolutely guaranteed to continue in the +long run, and even increase. + +2. The Bareos fork of Bacula: + +The Bacula repository has been on “hold” since our last release +in early February, because on 27 February 2013, I learned that there +was a fork of Bacula made by a former “consultant” of Bacula +Systems with a former reseller of Bacula Systems. Unfortunately, +despite the fact that Bareos hired one of the best German Open +Source lawyers , there were a number of serious copyright violations +with their code. Since the Bacula code is copyrighted by the Free +Software Foundation Europe (FSFE), I provided the technical support, +and the FSFE worked with Bareos to clean up their copyright +violations. That took a significant amount of time (many months), +and the Bareos code though significantly improved is still not +totally free of copyright infringements. I won't go into the +details here as all of you may not be interested, but will have much +more to say about Bareos in later blogs, and when the blog is setup +I will let you know. + +I find Bareos an unusual fork, because it wasn't done in what I +believe to be the normal Open Source way. Normally a fork is made +when a project is blocked or has serious disagreements with the +users. Its also normally done in open communication rather than +underhanded or in secret. In the case of Bacula, though some of the +development slowed down (I will go into this in detail later), it +certainly was by no means stopped. To complain about active +development in Bacula Systems, is, in my opinion, incorrect first +because adding features to the Enterprise version costs Bacula +Systems a lot mostly in salaries yet takes absolutely nothing from +Bacula. In fact, when you read the next section, you will see that +the more that Bacula Systems develops, the more features that Bacula +over time will have. Going back to what I find abnormal about the +Bareos fork is that they claim that they spent three years +developing a lot of new features, thus they are more feature rich +than Bacula. Yes, for the moment, they have a few features that +Bacula does not yet have, but not for long, and more importantly +over the three years of development of those features they never +offered these new features to Bacula nor to any Open Source project. +Instead they were developed in secret. I find that a very strange +behavior for a self-proclaimed Open Source company (actually, they +are “forced” to be Open Source because of the Bacula AGPLv3 +license). So as Bacula contributors and users, you would be within +your rights to feel very upset with Bareos, because they never +offered you the code they developed. + +I assure that I will do all in my power to ensure that any +worthwhile features that Bareos implements will be implemented in +Bacula, and most likely better integrated and more robust, and where +possible with even more functionality and growth potential. + + +3. Bacula Systems and the FSFE: + +If you have been a long time Bacula user you may recall that I +discussed the possibility in 2006-2007 of creating a company, now +called Bacula Systems, to ensure the continuation of Bacula when I +will no longer be able to personally contribute – say in 10 or 20 +years, as well as to provide the financial means to add high-end +features to Bacula (a fibre channel network costs about $50K to +setup). Much to my surprise 95% or more of the responses I got were +very positive. Bacula Systems was created in July 2008, and for the +first two years, the Enterprise code base and the Community code +base were identical. Unfortunately, that didn't work financially +for Bacula Systems. Companies willing to pay, were willing to pay +for features and support but not support alone, so Bacula Systems +embarked on development to continue maintenance and improvement of +Bacula while at the same time creating mostly plugins to add +differentiation to the Enterprise version. + +Now this may not sound very Open Source to you, and I understand, +because I feel the same way. Were it at all possible, I would give +you all of Bacula Systems code, unfortunately, that is not +economically feasible at the current time, and yet without Bacula +Systems, I fear the Bacula project will die or worse yet fall into +the hands of someone incapable of maintaining the high quality we +have created. + +While I was consulting with the Free Software Foundation Europe +(FSFE) on the Bareos copyright violations, Bacula Systems and I +began discussions with the FSFE on how to guarantee the long term +survival of Bacula. These discussions, extremely positive on both +sides and all points, recently lead to a formal written agreement +between myself, Bacula Systems, and the FSFE. There are a number of +points in the agreement, but probably the most important of all is +that Bacula Systems has now put in writing that it is an Open Source +company (at its heart), as it has always proclaimed, and will +contribute all the Enterprise code it creates to the Bacula +Community code base within at most a 5 year period. One exception +is that Bacula Systems is legally unable to contribute certain code +encumbered by third party proprietary license. The 5 year delay +gives Bacula Systems the chance to develop Enterprise features that +differentiate it, but ensures the continual growth of the Bacula +Community code. This model can possibly be used across the industry +to ensure the future of open source software in an environment where +development costs, particularly for hardware to do testing, are +prohibitive to the standard models of today. + +5. The future of Bacula (the Community version): + +If you have read section 4 above, hopefully if you were not already +convinced that Bacula is alive that you can now see that it will +have a long and successful future ahead of it. If you have any +doubts, please do not hesitate to either send me an email on the +bacula-users list or directly to me (if you want it private). +Hopefully, by mid-December I will have a blog setup (need a major +upgrade of bacula.org to do so), and I will then fill you in on the +details of the Bareos fork as well as more details on what next to +expect in Bacula. + +Thank you for contributing to and/or using Bacula ... + +Best regards, +Kern +;;;; + + +Kern;;;20 February 2013 at 13:43 CET +We have released Bacula Version 5.2.13 to Source Forge + +Version 5.2.13 version has 20 bug fixes since the last +release. Some of the bug fixes are important. + +Important changes since last release: + +19Feb13 + - Fix build/configure problems with bpluginfo.c + - Refactor lock_volumes so most lock a vol rather than globally + - Add virtualfull-extreme test + - Apply patch for chio-changer-openbsd from bug #1984 -- Implements listall + - Add bat Mac patch from bug #1953 + - Fix bug #1812 cannot run Copy/Migrate jobs from bat + - Fix loading of bat translations. Patch from bug #1890 + - Fix text input in bat. Fixe bug #1965 + - Remove Device from show command as it is not used or updated + - Fix #1982 update enabled keyword in help command + - Fix update_postgresql_tables 10 to 11 + - Clean after building doc binaries requested by Willem vd Akker for Debian + build + - Require Qt version 4.8.4 + - Fix bug #1955 that OK to run? does not retry on bad response + - Backport new lock calls + debug for SD + - Fix bug #1975 new label ignored when first is duplicate. + - Pull src/lib changes from master + - Pull SD files from master + - Turn off prune-test + - Add logdir and bsrdir to regression config + - Add bsrdir and logdir to regress-config + - Changed default directories from /var/bacula to /opt/bacula + - Make spool size max message more explicit + - Display more info when maximum spool size is reached + - Fix TERM defs for Solaris 11 in conio.c + - Allow group to cd to sysconfdir + - Make sysconfdir belong to Bacula user/group and exclude others + - Fix Virtual Full file close bug -- race condition + - Enhance mount message to include read/append + - Improve vol_mgr swap messages + - Enhance output of Using Device to include for read/write + - Add jobid to Pmsg output + - Fix bug #1959 input validation on delete of jobs. + - Fix bug #1956 Authorization Errors false + - Add workaround for #5507 where autoprune and reload are in deadlock + - Allow conf file quoted strings to be used in a list + - Make bfgets handle very long lines + - Avoid seg fault by checking for NULL client -- reported by Arno + - Final fix of bug #1943 + - Fix bug #1948 MailOnSuccess get executed on error. + - Fix missing index on Media table + - Fix bug #1943 no message storage on closed database connection. + - Fix bug #1946 Fix problem with MySQL with big Base jobs. + - Reduce wait time in regression RunScripts + - Implement a timeout on regression tests of 15 minutes + - Fix #5346 .bvfs_lsfiles and .bvfs_restore to handle deleted files + - Fix spooldata, accurate and ingnoreduplicate run argument + - Fix messages segfault + - Implement feature request #1939 + - fix #1938 about PATH_MAX on hurd + - fix #1938 about MAXPATHLEN on hurd + - Fix bug #1937 OpenBSD autochanger example script + - Fix #4996 about MaxRunTime canceling the job too early + - Fix bug #1934 Wrong values at media list in BAT + - Fix race condition in close_msg that causes seg fault + - Fix double unlink + - More tests for freeing NULL pointers + - Add MaximumConcurrentReadJobs directive to Storage resource + - Add ujobid to .bvfs_get_jobids + - Avoid problem when stripping a relative path + - Fix #4513 about HIDDEN attribute set to parent directory during restore + +Bugs fixed/closed since last release: +1812 1890 1934 1937 1938 1939 1943 1946 1948 1953 1955 1956 1959 1965 1975 +1982 1984 4513 4996 5346 5507 + + +;; +Kern;;;11 September 2012 at 13:43 CET +We have released Bacula Version 5.2.11 to Source Forge + +Version 5.2.11 version is a bug fix release. + + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.11 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Important changes since last release: + - A new bpluginfo utility tool + - Make dump_resource respect console ACL's (security fix) + - Add chio-changer-openbsd submitted in bug #1903 + - Important Copy job bug fixed. + +10Sep12 + - Add JobId to SD debug output + - Commit batch session every 800,000 files + - Simplify safer delete code using a single regex + - Clean up error message generation during update attributes -- should fix bug + #1823 + - Eliminate heartbeat error messages. Fixes bug #1925 + - Apply patch from bug #1911 to implement --with-logdir + - Add chio-changer-openbsd submitted in bug #1903 + - mysql database install/update scripts patch. Fixes bug #1901 + - Delegate uid/gid changing to systemd. Fixes bug #1905 + - Fix bug #1906 + - Fix manpage errors, fixes bug #1907 + - Final fix for #1859 missed one query. + - bpluginfo - Utility tool display various information + - Backport more master code + - Fix Makefile.in so that testfind builds with acl dependency + - Make dump_resource respect console ACL's (security fix) + - backport code from master + - Fix #1923 about MySQL 4 support for BVFS + - Fix #1902 about bad status slots output + - Fix Sigma Copy bug #4377 + - Fix bsmtp + - Reset bsmtp to only send to IPv4 mailhosts. + +Bugs fixed/closed since last release: +1823 1859 1901 1902 1903 1905 1906 1907 1911 1923 1925 4377 +;;;; + +Kern;;;28 June 2012 at 20:55 CET +We have released Bacula Version 5.2.10 to Source Forge + +General: +-------- +The 5.2.10 version is a significant bug fix release including +our first cut of AFS support. + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.9, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.9 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Major changes since 5.2.9: +- IPv6 networking should work correctly now +- SQL Counter problems resolved +- Bat hanging after running a number of jobs +- dbcheck rw_writelock problem +- Windows VSS restore bug + +New Feature: +- Add rudimentary support for backing up AFS + +28Jun12 + - Add extra test for unsupported Protocol Families. + - Fix inet_pton call. + - Real fix of bug #1897 5.2.9 breaks IPv6 connectivity + - Fix bug #1891 wrong daemon name printed with -? option + - Fix bug #1859 INSERT INTO Counters fails + - First try at fixing bug #1897 5.2.9 breaks IPv6 connectivity + - Convert write_rec_to_block() to a state machine + - Fix bat from eating all Dir connections -- fixes bug #1872 + - Back port fixes from BEE to fix Windows restore bug #1892 + - Add block checksum flag to version.h + - Add block checksum debug code + - Fix bug #1893 dbcheck -B gives rwl_writelock error. + - Remove restriction of network buffer size being multiple of a tape block + - Add rudimentary support for saving AFS acls. + - Send level before the fileset in estimate_cmd() like in backup() + +Bugs fixed/closed since last release: +1859 1872 1891 1892 1893 1897 + +;;;; + +Kern;;;11 June 2012 at 20:15 CET +We have released Bacula Version 5.2.8 to Source Forge + +This release is identical to release 5.2.8, with the exception +that the patch: + + "Apply readline without TERM_LIB dependency patch from bug #1871" + +has been removed because it is not portable to all systems or +versions of readline. + +;;;; +Kern;;;10 June 2012 at 12:04 CET + + Release Notes for Bacula 5.2.8 + + Bacula code: Total files = 1,110 Total lines = 231,439 (Using SLOCCount) + +General: +-------- +The 5.2.8 version is a bug fix release. + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.7, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.8 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Changes since 5.2.7: +-------------------- + +10Jun12 + - Apply readline without TERM_LIB dependency patch from bug #1871 + - Ensure cancel status is set for canceled duplicate jobs -- partial fix for + bug #1851 + - Enhance Win32 error messages -- fixes bug #1879 + - Apply patch from bug 1876 to permit dot commands in a runscript + - Remove config variables from examples/devices fixes bug #1884 + - Add more debug code for bug 1885 + - Add more debug to block.c when read rejected + - Fail job that attempts to read from closed device -- reported in bug 1885 but + not fixed + - Detect Python 2.7 fixes bug #1888 + - Apply systemd patch from bug #1886 + - Rewrite some SD subroutines as class members + - New overhaul of xattr code. + - Use IPPROTO_TCP for SOL_TCP if SOL_TCP isn't defined. + - Fix compiler warning. + +Bugs fixed/closed since last release: +1851 1871 1879 1884 1886 1888 + +;;;; +Kern;;;2 June 2012 at 13:01 CET + + Release Notes for Bacula 5.2.7 + + Bacula code: Total files = 1,110 Total lines = 231,439 (Using SLOCCount) + +General: +-------- +The 5.2.7 version is a significan bug fix release. (The +diff file between 5.2.6 and this one is over 25,000 lines). + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.7, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.3 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Changes since 5.2.6: +-------------------- +02Jun12 + - Correct Qt version needed for bat + - Turn on SD deadlock detection by default + - Fix bug #1880 ltdl.m4 missing for new feature. + - Fix bug #1870 GNU binutils-gold build failure. + - Add new empty netinet/tcp.h so Windows compile works + - Eliminate compiler warning in stored/dircmd.c + - Fix TCP Heartbeat code + - Align conio.c struct at 8 bytes to fix Sparc bus error. + - Fix get_basename() -- rewrite + - Apply James' exchange-fd.c regression fix + - Add first Dutch translation + - Ensure that StorageId is updated after write + - Fix updating of cached slots + error message numbering + - Remove old shared objects before install + - Add missing dummy bVarPrefixLinks variable retrieval. + - Fix mysql config for older versions of mysql+config. + - Fix problem in BVFS with concurrent queries + - Add more checks to bvfs + - Try to use multi-row insert statements for mysql. + - On some platforms intptr_t is in an other include file. + - Add support for fcntl(fd, F_CLOSEM) to close filedescriptors. + - Use closefrom if available instead of trying to close fds ourself. + - Fix class member shadowing. + - Move storages tls initialization to proper place. + - Fix Copyright on init scripts suse, add specific daemon user & group + - Add support for new POSIX getaddrinfo interface. + - Apply MSQL detection fixes from bug #1829 + - Fix bug #1805 cannot specify hostname in ./configure + - Fix restore_job restore keyword + - Allow plugin to backup a directory + - Print seconds in schedule time as noted in bug #1854 + - Fix #1857 about restore cd command that blocks on empty directory + - Backport new StorageId code + - More definitive fix for update slots bug + - Fix #4062 on make_catalog_backup.pl script when user/password are empty + - Fix bug #1853: bacula-sd dead but pid file exists. + - Fix bug #1841 estimate command level=differential goes wrong in accurate + mode. + - Fix bug #1842 Incorrect port number reported + - Fix bug #1848 bsock.c:335 Socket has errors=1 on call to client + - Force the re-initialization of BaseJobOpts, AccurateOpts and VerifyOpts + between two Include{} + +Bugs fixed/closed since last release: +1805 1829 1841 1842 1848 1853 1854 1857 1870 1874 1880 4062 + + +;;;; Kern;;;22 February 2012 at 15:03 CET We have released Bacula Version 5.2.6 to Source Forge diff --git a/docs/home-page/en/news.txt b/docs/home-page/en/news.txt index 8328c835..e0df4e40 100644 --- a/docs/home-page/en/news.txt +++ b/docs/home-page/en/news.txt @@ -1,3 +1,153 @@ +Kern;;;25 November 2013 at 17:53 CET +Bacula Status Report + +Hello, + +I would like to speak to you about the following points: + +1. The rumors of the death of Bacula (the Community version) +2. The Bareos fork of Bacula +3. Bacula Systems and the FSFE (Free Software Foundation Europe) +4. The future of Bacula (the Community version) + +1. The rumors of the death of Bacula (the Community version): + +I borrow words from a quote of Mark Twain : The rumors of the death +of Bacula are highly exaggerated! + +I began working on Bacula 14 years ago (in January 2000), and it has +been Open Source from the time it was publicly released in April +2002, and it will remain Open Source. I have been and am fully +devoted to Open Source, and in particular to Bacula, which is like +my “baby”. So to hear rumors that Bacula is dead or that I have +withheld commits because they are Enterprise features is shocking +and hurtful to me as well as not true. + +I did inform the Bacula Community several years ago that my personal +participation in Bacula would decrease a bit for several years to +allow me to focus more on getting Bacula Systems started. In my +opinion, that has not been a serious disadvantage for the Bacula +project since Bacula Systems over that period has contributed far +more code to Bacula than I could have alone over the same period, +and as you will see a bit later in this status report, Bacula +Systems contributions are absolutely guaranteed to continue in the +long run, and even increase. + +2. The Bareos fork of Bacula: + +The Bacula repository has been on “hold” since our last release +in early February, because on 27 February 2013, I learned that there +was a fork of Bacula made by a former “consultant” of Bacula +Systems with a former reseller of Bacula Systems. Unfortunately, +despite the fact that Bareos hired one of the best German Open +Source lawyers , there were a number of serious copyright violations +with their code. Since the Bacula code is copyrighted by the Free +Software Foundation Europe (FSFE), I provided the technical support, +and the FSFE worked with Bareos to clean up their copyright +violations. That took a significant amount of time (many months), +and the Bareos code though significantly improved is still not +totally free of copyright infringements. I won't go into the +details here as all of you may not be interested, but will have much +more to say about Bareos in later blogs, and when the blog is setup +I will let you know. + +I find Bareos an unusual fork, because it wasn't done in what I +believe to be the normal Open Source way. Normally a fork is made +when a project is blocked or has serious disagreements with the +users. Its also normally done in open communication rather than +underhanded or in secret. In the case of Bacula, though some of the +development slowed down (I will go into this in detail later), it +certainly was by no means stopped. To complain about active +development in Bacula Systems, is, in my opinion, incorrect first +because adding features to the Enterprise version costs Bacula +Systems a lot mostly in salaries yet takes absolutely nothing from +Bacula. In fact, when you read the next section, you will see that +the more that Bacula Systems develops, the more features that Bacula +over time will have. Going back to what I find abnormal about the +Bareos fork is that they claim that they spent three years +developing a lot of new features, thus they are more feature rich +than Bacula. Yes, for the moment, they have a few features that +Bacula does not yet have, but not for long, and more importantly +over the three years of development of those features they never +offered these new features to Bacula nor to any Open Source project. +Instead they were developed in secret. I find that a very strange +behavior for a self-proclaimed Open Source company (actually, they +are “forced” to be Open Source because of the Bacula AGPLv3 +license). So as Bacula contributors and users, you would be within +your rights to feel very upset with Bareos, because they never +offered you the code they developed. + +I assure that I will do all in my power to ensure that any +worthwhile features that Bareos implements will be implemented in +Bacula, and most likely better integrated and more robust, and where +possible with even more functionality and growth potential. + + +3. Bacula Systems and the FSFE: + +If you have been a long time Bacula user you may recall that I +discussed the possibility in 2006-2007 of creating a company, now +called Bacula Systems, to ensure the continuation of Bacula when I +will no longer be able to personally contribute – say in 10 or 20 +years, as well as to provide the financial means to add high-end +features to Bacula (a fibre channel network costs about $50K to +setup). Much to my surprise 95% or more of the responses I got were +very positive. Bacula Systems was created in July 2008, and for the +first two years, the Enterprise code base and the Community code +base were identical. Unfortunately, that didn't work financially +for Bacula Systems. Companies willing to pay, were willing to pay +for features and support but not support alone, so Bacula Systems +embarked on development to continue maintenance and improvement of +Bacula while at the same time creating mostly plugins to add +differentiation to the Enterprise version. + +Now this may not sound very Open Source to you, and I understand, +because I feel the same way. Were it at all possible, I would give +you all of Bacula Systems code, unfortunately, that is not +economically feasible at the current time, and yet without Bacula +Systems, I fear the Bacula project will die or worse yet fall into +the hands of someone incapable of maintaining the high quality we +have created. + +While I was consulting with the Free Software Foundation Europe +(FSFE) on the Bareos copyright violations, Bacula Systems and I +began discussions with the FSFE on how to guarantee the long term +survival of Bacula. These discussions, extremely positive on both +sides and all points, recently lead to a formal written agreement +between myself, Bacula Systems, and the FSFE. There are a number of +points in the agreement, but probably the most important of all is +that Bacula Systems has now put in writing that it is an Open Source +company (at its heart), as it has always proclaimed, and will +contribute all the Enterprise code it creates to the Bacula +Community code base within at most a 5 year period. One exception +is that Bacula Systems is legally unable to contribute certain code +encumbered by third party proprietary license. The 5 year delay +gives Bacula Systems the chance to develop Enterprise features that +differentiate it, but ensures the continual growth of the Bacula +Community code. This model can possibly be used across the industry +to ensure the future of open source software in an environment where +development costs, particularly for hardware to do testing, are +prohibitive to the standard models of today. + +5. The future of Bacula (the Community version): + +If you have read section 4 above, hopefully if you were not already +convinced that Bacula is alive that you can now see that it will +have a long and successful future ahead of it. If you have any +doubts, please do not hesitate to either send me an email on the +bacula-users list or directly to me (if you want it private). +Hopefully, by mid-December I will have a blog setup (need a major +upgrade of bacula.org to do so), and I will then fill you in on the +details of the Bareos fork as well as more details on what next to +expect in Bacula. + +Thank you for contributing to and/or using Bacula ... + +Best regards, +Kern +;;;; + + Kern;;;20 February 2013 at 13:43 CET We have released Bacula Version 5.2.13 to Source Forge diff --git a/docs/home-page/es/news.txt b/docs/home-page/es/news.txt index 6e82c8d1..e0df4e40 100644 --- a/docs/home-page/es/news.txt +++ b/docs/home-page/es/news.txt @@ -1,3 +1,529 @@ +Kern;;;25 November 2013 at 17:53 CET +Bacula Status Report + +Hello, + +I would like to speak to you about the following points: + +1. The rumors of the death of Bacula (the Community version) +2. The Bareos fork of Bacula +3. Bacula Systems and the FSFE (Free Software Foundation Europe) +4. The future of Bacula (the Community version) + +1. The rumors of the death of Bacula (the Community version): + +I borrow words from a quote of Mark Twain : The rumors of the death +of Bacula are highly exaggerated! + +I began working on Bacula 14 years ago (in January 2000), and it has +been Open Source from the time it was publicly released in April +2002, and it will remain Open Source. I have been and am fully +devoted to Open Source, and in particular to Bacula, which is like +my “baby”. So to hear rumors that Bacula is dead or that I have +withheld commits because they are Enterprise features is shocking +and hurtful to me as well as not true. + +I did inform the Bacula Community several years ago that my personal +participation in Bacula would decrease a bit for several years to +allow me to focus more on getting Bacula Systems started. In my +opinion, that has not been a serious disadvantage for the Bacula +project since Bacula Systems over that period has contributed far +more code to Bacula than I could have alone over the same period, +and as you will see a bit later in this status report, Bacula +Systems contributions are absolutely guaranteed to continue in the +long run, and even increase. + +2. The Bareos fork of Bacula: + +The Bacula repository has been on “hold” since our last release +in early February, because on 27 February 2013, I learned that there +was a fork of Bacula made by a former “consultant” of Bacula +Systems with a former reseller of Bacula Systems. Unfortunately, +despite the fact that Bareos hired one of the best German Open +Source lawyers , there were a number of serious copyright violations +with their code. Since the Bacula code is copyrighted by the Free +Software Foundation Europe (FSFE), I provided the technical support, +and the FSFE worked with Bareos to clean up their copyright +violations. That took a significant amount of time (many months), +and the Bareos code though significantly improved is still not +totally free of copyright infringements. I won't go into the +details here as all of you may not be interested, but will have much +more to say about Bareos in later blogs, and when the blog is setup +I will let you know. + +I find Bareos an unusual fork, because it wasn't done in what I +believe to be the normal Open Source way. Normally a fork is made +when a project is blocked or has serious disagreements with the +users. Its also normally done in open communication rather than +underhanded or in secret. In the case of Bacula, though some of the +development slowed down (I will go into this in detail later), it +certainly was by no means stopped. To complain about active +development in Bacula Systems, is, in my opinion, incorrect first +because adding features to the Enterprise version costs Bacula +Systems a lot mostly in salaries yet takes absolutely nothing from +Bacula. In fact, when you read the next section, you will see that +the more that Bacula Systems develops, the more features that Bacula +over time will have. Going back to what I find abnormal about the +Bareos fork is that they claim that they spent three years +developing a lot of new features, thus they are more feature rich +than Bacula. Yes, for the moment, they have a few features that +Bacula does not yet have, but not for long, and more importantly +over the three years of development of those features they never +offered these new features to Bacula nor to any Open Source project. +Instead they were developed in secret. I find that a very strange +behavior for a self-proclaimed Open Source company (actually, they +are “forced” to be Open Source because of the Bacula AGPLv3 +license). So as Bacula contributors and users, you would be within +your rights to feel very upset with Bareos, because they never +offered you the code they developed. + +I assure that I will do all in my power to ensure that any +worthwhile features that Bareos implements will be implemented in +Bacula, and most likely better integrated and more robust, and where +possible with even more functionality and growth potential. + + +3. Bacula Systems and the FSFE: + +If you have been a long time Bacula user you may recall that I +discussed the possibility in 2006-2007 of creating a company, now +called Bacula Systems, to ensure the continuation of Bacula when I +will no longer be able to personally contribute – say in 10 or 20 +years, as well as to provide the financial means to add high-end +features to Bacula (a fibre channel network costs about $50K to +setup). Much to my surprise 95% or more of the responses I got were +very positive. Bacula Systems was created in July 2008, and for the +first two years, the Enterprise code base and the Community code +base were identical. Unfortunately, that didn't work financially +for Bacula Systems. Companies willing to pay, were willing to pay +for features and support but not support alone, so Bacula Systems +embarked on development to continue maintenance and improvement of +Bacula while at the same time creating mostly plugins to add +differentiation to the Enterprise version. + +Now this may not sound very Open Source to you, and I understand, +because I feel the same way. Were it at all possible, I would give +you all of Bacula Systems code, unfortunately, that is not +economically feasible at the current time, and yet without Bacula +Systems, I fear the Bacula project will die or worse yet fall into +the hands of someone incapable of maintaining the high quality we +have created. + +While I was consulting with the Free Software Foundation Europe +(FSFE) on the Bareos copyright violations, Bacula Systems and I +began discussions with the FSFE on how to guarantee the long term +survival of Bacula. These discussions, extremely positive on both +sides and all points, recently lead to a formal written agreement +between myself, Bacula Systems, and the FSFE. There are a number of +points in the agreement, but probably the most important of all is +that Bacula Systems has now put in writing that it is an Open Source +company (at its heart), as it has always proclaimed, and will +contribute all the Enterprise code it creates to the Bacula +Community code base within at most a 5 year period. One exception +is that Bacula Systems is legally unable to contribute certain code +encumbered by third party proprietary license. The 5 year delay +gives Bacula Systems the chance to develop Enterprise features that +differentiate it, but ensures the continual growth of the Bacula +Community code. This model can possibly be used across the industry +to ensure the future of open source software in an environment where +development costs, particularly for hardware to do testing, are +prohibitive to the standard models of today. + +5. The future of Bacula (the Community version): + +If you have read section 4 above, hopefully if you were not already +convinced that Bacula is alive that you can now see that it will +have a long and successful future ahead of it. If you have any +doubts, please do not hesitate to either send me an email on the +bacula-users list or directly to me (if you want it private). +Hopefully, by mid-December I will have a blog setup (need a major +upgrade of bacula.org to do so), and I will then fill you in on the +details of the Bareos fork as well as more details on what next to +expect in Bacula. + +Thank you for contributing to and/or using Bacula ... + +Best regards, +Kern +;;;; + + +Kern;;;20 February 2013 at 13:43 CET +We have released Bacula Version 5.2.13 to Source Forge + +Version 5.2.13 version has 20 bug fixes since the last +release. Some of the bug fixes are important. + +Important changes since last release: + +19Feb13 + - Fix build/configure problems with bpluginfo.c + - Refactor lock_volumes so most lock a vol rather than globally + - Add virtualfull-extreme test + - Apply patch for chio-changer-openbsd from bug #1984 -- Implements listall + - Add bat Mac patch from bug #1953 + - Fix bug #1812 cannot run Copy/Migrate jobs from bat + - Fix loading of bat translations. Patch from bug #1890 + - Fix text input in bat. Fixe bug #1965 + - Remove Device from show command as it is not used or updated + - Fix #1982 update enabled keyword in help command + - Fix update_postgresql_tables 10 to 11 + - Clean after building doc binaries requested by Willem vd Akker for Debian + build + - Require Qt version 4.8.4 + - Fix bug #1955 that OK to run? does not retry on bad response + - Backport new lock calls + debug for SD + - Fix bug #1975 new label ignored when first is duplicate. + - Pull src/lib changes from master + - Pull SD files from master + - Turn off prune-test + - Add logdir and bsrdir to regression config + - Add bsrdir and logdir to regress-config + - Changed default directories from /var/bacula to /opt/bacula + - Make spool size max message more explicit + - Display more info when maximum spool size is reached + - Fix TERM defs for Solaris 11 in conio.c + - Allow group to cd to sysconfdir + - Make sysconfdir belong to Bacula user/group and exclude others + - Fix Virtual Full file close bug -- race condition + - Enhance mount message to include read/append + - Improve vol_mgr swap messages + - Enhance output of Using Device to include for read/write + - Add jobid to Pmsg output + - Fix bug #1959 input validation on delete of jobs. + - Fix bug #1956 Authorization Errors false + - Add workaround for #5507 where autoprune and reload are in deadlock + - Allow conf file quoted strings to be used in a list + - Make bfgets handle very long lines + - Avoid seg fault by checking for NULL client -- reported by Arno + - Final fix of bug #1943 + - Fix bug #1948 MailOnSuccess get executed on error. + - Fix missing index on Media table + - Fix bug #1943 no message storage on closed database connection. + - Fix bug #1946 Fix problem with MySQL with big Base jobs. + - Reduce wait time in regression RunScripts + - Implement a timeout on regression tests of 15 minutes + - Fix #5346 .bvfs_lsfiles and .bvfs_restore to handle deleted files + - Fix spooldata, accurate and ingnoreduplicate run argument + - Fix messages segfault + - Implement feature request #1939 + - fix #1938 about PATH_MAX on hurd + - fix #1938 about MAXPATHLEN on hurd + - Fix bug #1937 OpenBSD autochanger example script + - Fix #4996 about MaxRunTime canceling the job too early + - Fix bug #1934 Wrong values at media list in BAT + - Fix race condition in close_msg that causes seg fault + - Fix double unlink + - More tests for freeing NULL pointers + - Add MaximumConcurrentReadJobs directive to Storage resource + - Add ujobid to .bvfs_get_jobids + - Avoid problem when stripping a relative path + - Fix #4513 about HIDDEN attribute set to parent directory during restore + +Bugs fixed/closed since last release: +1812 1890 1934 1937 1938 1939 1943 1946 1948 1953 1955 1956 1959 1965 1975 +1982 1984 4513 4996 5346 5507 + + +;; +Kern;;;11 September 2012 at 13:43 CET +We have released Bacula Version 5.2.11 to Source Forge + +Version 5.2.11 version is a bug fix release. + + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.11 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Important changes since last release: + - A new bpluginfo utility tool + - Make dump_resource respect console ACL's (security fix) + - Add chio-changer-openbsd submitted in bug #1903 + - Important Copy job bug fixed. + +10Sep12 + - Add JobId to SD debug output + - Commit batch session every 800,000 files + - Simplify safer delete code using a single regex + - Clean up error message generation during update attributes -- should fix bug + #1823 + - Eliminate heartbeat error messages. Fixes bug #1925 + - Apply patch from bug #1911 to implement --with-logdir + - Add chio-changer-openbsd submitted in bug #1903 + - mysql database install/update scripts patch. Fixes bug #1901 + - Delegate uid/gid changing to systemd. Fixes bug #1905 + - Fix bug #1906 + - Fix manpage errors, fixes bug #1907 + - Final fix for #1859 missed one query. + - bpluginfo - Utility tool display various information + - Backport more master code + - Fix Makefile.in so that testfind builds with acl dependency + - Make dump_resource respect console ACL's (security fix) + - backport code from master + - Fix #1923 about MySQL 4 support for BVFS + - Fix #1902 about bad status slots output + - Fix Sigma Copy bug #4377 + - Fix bsmtp + - Reset bsmtp to only send to IPv4 mailhosts. + +Bugs fixed/closed since last release: +1823 1859 1901 1902 1903 1905 1906 1907 1911 1923 1925 4377 +;;;; + +Kern;;;28 June 2012 at 20:55 CET +We have released Bacula Version 5.2.10 to Source Forge + +General: +-------- +The 5.2.10 version is a significant bug fix release including +our first cut of AFS support. + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.9, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.9 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Major changes since 5.2.9: +- IPv6 networking should work correctly now +- SQL Counter problems resolved +- Bat hanging after running a number of jobs +- dbcheck rw_writelock problem +- Windows VSS restore bug + +New Feature: +- Add rudimentary support for backing up AFS + +28Jun12 + - Add extra test for unsupported Protocol Families. + - Fix inet_pton call. + - Real fix of bug #1897 5.2.9 breaks IPv6 connectivity + - Fix bug #1891 wrong daemon name printed with -? option + - Fix bug #1859 INSERT INTO Counters fails + - First try at fixing bug #1897 5.2.9 breaks IPv6 connectivity + - Convert write_rec_to_block() to a state machine + - Fix bat from eating all Dir connections -- fixes bug #1872 + - Back port fixes from BEE to fix Windows restore bug #1892 + - Add block checksum flag to version.h + - Add block checksum debug code + - Fix bug #1893 dbcheck -B gives rwl_writelock error. + - Remove restriction of network buffer size being multiple of a tape block + - Add rudimentary support for saving AFS acls. + - Send level before the fileset in estimate_cmd() like in backup() + +Bugs fixed/closed since last release: +1859 1872 1891 1892 1893 1897 + +;;;; + +Kern;;;11 June 2012 at 20:15 CET +We have released Bacula Version 5.2.8 to Source Forge + +This release is identical to release 5.2.8, with the exception +that the patch: + + "Apply readline without TERM_LIB dependency patch from bug #1871" + +has been removed because it is not portable to all systems or +versions of readline. + +;;;; +Kern;;;10 June 2012 at 12:04 CET + + Release Notes for Bacula 5.2.8 + + Bacula code: Total files = 1,110 Total lines = 231,439 (Using SLOCCount) + +General: +-------- +The 5.2.8 version is a bug fix release. + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.7, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.8 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Changes since 5.2.7: +-------------------- + +10Jun12 + - Apply readline without TERM_LIB dependency patch from bug #1871 + - Ensure cancel status is set for canceled duplicate jobs -- partial fix for + bug #1851 + - Enhance Win32 error messages -- fixes bug #1879 + - Apply patch from bug 1876 to permit dot commands in a runscript + - Remove config variables from examples/devices fixes bug #1884 + - Add more debug code for bug 1885 + - Add more debug to block.c when read rejected + - Fail job that attempts to read from closed device -- reported in bug 1885 but + not fixed + - Detect Python 2.7 fixes bug #1888 + - Apply systemd patch from bug #1886 + - Rewrite some SD subroutines as class members + - New overhaul of xattr code. + - Use IPPROTO_TCP for SOL_TCP if SOL_TCP isn't defined. + - Fix compiler warning. + +Bugs fixed/closed since last release: +1851 1871 1879 1884 1886 1888 + +;;;; +Kern;;;2 June 2012 at 13:01 CET + + Release Notes for Bacula 5.2.7 + + Bacula code: Total files = 1,110 Total lines = 231,439 (Using SLOCCount) + +General: +-------- +The 5.2.7 version is a significan bug fix release. (The +diff file between 5.2.6 and this one is over 25,000 lines). + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.7, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.3 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Changes since 5.2.6: +-------------------- +02Jun12 + - Correct Qt version needed for bat + - Turn on SD deadlock detection by default + - Fix bug #1880 ltdl.m4 missing for new feature. + - Fix bug #1870 GNU binutils-gold build failure. + - Add new empty netinet/tcp.h so Windows compile works + - Eliminate compiler warning in stored/dircmd.c + - Fix TCP Heartbeat code + - Align conio.c struct at 8 bytes to fix Sparc bus error. + - Fix get_basename() -- rewrite + - Apply James' exchange-fd.c regression fix + - Add first Dutch translation + - Ensure that StorageId is updated after write + - Fix updating of cached slots + error message numbering + - Remove old shared objects before install + - Add missing dummy bVarPrefixLinks variable retrieval. + - Fix mysql config for older versions of mysql+config. + - Fix problem in BVFS with concurrent queries + - Add more checks to bvfs + - Try to use multi-row insert statements for mysql. + - On some platforms intptr_t is in an other include file. + - Add support for fcntl(fd, F_CLOSEM) to close filedescriptors. + - Use closefrom if available instead of trying to close fds ourself. + - Fix class member shadowing. + - Move storages tls initialization to proper place. + - Fix Copyright on init scripts suse, add specific daemon user & group + - Add support for new POSIX getaddrinfo interface. + - Apply MSQL detection fixes from bug #1829 + - Fix bug #1805 cannot specify hostname in ./configure + - Fix restore_job restore keyword + - Allow plugin to backup a directory + - Print seconds in schedule time as noted in bug #1854 + - Fix #1857 about restore cd command that blocks on empty directory + - Backport new StorageId code + - More definitive fix for update slots bug + - Fix #4062 on make_catalog_backup.pl script when user/password are empty + - Fix bug #1853: bacula-sd dead but pid file exists. + - Fix bug #1841 estimate command level=differential goes wrong in accurate + mode. + - Fix bug #1842 Incorrect port number reported + - Fix bug #1848 bsock.c:335 Socket has errors=1 on call to client + - Force the re-initialization of BaseJobOpts, AccurateOpts and VerifyOpts + between two Include{} + +Bugs fixed/closed since last release: +1805 1829 1841 1842 1848 1853 1854 1857 1870 1874 1880 4062 + + +;;;; Kern;;;22 February 2012 at 15:03 CET We have released Bacula Version 5.2.6 to Source Forge diff --git a/docs/home-page/fr/news.txt b/docs/home-page/fr/news.txt index 6e82c8d1..e0df4e40 100644 --- a/docs/home-page/fr/news.txt +++ b/docs/home-page/fr/news.txt @@ -1,3 +1,529 @@ +Kern;;;25 November 2013 at 17:53 CET +Bacula Status Report + +Hello, + +I would like to speak to you about the following points: + +1. The rumors of the death of Bacula (the Community version) +2. The Bareos fork of Bacula +3. Bacula Systems and the FSFE (Free Software Foundation Europe) +4. The future of Bacula (the Community version) + +1. The rumors of the death of Bacula (the Community version): + +I borrow words from a quote of Mark Twain : The rumors of the death +of Bacula are highly exaggerated! + +I began working on Bacula 14 years ago (in January 2000), and it has +been Open Source from the time it was publicly released in April +2002, and it will remain Open Source. I have been and am fully +devoted to Open Source, and in particular to Bacula, which is like +my “baby”. So to hear rumors that Bacula is dead or that I have +withheld commits because they are Enterprise features is shocking +and hurtful to me as well as not true. + +I did inform the Bacula Community several years ago that my personal +participation in Bacula would decrease a bit for several years to +allow me to focus more on getting Bacula Systems started. In my +opinion, that has not been a serious disadvantage for the Bacula +project since Bacula Systems over that period has contributed far +more code to Bacula than I could have alone over the same period, +and as you will see a bit later in this status report, Bacula +Systems contributions are absolutely guaranteed to continue in the +long run, and even increase. + +2. The Bareos fork of Bacula: + +The Bacula repository has been on “hold” since our last release +in early February, because on 27 February 2013, I learned that there +was a fork of Bacula made by a former “consultant” of Bacula +Systems with a former reseller of Bacula Systems. Unfortunately, +despite the fact that Bareos hired one of the best German Open +Source lawyers , there were a number of serious copyright violations +with their code. Since the Bacula code is copyrighted by the Free +Software Foundation Europe (FSFE), I provided the technical support, +and the FSFE worked with Bareos to clean up their copyright +violations. That took a significant amount of time (many months), +and the Bareos code though significantly improved is still not +totally free of copyright infringements. I won't go into the +details here as all of you may not be interested, but will have much +more to say about Bareos in later blogs, and when the blog is setup +I will let you know. + +I find Bareos an unusual fork, because it wasn't done in what I +believe to be the normal Open Source way. Normally a fork is made +when a project is blocked or has serious disagreements with the +users. Its also normally done in open communication rather than +underhanded or in secret. In the case of Bacula, though some of the +development slowed down (I will go into this in detail later), it +certainly was by no means stopped. To complain about active +development in Bacula Systems, is, in my opinion, incorrect first +because adding features to the Enterprise version costs Bacula +Systems a lot mostly in salaries yet takes absolutely nothing from +Bacula. In fact, when you read the next section, you will see that +the more that Bacula Systems develops, the more features that Bacula +over time will have. Going back to what I find abnormal about the +Bareos fork is that they claim that they spent three years +developing a lot of new features, thus they are more feature rich +than Bacula. Yes, for the moment, they have a few features that +Bacula does not yet have, but not for long, and more importantly +over the three years of development of those features they never +offered these new features to Bacula nor to any Open Source project. +Instead they were developed in secret. I find that a very strange +behavior for a self-proclaimed Open Source company (actually, they +are “forced” to be Open Source because of the Bacula AGPLv3 +license). So as Bacula contributors and users, you would be within +your rights to feel very upset with Bareos, because they never +offered you the code they developed. + +I assure that I will do all in my power to ensure that any +worthwhile features that Bareos implements will be implemented in +Bacula, and most likely better integrated and more robust, and where +possible with even more functionality and growth potential. + + +3. Bacula Systems and the FSFE: + +If you have been a long time Bacula user you may recall that I +discussed the possibility in 2006-2007 of creating a company, now +called Bacula Systems, to ensure the continuation of Bacula when I +will no longer be able to personally contribute – say in 10 or 20 +years, as well as to provide the financial means to add high-end +features to Bacula (a fibre channel network costs about $50K to +setup). Much to my surprise 95% or more of the responses I got were +very positive. Bacula Systems was created in July 2008, and for the +first two years, the Enterprise code base and the Community code +base were identical. Unfortunately, that didn't work financially +for Bacula Systems. Companies willing to pay, were willing to pay +for features and support but not support alone, so Bacula Systems +embarked on development to continue maintenance and improvement of +Bacula while at the same time creating mostly plugins to add +differentiation to the Enterprise version. + +Now this may not sound very Open Source to you, and I understand, +because I feel the same way. Were it at all possible, I would give +you all of Bacula Systems code, unfortunately, that is not +economically feasible at the current time, and yet without Bacula +Systems, I fear the Bacula project will die or worse yet fall into +the hands of someone incapable of maintaining the high quality we +have created. + +While I was consulting with the Free Software Foundation Europe +(FSFE) on the Bareos copyright violations, Bacula Systems and I +began discussions with the FSFE on how to guarantee the long term +survival of Bacula. These discussions, extremely positive on both +sides and all points, recently lead to a formal written agreement +between myself, Bacula Systems, and the FSFE. There are a number of +points in the agreement, but probably the most important of all is +that Bacula Systems has now put in writing that it is an Open Source +company (at its heart), as it has always proclaimed, and will +contribute all the Enterprise code it creates to the Bacula +Community code base within at most a 5 year period. One exception +is that Bacula Systems is legally unable to contribute certain code +encumbered by third party proprietary license. The 5 year delay +gives Bacula Systems the chance to develop Enterprise features that +differentiate it, but ensures the continual growth of the Bacula +Community code. This model can possibly be used across the industry +to ensure the future of open source software in an environment where +development costs, particularly for hardware to do testing, are +prohibitive to the standard models of today. + +5. The future of Bacula (the Community version): + +If you have read section 4 above, hopefully if you were not already +convinced that Bacula is alive that you can now see that it will +have a long and successful future ahead of it. If you have any +doubts, please do not hesitate to either send me an email on the +bacula-users list or directly to me (if you want it private). +Hopefully, by mid-December I will have a blog setup (need a major +upgrade of bacula.org to do so), and I will then fill you in on the +details of the Bareos fork as well as more details on what next to +expect in Bacula. + +Thank you for contributing to and/or using Bacula ... + +Best regards, +Kern +;;;; + + +Kern;;;20 February 2013 at 13:43 CET +We have released Bacula Version 5.2.13 to Source Forge + +Version 5.2.13 version has 20 bug fixes since the last +release. Some of the bug fixes are important. + +Important changes since last release: + +19Feb13 + - Fix build/configure problems with bpluginfo.c + - Refactor lock_volumes so most lock a vol rather than globally + - Add virtualfull-extreme test + - Apply patch for chio-changer-openbsd from bug #1984 -- Implements listall + - Add bat Mac patch from bug #1953 + - Fix bug #1812 cannot run Copy/Migrate jobs from bat + - Fix loading of bat translations. Patch from bug #1890 + - Fix text input in bat. Fixe bug #1965 + - Remove Device from show command as it is not used or updated + - Fix #1982 update enabled keyword in help command + - Fix update_postgresql_tables 10 to 11 + - Clean after building doc binaries requested by Willem vd Akker for Debian + build + - Require Qt version 4.8.4 + - Fix bug #1955 that OK to run? does not retry on bad response + - Backport new lock calls + debug for SD + - Fix bug #1975 new label ignored when first is duplicate. + - Pull src/lib changes from master + - Pull SD files from master + - Turn off prune-test + - Add logdir and bsrdir to regression config + - Add bsrdir and logdir to regress-config + - Changed default directories from /var/bacula to /opt/bacula + - Make spool size max message more explicit + - Display more info when maximum spool size is reached + - Fix TERM defs for Solaris 11 in conio.c + - Allow group to cd to sysconfdir + - Make sysconfdir belong to Bacula user/group and exclude others + - Fix Virtual Full file close bug -- race condition + - Enhance mount message to include read/append + - Improve vol_mgr swap messages + - Enhance output of Using Device to include for read/write + - Add jobid to Pmsg output + - Fix bug #1959 input validation on delete of jobs. + - Fix bug #1956 Authorization Errors false + - Add workaround for #5507 where autoprune and reload are in deadlock + - Allow conf file quoted strings to be used in a list + - Make bfgets handle very long lines + - Avoid seg fault by checking for NULL client -- reported by Arno + - Final fix of bug #1943 + - Fix bug #1948 MailOnSuccess get executed on error. + - Fix missing index on Media table + - Fix bug #1943 no message storage on closed database connection. + - Fix bug #1946 Fix problem with MySQL with big Base jobs. + - Reduce wait time in regression RunScripts + - Implement a timeout on regression tests of 15 minutes + - Fix #5346 .bvfs_lsfiles and .bvfs_restore to handle deleted files + - Fix spooldata, accurate and ingnoreduplicate run argument + - Fix messages segfault + - Implement feature request #1939 + - fix #1938 about PATH_MAX on hurd + - fix #1938 about MAXPATHLEN on hurd + - Fix bug #1937 OpenBSD autochanger example script + - Fix #4996 about MaxRunTime canceling the job too early + - Fix bug #1934 Wrong values at media list in BAT + - Fix race condition in close_msg that causes seg fault + - Fix double unlink + - More tests for freeing NULL pointers + - Add MaximumConcurrentReadJobs directive to Storage resource + - Add ujobid to .bvfs_get_jobids + - Avoid problem when stripping a relative path + - Fix #4513 about HIDDEN attribute set to parent directory during restore + +Bugs fixed/closed since last release: +1812 1890 1934 1937 1938 1939 1943 1946 1948 1953 1955 1956 1959 1965 1975 +1982 1984 4513 4996 5346 5507 + + +;; +Kern;;;11 September 2012 at 13:43 CET +We have released Bacula Version 5.2.11 to Source Forge + +Version 5.2.11 version is a bug fix release. + + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.11 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Important changes since last release: + - A new bpluginfo utility tool + - Make dump_resource respect console ACL's (security fix) + - Add chio-changer-openbsd submitted in bug #1903 + - Important Copy job bug fixed. + +10Sep12 + - Add JobId to SD debug output + - Commit batch session every 800,000 files + - Simplify safer delete code using a single regex + - Clean up error message generation during update attributes -- should fix bug + #1823 + - Eliminate heartbeat error messages. Fixes bug #1925 + - Apply patch from bug #1911 to implement --with-logdir + - Add chio-changer-openbsd submitted in bug #1903 + - mysql database install/update scripts patch. Fixes bug #1901 + - Delegate uid/gid changing to systemd. Fixes bug #1905 + - Fix bug #1906 + - Fix manpage errors, fixes bug #1907 + - Final fix for #1859 missed one query. + - bpluginfo - Utility tool display various information + - Backport more master code + - Fix Makefile.in so that testfind builds with acl dependency + - Make dump_resource respect console ACL's (security fix) + - backport code from master + - Fix #1923 about MySQL 4 support for BVFS + - Fix #1902 about bad status slots output + - Fix Sigma Copy bug #4377 + - Fix bsmtp + - Reset bsmtp to only send to IPv4 mailhosts. + +Bugs fixed/closed since last release: +1823 1859 1901 1902 1903 1905 1906 1907 1911 1923 1925 4377 +;;;; + +Kern;;;28 June 2012 at 20:55 CET +We have released Bacula Version 5.2.10 to Source Forge + +General: +-------- +The 5.2.10 version is a significant bug fix release including +our first cut of AFS support. + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.9, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.9 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Major changes since 5.2.9: +- IPv6 networking should work correctly now +- SQL Counter problems resolved +- Bat hanging after running a number of jobs +- dbcheck rw_writelock problem +- Windows VSS restore bug + +New Feature: +- Add rudimentary support for backing up AFS + +28Jun12 + - Add extra test for unsupported Protocol Families. + - Fix inet_pton call. + - Real fix of bug #1897 5.2.9 breaks IPv6 connectivity + - Fix bug #1891 wrong daemon name printed with -? option + - Fix bug #1859 INSERT INTO Counters fails + - First try at fixing bug #1897 5.2.9 breaks IPv6 connectivity + - Convert write_rec_to_block() to a state machine + - Fix bat from eating all Dir connections -- fixes bug #1872 + - Back port fixes from BEE to fix Windows restore bug #1892 + - Add block checksum flag to version.h + - Add block checksum debug code + - Fix bug #1893 dbcheck -B gives rwl_writelock error. + - Remove restriction of network buffer size being multiple of a tape block + - Add rudimentary support for saving AFS acls. + - Send level before the fileset in estimate_cmd() like in backup() + +Bugs fixed/closed since last release: +1859 1872 1891 1892 1893 1897 + +;;;; + +Kern;;;11 June 2012 at 20:15 CET +We have released Bacula Version 5.2.8 to Source Forge + +This release is identical to release 5.2.8, with the exception +that the patch: + + "Apply readline without TERM_LIB dependency patch from bug #1871" + +has been removed because it is not portable to all systems or +versions of readline. + +;;;; +Kern;;;10 June 2012 at 12:04 CET + + Release Notes for Bacula 5.2.8 + + Bacula code: Total files = 1,110 Total lines = 231,439 (Using SLOCCount) + +General: +-------- +The 5.2.8 version is a bug fix release. + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.7, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.8 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Changes since 5.2.7: +-------------------- + +10Jun12 + - Apply readline without TERM_LIB dependency patch from bug #1871 + - Ensure cancel status is set for canceled duplicate jobs -- partial fix for + bug #1851 + - Enhance Win32 error messages -- fixes bug #1879 + - Apply patch from bug 1876 to permit dot commands in a runscript + - Remove config variables from examples/devices fixes bug #1884 + - Add more debug code for bug 1885 + - Add more debug to block.c when read rejected + - Fail job that attempts to read from closed device -- reported in bug 1885 but + not fixed + - Detect Python 2.7 fixes bug #1888 + - Apply systemd patch from bug #1886 + - Rewrite some SD subroutines as class members + - New overhaul of xattr code. + - Use IPPROTO_TCP for SOL_TCP if SOL_TCP isn't defined. + - Fix compiler warning. + +Bugs fixed/closed since last release: +1851 1871 1879 1884 1886 1888 + +;;;; +Kern;;;2 June 2012 at 13:01 CET + + Release Notes for Bacula 5.2.7 + + Bacula code: Total files = 1,110 Total lines = 231,439 (Using SLOCCount) + +General: +-------- +The 5.2.7 version is a significan bug fix release. (The +diff file between 5.2.6 and this one is over 25,000 lines). + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +Please note that with this version of Bacula, during the install +of the shared libraries from the src/cats, src/lib, and src/findlib +directories, any old Bacula libraries in the installation +destination directory will be removed. This should not cause +any problems and at least keeps the installation library cleaner. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +For packagers, if you change options, naming, and the way +we link our shared object files, as at least one of you does, +you are creating a situation where the user may not be able +to run multiple versions of Bacula on the same machine, which +is often very useful, and in addition, you create a configuration +that the project cannot properly support. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! +If you are upgrading from a version older than 5.2.7, please +read all the old ReleaseNotes of all versions released +after the version you are upgrading from. There area +many important notices sumarized here, and we do not +repeat them each bug fix release. +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + +Compatibility: +-------------- + As always, both the Director and Storage daemon must be upgraded at + the same time. + + Older 5.0.x and 3.0.x File Daemons are compatible with the 5.2.3 + Director and Storage daemons. There should be no need to upgrade older File + Daemons. + +Changes since 5.2.6: +-------------------- +02Jun12 + - Correct Qt version needed for bat + - Turn on SD deadlock detection by default + - Fix bug #1880 ltdl.m4 missing for new feature. + - Fix bug #1870 GNU binutils-gold build failure. + - Add new empty netinet/tcp.h so Windows compile works + - Eliminate compiler warning in stored/dircmd.c + - Fix TCP Heartbeat code + - Align conio.c struct at 8 bytes to fix Sparc bus error. + - Fix get_basename() -- rewrite + - Apply James' exchange-fd.c regression fix + - Add first Dutch translation + - Ensure that StorageId is updated after write + - Fix updating of cached slots + error message numbering + - Remove old shared objects before install + - Add missing dummy bVarPrefixLinks variable retrieval. + - Fix mysql config for older versions of mysql+config. + - Fix problem in BVFS with concurrent queries + - Add more checks to bvfs + - Try to use multi-row insert statements for mysql. + - On some platforms intptr_t is in an other include file. + - Add support for fcntl(fd, F_CLOSEM) to close filedescriptors. + - Use closefrom if available instead of trying to close fds ourself. + - Fix class member shadowing. + - Move storages tls initialization to proper place. + - Fix Copyright on init scripts suse, add specific daemon user & group + - Add support for new POSIX getaddrinfo interface. + - Apply MSQL detection fixes from bug #1829 + - Fix bug #1805 cannot specify hostname in ./configure + - Fix restore_job restore keyword + - Allow plugin to backup a directory + - Print seconds in schedule time as noted in bug #1854 + - Fix #1857 about restore cd command that blocks on empty directory + - Backport new StorageId code + - More definitive fix for update slots bug + - Fix #4062 on make_catalog_backup.pl script when user/password are empty + - Fix bug #1853: bacula-sd dead but pid file exists. + - Fix bug #1841 estimate command level=differential goes wrong in accurate + mode. + - Fix bug #1842 Incorrect port number reported + - Fix bug #1848 bsock.c:335 Socket has errors=1 on call to client + - Force the re-initialization of BaseJobOpts, AccurateOpts and VerifyOpts + between two Include{} + +Bugs fixed/closed since last release: +1805 1829 1841 1842 1848 1853 1854 1857 1870 1874 1880 4062 + + +;;;; Kern;;;22 February 2012 at 15:03 CET We have released Bacula Version 5.2.6 to Source Forge