CHANGES

   1 $Id: CHANGES,v 1.26 1999/12/17 10:14:27 tiniou Exp $
   2
   3 Changes between versions 0.4b11 and 0.4b12 (released ????????????????)
   4 ======================================================================
   5
   6 1.      Small fix in the dump man page. Thanks to Thorsten Kukuk
   7         <kukuk@suse.de> for submitting the patch.
   8
   9 2.      Fix for the exit code when using the size estimate option of
  10         dump. Thanks to Matti Taskinen <mkt@rni.helsinki.fi> for
  11         submitting the patch.
  12
  13 Changes between versions 0.4b10 and 0.4b11 (released December 5, 1999)
  14 ======================================================================
  15
  16 1.      Added a '--enable-kerberos' to configure.
  17
  18 2.      Added a 'S' option to dump which determines the amount of space
  19         that is needed to perform the dump without actually doint it, similar
  20         to the Sun's ufsdump 'S' option. Patch contributed by Rob Cermak
  21         <cermak@ahab.rutgers.edu>.
  22
  23 3.      Added a 'M' multi-volume option to dump and restore which enables
  24         dumping to multiple files (useful when dumping to an ext2
  25         partition to make several dump files in order to bypass the 2GB
  26         file size limitation). The 'f' argument is treated as a prefix and
  27         the output files will be named <prefix>001, <prefix>002 etc. With
  28         the 'M' flag, restore automatically selects the right file without
  29         asking to enter a new tape each time.
  30
  31 4.      Fixed a memory leak which was causing dump to grow very big
  32         (270MB when dumping a 10GB filesystem...). Thanks to Jason
  33         Fearon <jasonf@netrider.org.au> for reporting the bug.
  34
  35 Changes between versions 0.4b9 and 0.4b10 (released November 21, 1999)
  36 ======================================================================
  37
  38 1.      Make configure test if the system glob routines support
  39         extended syntax (ALTDIRFUNC). If not, use the internal glob
  40         routines instead of system ones. Thanks to Bernhard Sadlowski
  41         <sadlowsk@Mathematik.Uni-Bielefeld.DE> for reporting the bug
  42         and helping me resolve this and other minor libc5 compiling
  43         glitches.
  44
  45 2.      Fix a problem when dumping a ext2fs with the 'filetype'
  46         feature enabled. Thanks to Patrick J. LoPresti
  47         <patl@cag.lcs.mit.edu> for reporting the bug and to
  48         Theodore Y. Ts'o <tytso@mit.edu> for providing the patch.
  49
  50 3.      Made the nodump flag work on directories. A directory which
  51         has the nodump flag gets never dumped, regardless of its
  52         contents.
  53
  54 4.      Integrate a patch from Jeremy Fitzhardinge <jeremy@goop.org>
  55         which allows dump on an active ext3 filesystem. However, this
  56         is a "quick and dirty" patch which enables backup of an ext3
  57         filesystem through the ext2 compatibility (by ignoring the
  58         NEEDS_RECOVERY bit). The journal file is not recognized and
  59         it is dumped (it should not).
  60
  61 5.      Test the superblock compatibility flags when dumping, in order
  62         to be sure that we know how to deal with specific features.
  63
  64 Changes between versions 0.4b8 and 0.4b9 (released November 5, 1999)
  65 ====================================================================
  66
  67 1.      Use lchown instead of chown, fixing a possible security problem
  68         when restoring symlinks (a malicious user could use this
  69         to deliberately corrupt the ownership of important system files).
  70         Thanks to Chris Siebenmann <cks@utcc.utoronto.ca> for detecting
  71         this and providing the patch.
  72
  73 Changes between versions 0.4b7 and 0.4b8 (released November 3, 1999)
  74 ====================================================================
  75
  76 1.      Put dump sources under CVS, added Id tags in all files so
  77         one can use 'ident' on binary files.
  78
  79 2.      Added the dump/restore version in the usage text so one can
  80         easily verify the version he is using.
  81
  82 3.      Small patch from Nuno Oliveira <nuno@eq.uc.pt> which fixes
  83         a va_start/va_end problem on linux-ppc (always call va_start
  84         va_end in pairs each time we use a vararg function).
  85
  86 4.      Added again the DT_* constants because old libc does not
  87         contain them :(. Thanks to Eric Maisonobe <virnet@nat.fr>
  88         for submitting the bug report.
  89
  90 5.      Use ext2fs_llseek instead of llseek. With recent e2fsprogs
  91         this should enable dumping big (huge) filesystems.
  92
  93 6.      Added the RSH environment variable in order to be able to
  94         use a rsh replacement like ssh when doing remote backups (and
  95         bypass the security limitations of rcmd). Now you can do remote
  96         backups without being root (or making dump setuid root).
  97
  98 7.      Modified again the way dumpdates works. For incremental dumps,
  99         we need to read dumpdates even if we are not using 'u' option.
 100         Thanks to Bdale Garbee <bdale@gag.com> for his ideas on how
 101         this should work.
 102
 103 Changes between versions 0.4b6 and 0.4b7 (released October 8, 1999)
 104 ===================================================================
 105
 106 1.      Removed the 'k' flag from the restore 'about' text if kerberos
 107         was not compiled in.
 108
 109 2.      Prototyped (f)setflags from e2fsprogs and corrected the calls
 110         to them (fsetflags takes a char*, setflags an open fd!).
 111
 112 3.      (f)setflags is called only if the flags aren't empty. If the
 113         file is a special file, a warning is printed, because changing
 114         flags implies opening the device. Normally, a special file
 115         should not have any flag... (Debian bug #29775, patch provided
 116         by Abhijit Dasgupta <abhijit@ans.net>).
 117
 118 4.      Made possible to dump a file system not mentioned in /etc/fstab.
 119         (Debian bug #11904, patch provided by Eirik Fuller <eirik@netcom.com>).
 120
 121 5.      Changed the default behaviour to not create dumpdates
 122         unless 'u' option is specified. Removed the old "debian-patch"
 123         which provided the same thing. (Debian bug #38136, #33818).
 124
 125 6.      Removed all those dump*announce, since they were getting old...
 126
 127 7.      Added warning messages if dumpdates does not exist and
 128         when an inferior level dump does not exist (except for a level 0
 129         dump).
 130
 131 8.      Debugged the glob calls in interactive mode: restore used a
 132         dirent struct which was different from the /usr/include/dirent.h
 133         one (this used to work, is it a glibc2 change?), so none of the
 134         compat glob (which used /usr/include/dirent.h) or the system glob
 135         worked. Restore use now the system dirent (and the system
 136         DT_* constants), which are compatible with BSD ones.
 137
 138 9.      Added a configure flag (--with-dumpdatespath) to specify
 139         the location of dumpdates file. By default, it is
 140         /etc/dumpdates.
 141
 142 10.     Added the "AUTHOR" and "AVAILABILITY" sections and
 143         included the current date/version in man pages.
 144
 145 11.     Corrected the estimation of remaining time when
 146         the operator doesn't change the tapes quickly enough. This
 147         was an old bug, I thought I corrected it, and discovered
 148         that in fact it was corrected in two different places, so
 149         the results canceled each other...
 150
 151 Changes between versions 0.4b5 and 0.4b6 (released October 1, 1999)
 152 ===================================================================
 153
 154 1.      Integrated multiple patches from RedHat, Debian and SuSE:
 155
 156         - tweak dump/itime.c to not try to read dumpdates if the 'u' option
 157           isn't specified.
 158         - several fixes in the man pages.
 159         - update the default tape device to /dev/st0.
 160         - many updates for Linux Alpha (byte ordering, size_t etc).
 161         - buffer overruns.
 162         - use environment variable for TMPDIR (instead of /tmp).
 163         - use sigjmp_buf instead of jmp_buf (RedHat bug #3260).
 164         - workaround egcs bug (RedHat bugs #4281 and #2989).
 165         - wire $(OPT) throughout Makefile's.
 166
 167 2.      Upgrade the dump revision to 1, making possible to dump filesystems
 168         made with e2fsprogs-1.15 or newer. Nothing seems to break...
 169
 170 3.      Fix some compile warnings, prototype all functions.
 171
 172 4.      Use glibc err/glob instead of internal compatibility
 173         routines (only if available).
 174
 175 5.      Fix a compile error on Linux 2.2.7 / libc5 (5.4.44) (patch provided
 176         by Bernhard Sadlowski <sadlowsk@mathematik.uni-bielefeld.de>).
 177
 178 Changes between versions 0.4b4 and 0.4b5 (released September 22, 1999)
 179 ======================================================================
 180
 181 1.      Integrated the changes from FreeBSD-3.1-RELEASE
 182         (mostly bug fixes, buffer overruns, dump has now an "automatic
 183         tape length calculation" flag, dump/restore can use kerberos now
 184         (this is NOT tested), use environment variables for TAPE and
 185         RMT etc.).
 186
 187 2.      Integrated three RedHat patches ("glibc", "kernel" and "bread" patches)
 188
 189 3.      Corrected a bug in restore when using 'C' option with multi-volumes
 190         tapes (files splited accros two tapes give "size changed" errors
 191         when comparing).
 192
 193 4.      Corrected the long standing bug when dumping multiple tapes.
 194         This works for me, needs further testing.
 195
 196 Changes between versions 0.4b3 and 0.4b4 (released January 17, 1997)
 197 ====================================================================
 198
 199 1.      Dump now runs correctly on kernels 2.1.x
 200         Fix made by Gerald Peters <gapeters@worldnet.att.net>
 201
 202 Changes between versions 0.4b2 and 0.4b3
 203 ========================================
 204
 205 1.      Use realpath() if available
 206
 207 2.      Report statistics
 208
 209 Changes between versions 0.4b1 and 0.4b2
 210 ========================================
 211
 212 1.      Fixed the bug fix from Greg Lutz (I had made a mistake when integrating
 213         the patch)
 214
 215 2.      Fixed restore to make it able to read FreeBSD 2.x dumps again
 216
 217 3.      Fixed configure.in to correctly handle --enable-rmt
 218
 219 Changes between versions 0.3 and 0.4b1
 220 ======================================
 221
 222 1.      Integrated the changes from 4.4BSD-Lite2
 223
 224 2.      Integrated the patches from Debian and Red Hat
 225
 226 3.      Portability changes: use the __u32, __u16, __s32, and __s16 types
 227
 228 4.      Changed dump to use the Ext2fs library to get block addresses.  This
 229         should solve the endianness problem on SparcLinux.
 230
 231 5.      Created a configure.in file (shamelessly stolen from the e2fsprogs
 232         distribution's one) to use autoconf
 233
 234 6.      Fixed a few minor bugs
 235
 236 Changes between versions 0.2e and 0.2f
 237 ======================================
 238
 239 1.      Added the creation of named pipes (fifos) in restore.
 240
 241 2.      Added the -N flag in the restore manual page.
 242
 243 3.      Added the file kernel.patch which contains the llseek() optimization
 244         patch for 1.2.x kernels.
 245
 246 4.      Fixed a bug in the restoration of symbolic links: owner and group were
 247         not restored.
 248
 249 5.      Integrated some changes from FreeBSD 2.2.
 250
 251 6.      Added a call to ftruncate() after restoring each file to restore
 252         correctly files ending by a hole.
 253
 254 Changes between versions 0.2d and 0.2e
 255 ======================================
 256
 257 1.      Fixed a bug in the "set owner/mode" process.  Every file was restored
 258         with owner = root (0) and group = root/wheel/whatever (0).
 259
 260 Changes between versions 0.2c and 0.2d
 261 ======================================
 262
 263 1.      Dump is now able to backup 2GB+ filesystems.
 264
 265 2.      Dump and restore can now be linked as static binaries.
 266
 267 Changes between versions 0.2b and 0.2c
 268 ======================================
 269
 270 1.      Fixed a bug when dumping ``slow'' (i.e. normal) symbolic links.
 271
 272 Changes between versions 0.2a and 0.2b
 273 ======================================
 274
 275 1.      Really fixed the bug that I should have corrected in 0.2a.
 276
 277 2.      Enabled optimization again.
 278
 279 Changes between versions 0.2 and 0.2a
 280 =====================================
 281
 282 1.      Disabled the optimization during compilation.
 283
 284 Changes between versions 0.1 and 0.2
 285 ====================================
 286
 287 1.      Fixed a bug in fstab.c which caused a null pointer to be stored in
 288         the fs_type field (actually, I modified the file fstab.c to make it
 289         use the mntent functions).
 290
 291 2.      Dump and restore now use a 4.3 BSD compatible dump format.  Backups
 292         made by dump should be readable by the BSD restore and backups made
 293         by the BSD dump should be readable by restore.  Unfortunately, this
 294         means that the dump format has changed between version 0.1 and version
 295         0.2 :-(
 296
 297 3.      Dump is now able to backup a subtree, it is no longer limited to whole
 298         filesystems like the BSD version.
 299
 300 4.      Dump now uses ext2_llseek() so it is able to backup filesystems bigger
 301         than 2 GB.
 302
 303 Changes between versions 0.0 and 0.1
 304 ====================================
 305
 306 1.      Now create links rdump and rrestore during the `make install' step.
 307
 308 2.      Linux port specific bugs added to the manual pages
 309
 310 3.      Incorrect estimation of the number of tapes blocks fixed when doing
 311         incremental backups.
 312
 313 4.      Better ls-like format in restore in interactive mode.