------------------------------------------------------------------- Tue Feb 24 12:56:06 CET 2009 - agruen@suse.de - Include patch headers in reject files so that they form proper patches themselves. - Rewrite the unified reject files patch; this is much cleaner now. - Add an improved strategy for locating merges. ------------------------------------------------------------------- Tue Feb 3 06:10:49 CET 2009 - agruen@suse.de - Implement diff3-style merges (including several fixes and improvements). ------------------------------------------------------------------- Wed Apr 23 10:53:19 CEST 2008 - agruen@suse.de - remember-backup-files.diff: Fix bug when a file is touched by the same patch more than twice. Move the test cases from the patch header into separate files. - patch-2.5.9-cat_if_device.diff: No need to remember the device we write to as a backup file. ------------------------------------------------------------------- Fri Nov 2 01:57:03 CET 2007 - agruen@suse.de - Patch fails to apply hunks with asymmetric context correctly. ------------------------------------------------------------------- Wed Jan 25 21:30:47 CET 2006 - mls@suse.de - converted neededforbuild to BuildRequires ------------------------------------------------------------------- Thu Jan 12 07:52:31 CET 2006 - agruen@suse.de - remember-backup-files.diff: Fix case where a patch modifies a read-only file more than once while --backup is used (test case in the patch header). ------------------------------------------------------------------- Thu Aug 18 18:09:59 CEST 2005 - mmj@suse.de - --reject-unified is called --unified-reject-files [#105151] ------------------------------------------------------------------- Wed Jun 29 13:50:06 CEST 2005 - mmj@suse.de - Don't compile with -f-signed-char [#93883] - Don't strip explicitly ------------------------------------------------------------------- Tue Jan 4 21:11:54 CET 2005 - mmj@suse.de - Add patch to make patch able to write reject files to devices, named pipes and sockets [#45794] ------------------------------------------------------------------- Sat Jan 25 02:30:00 CET 2004 - agruen@suse.de - Add --reject-unified option to produce unified reject files: Before, unified reject files were produced if the patch itself was unified; this could sometimes have been annoying. - If a patch with C function names (diff -p) has rejects, include the function names in the reject files. ------------------------------------------------------------------- Sun Jan 11 11:17:55 CET 2004 - adrian@suse.de - add %defattr ------------------------------------------------------------------- Tue Oct 7 15:03:22 CEST 2003 - agruen@suse.de - remember-backup-files: Also include the file timestamps in the hash table for non-POSIX-compliant systems that don't guarantee that i_dev + i_ino uniquely identifies a file. - #32031: Create --global-reject-file even if --dry-run is specified. Add a missing '\n'. ------------------------------------------------------------------- Wed Jul 16 15:21:44 CEST 2003 - agruen@suse.de - Replace trailing-cr-fix.diff with official upstream version that fixes this bug differently. ------------------------------------------------------------------- Mon Jun 30 13:30:33 CEST 2003 - agruen@suse.de - trailing-cr-fix.diff: Fix a bug in carriage return detection (DOS files) in the unified diff path. - Disable patch that adds comments to C preprocessor style merges, as specified by POSIX.1-2001. ------------------------------------------------------------------- Wed Jun 25 16:46:41 CEST 2003 - agruen@suse.de - Fix bug introduced on June 5 that broke remember-backup-files. ------------------------------------------------------------------- Tue Jun 24 17:35:09 CEST 2003 - agruen@suse.de - smart-reject-file-format.diff: Work around a special case in which patches are not terminated with a '^' in the internal representation. The resulting error message was "internal error in abort_hunk". ------------------------------------------------------------------- Thu Jun 5 09:38:17 CEST 2003 - agruen@suse.de - Temporary reject file logic: Ooops, now must only close the temporary reject file after processing all patches, instead of after each patch. ------------------------------------------------------------------- Thu Jun 5 01:00:38 CEST 2003 - agruen@suse.de - Upgrade to 2.5.9: Several fixes, obsoletes rename-same-file.patch. - Fix and adapt global-reject-file patch: + The global reject file included corrupted headers for each rejected hunk, instead of one header for each file with rejects. + Rename --global-reject to --global-reject-file. + Simplify temporary reject file logic. + Adapt to unified-reject-files patch (which has different headers). + Add entry in man page. - Fix a bug in smart-reject-file-format: Files that are created are identified by `-0,0'; before the patch was generating `-1,0' in reject files. - Switch remember-backup-files.diff over to use gnulib's hash tables instead of glibc's binary trees, requested from upstream to ensure greater portability. - Add /* SYM */ comment to #endif lines for patch -D SYM, too. ------------------------------------------------------------------- Wed Apr 9 19:33:59 CEST 2003 - agruen@suse.de - Fix another bug with hard links and backup file generation. - Fix backup file generation if the same file appears in the patch more than once. ------------------------------------------------------------------- Wed Mar 26 14:00:55 CET 2003 - mmj@suse.de - Update to 2.5.8: + Bugfixes + patch -D now outputs preprocessor lines without comments, as required by POSIX 1003.1-2001 + File names in context patches may now contain spaces, so long as the context patch headers use a tab to separate the file name from the time stamp + Perforce is now supported + Patch lines beginning with "#" are comments and are ignored ------------------------------------------------------------------- Wed Jan 15 16:50:57 CET 2003 - agruen@suse.de - Fix a bug with hardlinks (see rename-same-file.patch) ------------------------------------------------------------------- Tue Sep 17 17:34:28 CEST 2002 - ro@suse.de - removed bogus self-provides ------------------------------------------------------------------- Wed Feb 6 12:05:34 CET 2002 - coolo@suse.de - called suse_update_config ------------------------------------------------------------------- Mon Jun 25 12:46:18 CEST 2001 - uli@suse.de - added patch adding "--global-reject" option by ak@suse.de - bzipped tarball ------------------------------------------------------------------- Tue Mar 6 20:36:46 CET 2001 - bk@suse.de - update to 2.5.4, added 2 patches from PLD and use buildroot ------------------------------------------------------------------- Wed Nov 8 17:41:23 CET 2000 - uli@suse.de - now builds with -D_GNU_SOURCE, should avoid miscompilation that breaks LFS support - added fix for offset output by Alessandro Rubini - added fix and enhancement for --ifdef by Pete Buechler ------------------------------------------------------------------- Sun Feb 27 20:16:11 CET 2000 - @suse.de - added missing CFLAGS quotes. ------------------------------------------------------------------- Sun Feb 27 19:07:58 CET 2000 - bk@suse.de - added PPC fixes by Uli back again (-fsigned-char, CPPFLAGS) ------------------------------------------------------------------- Fri Feb 25 17:44:41 CET 2000 - schwab@suse.de - Specfile cleanup, get rid of Makefile.Linux - /usr/man -> /usr/share/man - Add group tag. ------------------------------------------------------------------- Mon Sep 13 17:23:57 CEST 1999 - bs@suse.de - ran old prepare_spec on spec file to switch to new prepare_spec. ------------------------------------------------------------------- Wed Aug 18 12:13:35 MEST 1999 - uli@suse.de - fixed for PPC (-fsigned-char, CPPFLAGS) ------------------------------------------------------------------- Tue Mar 16 14:58:42 MET 1999 - ro@suse.de - update to 2.5.3 using diff from jurix ------------------------------------------------------------------- Wed Jan 13 15:45:28 MET 1999 - bs@suse.de - applied patch from Egbert Eich (problems with non existing files fixed) ------------------------------------------------------------------- Thu Nov 5 00:19:13 MET 1998 - ro@suse.de - use libc's basename() for glibc