head	1.2;
access;
symbols
	pkgsrc-2014Q1:1.1.0.8
	pkgsrc-2014Q1-base:1.1
	pkgsrc-2013Q4:1.1.0.6
	pkgsrc-2013Q4-base:1.1
	pkgsrc-2013Q3:1.1.0.4
	pkgsrc-2013Q3-base:1.1
	pkgsrc-2013Q2:1.1.0.2
	pkgsrc-2013Q2-base:1.1;
locks; strict;
comment	@// @;


1.2
date	2014.05.14.04.35.20;	author rodent;	state dead;
branches;
next	1.1;
commitid	dFz6ZBNjMGAoFsAx;

1.1
date	2013.06.26.15.49.31;	author joerg;	state Exp;
branches;
next	;
commitid	Miud656iitEgK8Vw;


desc
@@


1.2
log
@Updated to latest release, 2.4.13, as requested by doc/TODO. Removed
patches which were applied upstream or deprecated. USE_TOOLS+=groff to fix
generation of a few .html files on OpenBSD. Depends on libgcrypt instead of
openssl now. From ChangeLog:

from 2.4.12 to 2.4.13
- adding initialization value for two variables to avoid inappropriate
  warning when compiling with -Wall option
- reducing UNIX_PATH_MAX by the system when not defined from 108 to 104
  bytes to accommodate BSD systems
- fixing assignment operator of class criterium that was not returning
  any value as it should
- removing useless boolean expression that always succeeds in logical
  AND expression
- adding support for back-slash of quoting characters in DCF files
- fixed compilation issues with clang / FreeBSD, Thanks to Neil
  Darlow's server ;-)
- fixed compilation warning due to deprecated symbols in libgcrypt
  header files
- replaced gnu make specific rules by legacy ones to avoid automake
  warning about them
- removed old unused stuff from misc sub-directory
- adding warning at compilation time if libgcrypt used is older than
  1.6.0
- adding warning at execution time if hash computation is requested
  with slices greater than 256 Gio and ligbcrypt dynamically or
  statically linked is older than 1.6.0
- adding alternative methods in list_entry API class to return dates as
  number of seconds
- fixed bug in hour-shift (-H option) when comparing dates from an old
  extracted catalogue (archive format 7 or older).
- fixed documentation bug about the meaning of the compression ratio
- fixed a display bug about the "compression flag" wrongly displayed
  for uncompressed files
- fixed unhandled exception when giving non number argument to -1 option

from 2.4.11 to 2.4.12
- for correctness fixed delete vs delete[] on vector of char (not incidence
  reported)
- fixed out of range access in routine used to read very old archive format
- fixed error in logical expression leading a sanity test to be useless
- removed duplicated variable assignment
- updated FAQ
- fixed typo and spelling errors
- fixed bug (reported by Torsten Bronger) in the escape layer leading libdar
  to wrongly reporting a file as corrupted at reading time
- fixed bug in the sparse file detection mechanism that lead the minimum size
  hole detection to become a multiple of the default value or specified one.
  This implied a less efficient reduction of sparse files because smaller
  holes in files were ignored
- fixed and updated man page about --go-into option
- updated full-from-diff target in /etc/darrc default file
- added a debug option in hash_file class (option only used from testing
  tools) to troubleshoot sha1/md5 hash problem on slices larger than
  (2**38)+63 bytes, bug reported by Mike Lenzen and understood by Yuriy
  Kaminskiy at libgcrypt. Note: This bug is still open due to an integer
  overflow in libgcrypt.
- backported from current development code an additional and more simple
  way to read an archive using the libdar API. This API extension is not used
  by dar command-line tools for now.
- Fixing installation of libdar header files on Darwin, where "DARwin" macros
  were not filtered out from the generated libdar header files.
- Fixing self reported bug 'generic_file.cpp line 309' met while comparing an
  archive with a filesystem
- Update code in order to compile with gcc-4.8.2 in g++11 mode (partial
  implementation and adaptation of Fabian Stanke's patch)
- Fixing bug met while performing a verbose archive listing in sequential
  read mode
- Added Ryan Schmidt's Patch to properly display status at end of ./configure
  script under BSD systems (in particular Mac OS X)
- Updating configure.ac script to fix warning reported by autoconf when
  generating the ./configure script
- Addressed portability problem with BSD systems that do not provide a -d
  option to the 'cp' command, preventing proper installation of the Doxygen
  documentation. Fix based on patch provided by Jan Gosmann.

from 2.4.10 to 2.4.11
- Modified behavior of 'dar -h' and 'dar -V', both now return 0 as exist
  status instead of 1 (which means syntax error).
- Fixed bug: -Q is now available with -V under the collapsed form -QV or -VQ
- fixed typo in documentation
- fixed memory leakage met when dar fails a merging operation because the
  resulting archive is specified in an directory that does not exist.
- fixed bug met when isolating a differential backup in sequential read mode
- fixed bug about slice file permission not taking care about umask variable
  when the --hash feature is used.
- fixed performance issue when reading an archive over a pair of piles using
  dar_slave (possibly over ssh) when the archive makes use of escape marks and
  when no encryption is used
- added target "full-from-diff" in /etc/darrc default file
- fixed bug avoiding reading an truncated archive in direct access mode with
  the help of an external catalogue.
- new and better implementation of archive extraction in sequential read mode
- fixing bug (segfault) met when hitting CTRL-C while reading an archive in
  sequential mode
- fixing libdar.pc for pkg-config for the cflags given to external applications
- fixed memory allocation/desallocation mismatches (delete vs delete [] )
  concerning four vector of chars.
- fixed error in logical expression leading a sanity test to be useless
@
text
@$NetBSD: patch-src_testing_test__infinint.cpp,v 1.1 2013/06/26 15:49:31 joerg Exp $

--- src/testing/test_infinint.cpp.orig	2013-06-20 19:28:49.000000000 +0000
+++ src/testing/test_infinint.cpp
@@@@ -83,9 +83,9 @@@@ static void routine1()
     infinint f2 = f1;
     infinint f3 = 0;
 
-    deci d1 = f1;
-    deci d2 = f2;
-    deci d3 = f3;
+    libdar::deci d1 = f1;
+    libdar::deci d2 = f2;
+    libdar::deci d3 = f3;
 
     ui->warning(d1.human() + " " + d2.human() + " " + d3.human());
 
@@@@ -98,7 +98,7 @@@@ static void routine1()
         if(fd >= 0)
         {
             f3 = infinint(*ui, fd);
-            d3 = deci(f3);
+            d3 = libdar::deci(f3);
             ui->warning(d3.human());
         }
         close(fd);
@@@@ -106,33 +106,33 @@@@ static void routine1()
     }
 
     f1 += 3;
-    d1 = deci(f1);
+    d1 = libdar::deci(f1);
     ui->warning(d1.human());
 
     f1 -= 2;
-    d1 = deci(f1);
+    d1 = libdar::deci(f1);
     ui->warning(d1.human());
 
     f1 *= 10;
-    d1 = deci(f1);
+    d1 = libdar::deci(f1);
     ui->warning(d1.human());
 
     f2 = f1;
     f1 /= 3;
-    d1 = deci(f1);
+    d1 = libdar::deci(f1);
     ui->warning(d1.human());
 
     f2 %= 3;
-    d2 = deci(f2);
+    d2 = libdar::deci(f2);
     ui->warning(d2.human());
 
     f2 >>= 12;
-    d2 = deci(f2);
+    d2 = libdar::deci(f2);
     ui->warning(d2.human());
 
     f1 = 4;
     f2 >>= f1;
-    d2 = deci(f2);
+    d2 = libdar::deci(f2);
     ui->warning(d2.human());
 
     f1 = 4+12;
@@@@ -143,9 +143,9 @@@@ static void routine1()
     cout << f3 << endl;
     cout << (123 << 16) << endl;
     f2 <<= 4+12;
-    d2 = deci(f2);
+    d2 = libdar::deci(f2);
     ui->warning(d2.human());
-    d3 = deci(f3);
+    d3 = libdar::deci(f3);
     ui->warning(d3.human());
 
     try
@@@@ -184,12 +184,12 @@@@ static void routine1()
     {
         for(f3 = 2; f3 <= f1; f3++)
         {
-            d1 = deci(f1);
-            d2 = deci(f2);
-            d3 = deci(f3);
+            d1 = libdar::deci(f1);
+            d2 = libdar::deci(f2);
+            d3 = libdar::deci(f3);
             ui->warning(d1.human() + " " + d2.human() + " " + d3.human());
             f2 *= f3;
-            d2 = deci(f2);
+            d2 = libdar::deci(f2);
             ui->warning(d2.human());
         }
     }
@@@@ -197,17 +197,17 @@@@ static void routine1()
     {
         ui->warning(e.get_message());
     }
-    d2 = deci(f2);
-    d1 = deci(f1);
+    d2 = libdar::deci(f2);
+    d1 = libdar::deci(f1);
     ui->warning(string("factoriel(") + d1.human() + ") = " + d2.human());
 }
 
 static void routine2()
 {
-    ui->warning(deci(infinint(2).power((U_I)0)).human());
-    ui->warning(deci(infinint(2).power(infinint(0))).human());
-    ui->warning(deci(infinint(2).power((U_I)1)).human());
-    ui->warning(deci(infinint(2).power(infinint(1))).human());
-    ui->warning(deci(infinint(2).power((U_I)2)).human());
-    ui->warning(deci(infinint(2).power(infinint(2))).human());
+    ui->warning(libdar::deci(infinint(2).power((U_I)0)).human());
+    ui->warning(libdar::deci(infinint(2).power(infinint(0))).human());
+    ui->warning(libdar::deci(infinint(2).power((U_I)1)).human());
+    ui->warning(libdar::deci(infinint(2).power(infinint(1))).human());
+    ui->warning(libdar::deci(infinint(2).power((U_I)2)).human());
+    ui->warning(libdar::deci(infinint(2).power(infinint(2))).human());
 }
@


1.1
log
@deci and move are part of a C++11 STL implementation. Explicitly include
the namespace when using the former and rename the latter.
@
text
@d1 1
a1 1
$NetBSD$
@