| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
New in v0.7.16 (2018/01/12)
---------------------------
* Fixed bug #1733057 AttributeError: 'GPGError' object has no attribute 'decode'
- Replaced call to util.ufn() with call to util.uexc(). Stupid typo!
* More fixes for Unicode handling
- Default to 'utf-8' if sys.getfilesystemencoding() returns 'ascii' or None
- Fixed bug #1386373 with suggestion from Eugene Morozov
* Patched in lp:~crosser/duplicity/fix-oauth-flow
- Fixed bug #1638236 "BackendException with oauth2client 4.0.0"
* Patched in lp:~crosser/duplicity/dpbx-fix-file-listing
- Fixed bug #1639664 "Dropbox support needs to be updated for Dropbox SDK v7.1"
* Patched in lp:~crosser/duplicity/fix-small-file-upload
- Fixed small file upload changes made in Dropbox SDK v7.1
* Fix pylint error in webdavbackend.py
New in v0.7.15 (2017/11/13)
---------------------------
* Fixed bug introduced in new megabackend.py where process_commandline()
takes a string not a list. Now it takes both.
* Updated web page for new megabackend requirements.
* Patched in lp:~mterry/duplicity/more-decode-issues
- Here's some fixes for another couple UnicodeDecodeErrors.
- The duplicity/dup_time.py fixes when a user passes a utf8 date string (or a string with bogus
utf8 characters, but they have to really try to do that). This is bug 1334436.
- The bin/duplicity change from str(e) to util.uexc(e) fixes bug 1324188.
- The rest of the changes (util.exception_traceback and bin/duplicity changes to use it) are to
make the printing of exceptions prettier. Without this, if you see a French exception, you see
"accept\xe9es" instead of "acceptées".
- You can test all of these changes in one simple line:
$ LANGUAGE=fr duplicity remove-older-than $'accept\xffées'
* Fix backend.py to allow string, list, and tuple types to support megabackend.py.
* Fixed bug #1715650 with patch from Mattheww S
- Fix to make duplicity attempt a get first, then create, a container
in order to support container ACLs.
* Fixed bug #1714663 "Volume signed by XXXXXXXXXXXXXXXX, not XXXXXXXX"
- Normalized comparison length to min length of compared keys before comparison
- Avoids comparing mix of short, long, or fingerprint size keys.
* Merged in lp:~mterry/duplicity/rename-dep
- Make rename command a dependency for LP build
* Fixed bug #1654756 with new b2backend.py module from Vincent Rouille
- Faster (big files are uploaded in chunks)
- Added upload progress reporting support
* Fixed bug #1448094 with patch from Wolfgang Rohdewald
- Don't log incremental deletes for chains that have no incrementals
* Fixed bug #1724144 "--gpg-options unused with some commands"
- Add --gpg-options to get version run command
* Fixed bug #1720159 - Cannot allocate memory with large manifest file since 0.7.03
- filelist is not read if --file-changed option in collection-status not present
- This will keep memory usage lower in non collection-status operations
* Fixed bug #1723890 with patch from Killian Lackhove
- Fixes error handling in pydrivebackend.py
* Fixed bug #1730902 GPG Error Handling
- use util.ufn() not str() to handle encoding
New in v0.7.14 (2017/08/31)
---------------------------
* Merged in lp:~dawgfoto/duplicity/skip_sync_collection_status
- collection-status should not sync metadata
- up-to-date local metadata is not needed as collection-status is
generated from remote file list
- syncing metadata might require to download several GBs
* Fixed slowness in 'collection-status' by basing the status on the
remote system only. The local cache is treated as empty.
* Fixed encrypted remote manifest handling to merely put out a non-fatal
error message and continue if the private key is not available.
* Patched in lp:~mterry/duplicity/giobackend-display-name
- giobackend: handle a wider variety of gio backends by making less assumptions;
in particular, this fixes the google-drive: backend
* Fixed bug #1709047 with suggestion from Gary Hasson
- fixed so default was to use original filename
* Fixed PEP8 errors in bin/duplicity
* Merged in lp:~mterry/duplicity/gio_child_for_display_name_0.7
- gio: be slightly more correct and get child GFiles based on display name
* Fixed bug #1711905 with suggestion from Schneider
- log.Warn was invoked with log.warn in webdavbackend.py
* Merged in lp:~mterry/duplicity/gpg-tag-versions
- Support gpg versions numbers that have tags on them.
- This can happen if you build gpg from git trunk (e.g. 2.1.15-beta20). Or if you run
against the freedesktop flatpak runtime (e.g. 2.1.14-unknown).
* Fixed bug #1394386 with new module megabackend.py from Tomas Vondra
- uses megatools from https://megatools.megous.com/ instead of mega.py library
which has been deprecated
- fixed copyright and PEP8 issues
- replaced subprocess.call() with self.subprocess_popen() to standardize
* Fixed bug #1713640 with patch from Aleksandar Ivanisevic
- replace 2.7 syntax with 2.6 equivalent
* Fixed bug #1538333 Assertion error in manifest.py: assert filecount == ...
- Made sure to never pass .part files as true manifest files
- Changed assert to log.Error to warn about truncated/corrupt filelist
- Added unit test to make sure detection works
- Note: while this condition is serious, it will not affect the basic backup and restore
functions. Interactive options like --list-files-changed and --file-changed will not
work correctly for this backup set, so it is advised to run a full backup as soon as
possible after this error occurs.
* Fixed bug #1638033 Remove leading slash on --file-to-restore
- code already used rstrip('/') so change to just strip('/')
New in v0.7.13.1 (2017/06/18)
-----------------------------
* Fixed problem in dist/makedist when building on Mac where AppleDouble
files were being created in the tarball. See:
https://superuser.com/questions/61185/why-do-i-get-files-like-foo-in-my-tarball-on-os-x
New in v0.7.13 (2017/06/12)
---------------------------
* Fixed bug #1680682 with patch supplied from Dave Allan
- Only specify --pinentry-mode=loopback when --use-agent is not specified
* Fixed man page that had 'cancel' instead of 'loopback' for pinentry mode
* Fixed bug #1684312 with suggestion from Wade Rossman
- Use shutil.copyfile instead of os.system('cp ...')
- Should reduce overhead of os.system() memory usage.
* Fixed bug #1320832 with suggestion from Oskar Wycislak
- Use chunks instead of reading it all in swiftbackend
* Fixed bug #1689632 with patch from Howard Kaye
- On MacOS, the tempfile.TemporaryFile call erroneously raises an
IOError exception saying that too many files are open. This causes
restores to fail randomly, after thousands of files have been restored.
* Fixed bug #1320641 and others regarding lockfile
- swap from lockfile to fasteners module
- use an fcntl() style lock for process lock of duplicity cache
- lockfile will now clear if duplicity is killed or crashes
* May have finally fixed bug #1556553, "Too many open files...".
- Applied patch from Howard Kaye, question #631423. The fix is to dup
the file descriptor, and then close the file in the deallocator
routine in the glue code. Duping the file lets the C code and the Python
code each close the file when they are done with it.
- Invalidated and removed the fix put in for bug #1320832.
- Caveat: long incremental chains will still eat up a large number of file
descriptors. It's a very risky practice, so I'm not inclined to fix it.
* Revisited bug #670891 with patch from Edgar Soldin
- Forced librsync.PatchedFile() to extract file object from TemporaryFile()
object when on Windows or Cygwin systems. This allows us to avoid the
problem of tmpfile() use which creates temp files in the wrong place.
- See discussion at https://bugs.launchpad.net/duplicity/+bug/670891
* Fix bug #1672540 with patch from Benoit Nadeau
- Rename would fail to move par files when moving across filesystems.
- Patch uses shutil.move() to do the rename instead.
* Fixed bug #1265765 with patches from Matthias Larisch and Edgar Soldin
- SSH Paramiko backend now uses BufferedFile implementation to enable
collecting the entire list of files on the backend.
* Copy.com is gone so remove copycombackend.py.
* Merged in lp:~xlucas/duplicity/swift-multibackend-bug
- Fix a bug when swift backend is used in a multibackend configuration.
* Merged in lp:~duplicity-team/duplicity/po-updates
New in v0.7.12 (2017/03/21)
---------------------------
* Fixed bug #1623342 with patch supplied by Daniel Jakots
- Failing test on OpenBSD because tar/gtar not found
* Fixed bug #1654220 with patch supplied by Kenneth Newwood
- Duplicity fails on MacOS because GPG version parsing fails
* Fixed bug #1655268 "--gpg-binary option not working"
- If gpg binary is specified rebuild gpg profile using new binary location
* Fixed bug #1658283 "Duplicity 0.7.11 broken with GnuPG 2.0"
- Made gpg version check more robust than just major version
- Now use --pinentry-mode=loopback on gpg 2.1 and greater
- Removed check for non-Linux systems, a false problem
* Merged in lp:~matthew-t-bentley/duplicity/duplicity
- Sets a user agent. Backblaze asked for this in case there are errors that originate
from the Duplicity B2 backend
- Only retrieves a new upload URL when the current one expires, to bring it in line
with their best practices for integrations: https://www.backblaze.com/b2/docs/integration_checklist.html
* Add detail about import exceptions in onedrivebackend.py
* Fixed bug #1657916 with patch supplied by Daniel Harvey
- B2 provider cannot handle two backups in the same bucket
* Fixed bug #1603704 with patch supplied by Maciej Bliziński
- Crash with UnicodeEncodeError
* Some fixes to gpg.py to handle gpg1 & gpg2 & gpg2.1 commandline issues
- --gpg-agent is optional on gpg1, but on gpg2 it is used automatically
- --pinentry-mode is not a valid opt until gpg2.1, so condition on that
* Fixed bug #1367675 - IMAP Backend does not work with Yahoo server
- added the split() as needed in 'nums=list[0].strip().split(" ")'
- the other fixes mentioned in the bug report comments were already done
* Fixed bug #1671852 - Code regression caused by revision 1108
- change util.uexc() back to bare uexc()
* Fixed bug #1668750 - Don't mask backend errors
- added exception prints to module import errors
New in v0.7.11 (2016/12/31)
---------------------------
* Fixed bugs #815510 and #1615480
- Changed default --volsize to 200MB
* Merged in lp:~mstoll-de/duplicity/duplicity
- Backblaze announced a new domain for the b2 api
* Merged in lp:~aaron-whitehouse/duplicity/bug_1620085_exclude-if-present-locked-folder
- Fixes Bug #1620085: --exclude-if-present gives OSError looking for tag in locked folders
* Fixed bug #1623342 with patch from Daniel Jakots
- failing test on OpenBSD because tar/gtar not found
* Merged in lp:~mwilck/duplicity/duplicity
- GPG: enable truly non-interactive operation with gpg2
- This patch fixes the IMO unexpected behavior that, when using GnuPG2, a pass phrase dialog always pops up for
saving backups. This is particularly annoying when trying to do unattended / fully automatic backups.
* Fixed bug #1621194 with code from Tornhoof
- Do backup to google drive working without a service account
* Fixed bug #1642098 - does not create PAR2 archives when '--par2-options' is used
- Missing space between par2-options plus default options
* Fix bug using 40-char sign keys, from Richard McGraw on mail list
- Remove truncation of argument and adjust comments
* Merged in lp:~dernils/duplicity/robust-dropbox-backend
- Added new command line option --backend-retry-delay
that allows to determine the time that duplicity sleeps
before retrying after an error has occured.
- Added some robustness to dpbxbackend.py that ensures re-authentication
happens in case that a socket is changed (e.g. due to a forced reconnect
of a dynamic internet connection).
* Merged in lp:~ed.so/duplicity/manpage.fixes
- Fix html output via rman on the website
* Merged in lp:~horgh/duplicity/copy-symlink-targets-721599
- Add --copy-links to copy symlink contents, not just the link itself.
* Merged in lp:~aaron-whitehouse/duplicity/Bug_1624725_files_within_folder_slash
- Fixed Bug #1624725, so that an include glob ending in "/" now includes folder contents (for globs with
and without special characters). This preserves the behaviour that an expression ending in "/" only
matches a folder, but now the contents of any matching folder is included.
* Fix problem with gpg2 in yakety and zesty
* Fix Bug #1642813 with patch from Ravi
- If stat() returns None, don't attempt to set perms.
* Merged in lp:~breunigs/duplicity/amazondrive3
- As reported on the mailinglist, if a space is entered while duplicity asks for the URL, it fails.
Since all important spaces are URL encoded anyway, this should be fine even if there are spaces in
the URL at all. I also patched it in the onedrive backend, because it must have similar issues.
* Prep for 0.7.11
New in v0.7.10 (2016/08/20)
---------------------------
* Merged in lp:~mwilck/duplicity/0.7-series
- Speedup of path_matches_glob() by about 8x. See
https://code.launchpad.net/~mwilck/duplicity/0.7-series/+merge/301332
for more details.
* Remove -w from setsid in functional tests.
* Fixed conflict in merge from Martin Wilck and applied
- https://code.launchpad.net/~mwilck/duplicity/0.7-series/+merge/301492
- merge fixes setsid usage in functional testing.
* Fixed bug #1612472 with patch from David Cuthbert
- Restore from S3 fails with --with-prefix-archive if prefix includes '/'
* Merged in lp:~arashad.ahamad/duplicity/duplicity_latest
- Changes for connecting to IBM Bluemix ObjectStorage. See man page.
* Merged in lp:~fenisilius/duplicity/acd_init_mkdir
- Allow duplicity to create remote folder
|
|
to microcode-20171117.tgz because microcode-20180108.tgz is removed.
|
|
|
|
all security advisory up to and including XSA254.
While there pass XEN_VENDORVERSION=nb${PKGREVISION} to make so that
'xl info' shows the NetBSD PKGREVISION. If PKGREVISION is not available,
define this as 'nb0'.
|
|
as it breaks the plugin mechanism due to not being able to
resolve various symbols
|
|
msgpack distribution changed name from 'msgpack-python' to 'msgpack'.
Bump PKGREVISION.
|
|
6.1.2
Auditbeat
- Add an error check to the file integrity scanner to prevent a panic
when there is an error reading file info via lstat.
Filebeat
- Switch to docker prospector in sample manifests for Kubernetes
deployment
|
|
Patch from Travis Paul in PR pkg/52930.
Bump revision.
|
|
In pbulk with PKGSRC_RUN_TESTS, test dependencies are not installed
but expected.
|
|
> This README gives references for one of three mitigation strategies
> for Meltdown.
> This series is a first-class migitation pagetable isolation series for
> Xen. It is available for Xen 4.6 to Xen 4.10 and later.
bump PKGREVISION
|
|
|
|
16 months worth of updates. requires qt5>=5.7
|
|
|
|
|
|
Fixes at least joyent/pkgsrc#60.
|
|
Use TEST_DEPENDS.
0.40 Tue Jan 16 10:00:09 2017
- github Issue #5 - Deep directories pathmk
- rt 123966 - switch to bsd_glob() since glob() will disappear in perl 5.30
- rt 123971 - skip symlink tests when the OS does not support symlinks
- rt 123970 - use Path::Tiny instead of File::Slurp in tests
- rt 117241 - add test for read only directories
|
|
|
|
- rsyslogd: add capability to specify that no pid file shall be
written
- ompgsql: considerable enhancements
- build system: removed --enable-rtinst configure option
- pmrfc3164: support for headerless messages
- omhiredis: add option to use RPUSH instead of LPUSH
- mmexternal improvements
- omprog: refactored, code shared with mmexternal moved to common
object
- logctl tool: refactor to support newer rsyslog standards
- imfile: added support for Solaris File Event notification (FEN)
- core/action: new parameter "action.errorfile"
- imfile: added new module parameter "sortFiles"
- imuxsock: improved status reporting: socket name received from
systemd
- build system: added new testbench configure switches
- mmpstrucdata: new parameter "sd_name.lowercase"
- omfile: add module-global option "dynafile.donotsuspend"
- testbench: add a capability to turn off libfaketime tests via
configure
- testbench: name valgrind tests consistently
- RainerScript: add function parse_json()
- RainerScript: add function substring()
- RainerScript: add function http_request()
- RainerScript: add function previous_is_suspended()
- Patches from BSD projects have been imported
- script bugfix: invalid function names were silently ignored
- rainerscript: add int2hex() function
- rainerscript: add is_time() function
- RainerScript: add function script_error() and error-reporting
support
- testbench: fixed build problem of testbench tools under Alpine Linux
- added --enable-libsystemd configure option to enforce use of
libsystemd
- core/glbl: remove long-unused option $optimizeforuniprocessor
- core/queue: emit better status messages at rsyslog shutdown
- fixed a couple of build issues with gcc-7 (in less frequently used
modules)
- fixed a couple of build issues on the arm platform (actually
raspbian)
- impstats: fix invalid counter definitions for getrusage() reporting
- imudp bugfix: potential segfault in ratelimiting
- imptcp bugfix: access to free'ed memory
- mmanon bugfix: fix wrong ipv6 embedded recognition
- imfile bugfix: not detecting files in directory when wildcards are
used.
- script bugfix: improper string-to-number conversion for negative
numbers
- core/action bugfix: 100% CPU utilization on suspension of output
module
- core/variables bugfix: bare $! cannot be used in set statement
- core bugfix: auto commit of actions improperly handled
- core bugfix: filename length limitation of 199 bytes
- core bugfix: undefined behavior due to integer overflow
- core bugfix: race on LocalHostIP property during startup
- bugfix: potential segfault on startup
- omhiredis bugfix: rsyslog segfault on startup if no template is
specified
- omprog bugfix: argv[0] not set when using binary without arguments
- core: refactoring of rsyslog's cstr "class"
- parent directory creation function refactored
- mmsnmptrapd bugfix: potential misadressing
- imkafka: fix potential small ressource leak
- imkafka bugfix: do not emit error message on regular state
- omkafka: expose operational status to user where useful
- omkafka bugfix: potential message duplication
- omkafka: fix multithreading
- omkafka bugfix: potential misadressing
- omkafka bugfix: build fails with older versions of librdkafka
- omgssapi bugfix: fix compiler warnings with gcc-7
- dnscache bugfix: entries were cached based on IP AND port number
- omkafka bugfix: fixed memory leak
- mmdblookup bugfix: replace thread-unsafe strtok() by thread-safe
counterpart
- pmnormalize bugfix: remove unsave "strcat" implementation
- rainerscript bugfix: ltrim() and rtrim function misadressing
- imklog bugfix: local host IP was hardcoded to 127.0.0.1
- cleanup: remove obsolete pre-KSI GuardTime signature interface
- cleanup: obsolete defintion SOL_TCP replaced by newer IPPROTO_TCP
- lookup tables: fixed undefined behavior detected by UBSan
|
|
|
|
|
|
|
|
|
|
|
|
0.4.15:
Bug fixes.
|
|
the SMF manifest. Bump PKGREVISION.
|
|
|
|
|
|
This module permits to integrate arbitrary external programs into
rsyslog's logging. It is similar to the "execute program (^)" action,
but offers better security and much higher performance. While "execute
program (^)" can be a useful tool for executing programs if rare events
occur, omprog can be used to provide massive amounts of log data to a
program.
|
|
|
|
|
|
|
|
|
|
|
|
fixes for Meltdown and Spectre vulnerabilites.
|
|
|
|
Backported from u-boot master, with some added adjustments.
|
|
|
|
Plus, the file commit.d/20store-metadata had not been included in the
package, probably by accident. Regenerating cf-files.mk restored it.
|
|
|
|
|
|
Fixes:
- check: data corruption fix: fix for borg check --repair malfunction, #3444.
See the more detailled notes close to the top of this document.
- delete: also delete security dir when deleting a repo, #3427
- prune: fix building the "borg prune" man page, #3398
- init: use given --storage-quota for local repo, #3470
- init: properly quote repo path in output
- fix startup delay with dns-only own fqdn resolving, #3471
New features:
- added zstd compression. try it!
- added placeholder {reverse-fqdn} for fqdn in reverse notation
- added BORG_BASE_DIR environment variable, #3338
Other changes:
- list help topics when invalid topic is requested
- fix lz4 deprecation warning, requires lz4 >= 1.7.0 (r129)
- add parens for C preprocessor macro argument usages (did not cause malfunction)
- exclude broken pytest 3.3.0 release
- updated fish/bash completions
- init: more clear exception messages for borg create, #3465
- docs:
- add auto-generated docs for borg config
- don't generate HTML docs page for borgfs, #3404
- docs update for lz4 b2 zstd changes
- add zstd to compression help, readme, docs
- update requirements and install docs about bundled lz4 and zstd
- refactored build of the compress and crypto.low_level extensions, #3415:
- move some lib/build related code to setup_{zstd,lz4,b2}.py
- bundle lz4 1.8.0 (requirement: >= 1.7.0 / r129)
- bundle zstd 1.3.2 (requirement: >= 1.3.0)
- blake2 was already bundled
- rename BORG_LZ4_PREFIX env var to BORG_LIBLZ4_PREFIX for better consistency:
we also have BORG_LIBB2_PREFIX and BORG_LIBZSTD_PREFIX now.
- add prefer_system_lib* = True settings to setup.py - by default the build
will prefer a shared library over the bundled code, if library and headers
can be found and meet the minimum requirements.
Pre-1.1.4 potential data corruption issue
-----------------------------------------
A data corruption bug was discovered in borg check --repair, see issue #3444.
This is a 1.1.x regression, releases < 1.1 (e.g. 1.0.x) are not affected.
To avoid data loss, you must not run borg check --repair using an unfixed version
of borg 1.1.x. The first official release that has the fix is 1.1.4.
Package maintainers may have applied the fix to updated packages of 1.1.x (x<4)
though, see the package maintainer's package changelog to make sure.
If you never had missing item metadata chunks, the bug has not affected you
even if you did run borg check --repair with an unfixed version.
When borg check --repair tried to repair corrupt archives that miss item metadata
chunks, the resync to valid metadata in still present item metadata chunks
malfunctioned. This was due to a broken validator that considered all (even valid)
item metadata as invalid. As they were considered invalid, borg discarded them.
Practically, that means the affected files, directories or other fs objects were
discarded from the archive.
Due to the malfunction, the process was extremely slow, but if you let it
complete, borg would have created a "repaired" archive that has lost a lot of items.
If you interrupted borg check --repair because it was so strangely slow (killing
borg somehow, e.g. Ctrl-C) the transaction was rolled back and no corruption occurred.
The log message indicating the precondition for the bug triggering looks like:
item metadata chunk missing [chunk: 001056_bdee87d...a3e50d]
If you never had that in your borg check --repair runs, you're not affected.
But if you're unsure or you actually have seen that, better check your archives.
By just using "borg list repo::archive" you can see if all expected filesystem
items are listed.
|
|
The actual fix as been done by "pkglint -F */*/buildlink3.mk", and was
reviewed manually.
There are some .include lines that still are indented with zero spaces
although the surrounding .if is indented. This is existing practice.
|
|
1.4.6:
Bug fixes.
|
|
|
