This package can use either setuptools (egg.mk) or distutils
(distutils.mk), but it prefers setuptools and thus should be built
with our egg.mk support. Adjust PLIST for egg.mk (different)
variables and drop eggfile conditionalization, as eggs always have
egg-info directories.
Thanks to dholland@ and obache@ for noticing/suggesting.
"fuse-python".
Upstream does not provide NEWS. Perusal of Changelog leads to the
conclusion that there are many bugfixes and minor enhancements, but no
major changes.
(Tested with "bup fuse" and ../fuse-wikipediafs on NetBSD 5.1_STABLE.)
- fix lookup cache that was not really used
- fix advlock for read-only files
- return EBADF on directory read/write instead of passing to the filesystem
- The most significant new feature in this release is MDMF: "Medium-size
Distributed Mutable Files". Unlike standard SDMF files, these provide
efficient partial-access (reading and modifying small portions of the file
instead of the whole thing). MDMF is opt-in (it is not yet the default
format for mutable files), both to ensure compatibility with previous
versions, and because the algorithm does not yet meet memory-usage goals.
Enable it with ``--format=MDMF`` in the CLI (``tahoe put`` and ``tahoe
mkdir``), or the "format" radioboxes in the web interface. See
`<docs/specifications/mutable.rst>`_ for more details (`#393`_, `#1507`_)
- A "blacklist" feature allows blocking access to specific files through
a particular gateway. See the "Access Blacklist" section of
`<docs/configuration.rst>`_ for more details. (`#1425`_)
- A "drop-upload" feature has been added, which allows you to upload
files to a Tahoe-LAFS directory just by writing them to a local
directory. This feature is experimental and should not be relied on
to store the only copy of valuable data. It is currently available
only on Linux. See `<docs/frontends/drop-upload.rst>`_ for documentation.
(`#1429`_)
- The timeline of immutable downloads can be viewed using a zoomable and
pannable JavaScript-based visualization. This is accessed using the
'timeline' link on the File Download Status page for the download, which
can be reached from the Recent Uploads and Downloads page.
- do not confuse FUSE fileno with inode numbers
- do not use kernel name cache, implement our own instead.
- honour cache directive from FUSE filesystem for name and attrbiutes
and the amount of data memory involved is not easy to forcast. We therefore
raise the limit to the maximum.
Patch from Manuel Bouyer. It helps completing a cvs update on a glusterfs
volume.
This maintenance release with no new features. Majot bug fixes are:
Bug 2464 Fixed all the issues caused by GFID mismatch during
distribute rename.
Bug 2988 Fixed the issue of high CPU usage when Directory Quota
is enabled.
Bug 3122 Enhanced the volume set interface to support io-threads
on the client.
Bug 3210 Fixed the issue of modified mtime/atime of the files after
rebalance operation.
Bug 3191 Fixed the issue with symlinks during rebalance operation.
** Security-related Bugfix
- Fix flaw that would allow a person who knows a storage index of a file to
delete shares of that file (#1528).
- Remove corner cases in mutable file bounds management which could expose
extra lease info or old share data (from prior versions of the mutable
file) if someone with write authority to that mutable file exercised these
corner cases in a way that no actual Tahoe-LAFS client does. (Probably not
exploitable.) (#1528).
Major new features according to http://www.gluster.org/
- Elastic Volume Management
- New Gluster Console Manager
- Dynamic Storage for the data center and cloud
Packaging changes:
Use gmake and test TEST_TARGET, so make test runs make check.
Add dependency on py-sqlite2 for "tahoe backup".
Add py-mock dependency because tests are now runnable by the user.
Withdraw fixing of distfiles modes; this one is fine.
* Release 1.8.2 (2011-01-30)
** Compatibility and Dependencies
- Tahoe is now compatible with Twisted-10.2 (released last month), as well
as with earlier versions. The previous Tahoe-1.8.1 release failed to run
against Twisted-10.2, raising an AttributeError on
StreamServerEndpointService (#1286)
- Tahoe now depends upon the "mock" testing library, and the foolscap
dependency was raised to 0.6.1 . It no longer requires pywin32 (which
was used only on windows). Future developers should note that
reactor.spawnProcess and derivatives may no longer be used inside
Tahoe code.
** Other Changes
- the default reserved_space value for new storage nodes is 1 GB (#1208)
- documentation is now in reStructuredText (.rst) format
- "tahoe cp" should now handle non-ASCII filenames
- the unmaintained Mac/Windows GUI applications have been removed (#1282)
- tahoe processes should appear in top and ps as "tahoe", not "python",
on some unix platforms. (#174)
- "tahoe debug trial" can be used to run the test suite (#1296)
- the SFTP frontend now reports unknown sizes as "0" instead of "?",
to improve compatibility with clients like FileZilla (#1337)
- "tahoe --version" should now report correct values in situations where
1.8.1 might have been wrong (#1287)
** Bugfixes and Improvements
- Allow the repairer to improve the health of a file by uploading
some shares, even if it cannot achieve the configured happiness
threshold. This fixes a regression introduced between v1.7.1 and
v1.8.0. (#1212)
- Fix a memory leak in the ResponseCache which is used during mutable
file/directory operations. (#1045)
- Fix a regression and add a performance improvement in the downloader.
This issue caused repair to fail in some special cases. (#1223)
- Fix a bug that caused 'tahoe cp' to fail for a grid-to-grid copy
involving a non-ASCII filename. (#1224)
- Fix a rarely-encountered bug involving printing large strings to
the console on Windows. (#1232)
- Perform ~ expansion in the --exclude-from filename argument to
'tahoe backup'. (#1241)
- The CLI's 'tahoe mv' and 'tahoe ln' commands previously would try
to use an HTTP proxy if the HTTP_PROXY environment variable was set.
These now always connect directly to the WAPI, thus avoiding giving
caps to the HTTP proxy (and also avoiding failures in the case that
the proxy is failing or requires authentication). (#1253)
- The CLI now correctly reports failure in the case that 'tahoe mv'
fails to unlink the file from its old location. (#1255)
- 'tahoe start' now gives a more positive indication that the node
has started. (#71)
- The arguments seen by 'ps' or other tools for node processes are
now more useful (in particular, they include the path of the
'tahoe' script, rather than an obscure tool named 'twistd'). (#174)
** Removed Features
- The tahoe start/stop/restart and node creation commands no longer
accept the -m or --multiple option, for consistency between platforms.
(#1262)
** Packaging
- We now host binary packages so that users on certain operating systems
can install without having a compiler.
<http://tahoe-lafs.org/source/tahoe-lafs/deps/tahoe-lafs-dep-eggs/README.html>
- Use a newer version of a dependency if needed, even if an older
version is installed. This would previously cause a VersionConflict
error. (#1190)
- Use a precompiled binary of a dependency if one with a sufficiently
high version number is available, instead of attempting to compile
the dependency from source, even if the source version has a higher
version number. (#1233)
** Documentation
- All current documentation in .txt format has been converted to
.rst format. (#1225)
- Added docs/backdoors.rst declaring that we won't add backdoors to
Tahoe-LAFS, or add anything to facilitate government access to data.
(#1216)
* Release 1.8.0 (2010-09-23)
** New Features
- A completely new downloader which improves performance and
robustness of immutable-file downloads. It uses the fastest K
servers to download the data in K-way parallel. It automatically
fails over to alternate servers if servers fail in mid-download. It
allows seeking to arbitrary locations in the file (the previous
downloader which would only read the entire file sequentially from
beginning to end). It minimizes unnecessary round trips and
unnecessary bytes transferred to improve performance. It sends
requests to fewer servers to reduce the load on servers (the
previous one would send a small request to every server for every
download) (#287, #288, #448, #798, #800, #990, #1170, #1191)
- Non-ASCII command-line arguments and non-ASCII outputs now work on
Windows. In addition, the command-line tool now works on 64-bit
Windows. (#1074)
** Bugfixes and Improvements
- Document and clean up the command-line options for specifying the
node's base directory. (#188, #706, #715, #772, #1108)
- The default node directory for Windows is ".tahoe" in the user's
home directory, the same as on other platforms. (#890)
- Fix a case in which full cap URIs could be logged. (#685, #1155)
- Fix bug in WUI in Python 2.5 when the system clock is set back to
1969. Now you can use Tahoe-LAFS with Python 2.5 and set your
system clock to 1969 and still use the WUI. (#1055)
- Many improvements in code organization, tests, logging,
documentation, and packaging. (#983, #1074, #1108, #1127, #1129,
#1131, #1166, #1175)
Changes since previous verion, from the NEWS file
2.8.x:
- More scalable directory tree locking
- Atomic open(O_TRUNC) support
- Support big write requests on kernels 2.6.26 and newer
- Out-of-tree fuse module removed
- Better NFS exporting support
- New ioctl and poll requests
- New CUSE (Character Device in Userspace) interface
- Allow umask processing in userspace
- Added cache invalidation notifications
- Bugfixes and small improvements
2.7.x:
- Stacking support for the high level API
- Add filename charset conversion module
- Improved mounting
2.6.x!
- Improved read characteristics (asynchronous reads)
- Support for aborting filesystem connection
- POSIX file locking support
- Request interruption support
- Building module for Linux kernels earlier than 2.6.9 not supported
- Allow block device based filesystems to support swap files
- Several bugs fixed, including a rare system hang on SMP
- a daemon that creates a /dev/fuse socket and that performs PUFFS
relay to FUSE.
- a library and a header file to use in filesystems and libraries that
open /dev/fuse, so that they are directed to the socket.
Tahoe-LAFS is a Free Software/Open Source decentralized data store.
It distributes your filesystem across multiple servers, and even if
some of the servers fail or are taken over by an attacker, the entire
filesystem continues to work correctly and to preserve your privacy
and security.
bindfs is a FUSE filesystem for mounting a directory to another location,
similar to 'mount --bind' in Linux or unionfs in BSD. The permissions in
the mountpoint can be altered through various rules.
MP3FS is a read-only FUSE filesystem which transcodes audio formats
(currently FLAC) to MP3 on the fly when opened and read. This was
written to enable using a FLAC collection with software and/or
hardware which only understands MP3.
It is also a novel alternative to traditional MP3 encoder applications.
Just use your favorite file browser to select the files you want
encoded and copy them somewhere!
From changelog:
Bugs fixed since ChironFS 1.1.0:
o ChironFS could not locate the correct place where chirctl was
when called using the environment variable $PATH to find it.
Now chirctl is called in the same way as ChironFS, leaving
this work to $PATH searches.
Changes since ChironFS 1.1.0
---------------------
o ChironFS could not locate the correct place where chirctl was
when called using the environment variable $PATH to find it.
Now chirctl is called in the same way as ChironFS, leaving
this work to $PATH searches.
Changes since ChironFS 1.0.0
---------------------
o Added the --ctl (or -c) option which allows to mount a pseudo-filesystem
(like /proc) to control the behavior of the Chiron filesystem being mounted.
Currently, it's only possible to show the status of the replicas and change
their status. Dynamically generated nagios plugin scripts are provided too.
o Updated the howto in the chapter 5, explaining the use of the --ctl option.
This changes the buildlink3.mk files to use an include guard for the
recursive include. The use of BUILDLINK_DEPTH, BUILDLINK_DEPENDS,
BUILDLINK_PACKAGES and BUILDLINK_ORDER is handled by a single new
variable BUILDLINK_TREE. Each buildlink3.mk file adds a pair of
enter/exit marker, which can be used to reconstruct the tree and
to determine first level includes. Avoiding := for large variables
(BUILDLINK_ORDER) speeds up parse time as += has linear complexity.
The include guard reduces system time by avoiding reading files over and
over again. For complex packages this reduces both %user and %sys time to
half of the former time.
GSTFS is a filesystem for on-demand transcoding of music files between
different formats. It utilizes the gstreamer library for conversion so any
formats supported by gstreamer should also be supported by gstfs. The
filesystem's only requirement is that the gstreamer pipeline begin with a
filesrc with the name "_source" and end with an fdsink with the name "_dest".
The filesystem will automatically substitute the filename and fd number in
these pipelines.
Tools to access and modify a file system image without having to
mount it.
Classic commands included (prefixed with fsu_):
* cat, diff, du, find, ls
* chflags, chmod, chown, mv, rm, rmdir
* cp, ln, mkdir, mkfifo, mknod, touch
New commands:
* fsu_ecp (cp-like which allows the user to get/put files from/to the image)
* fsu_exec (executes a local command like text editor on a file from the image)
* fsu_write (writes its input in a file on the image)
* fsu_console (a command console tool that uses previous commands and adds
directory browsing)
This package is enabled only for NetBSD 4.99.72 and newer.
- assume that Python 2.4 and 2.5 are compatible and allow checking for
fallout.
- remove PYTHON_VERSIONS_COMPATIBLE that are obsoleted by the 2.3+
default. Modify the others to deal with the removals.
djmount is a UPnP AV client. It mounts the media content of
compatible UPnP AV devices as a Unix filesystem.
Djmount automatically discovers all UPnP AV Media Servers on the
network, and makes the content available in a directory tree. All
shared files (e.g. Audio or Video files) are directly visible and can
be played using your favorite media player.
[15:53:42] agc@inspiron1300 ~ 190 > df /mnt
Filesystem 1K-blocks Used Avail %Cap Mounted on
/dev/puffs 28101396 20741466 5954862 77% /mnt
[15:53:44] agc@inspiron1300 ~ 191 > l /mnt
total 5
dr-xr-xr-x 5 root wheel 512 Jan 1 2000 .
drwxr-xr-x 37 root wheel 1536 Mar 15 13:22 ..
dr-xr-xr-x 5 root wheel 512 Jan 1 2000 .debug
dr-xr-xr-x 8 root wheel 512 Jan 1 2000 INSPIRON1300: Alistair Crooks:
dr-xr-xr-x 7 root wheel 512 Jan 1 2000 MediaTomb
-r--r--r-- 1 root wheel 41 Jan 1 2000 devices
[15:54:23] agc@inspiron1300 ~ 192 >
- Ask for a password in the terminal so that it doesn't show in ps.
- Support different codepages.
- Support for SOCKS proxy.
- Bug fixes.
Add CFLAGS.NetBSD+=-D_NETBSD_SOURCE to make it build.
I'm not sure about FUSE_VERSION vs. FUSE_USE_VERSION in different
implementations, so define FUSE_USE_VERSION from FUSE_VERSION if it
does not exist and use that.
2007-04-17 CHANGES:
- fixed a stupid bug that avoids mounting the root of a server
- documented the optional port for mounting a webdav server
- improved attribute caching for some special cases
While I can find little documentation out there, it seems to work
quite well for me (Subversion 1.4.4 + apr1). Thusfar it's seemed
quite speedy and useful. I'm thinking that fuse might be an interesting
way to get distfiles (curlftpfs) or source snapshots into pkgsrc.
* Fix: successful hard link removal may reported "Input/Output error".
* Fix: listing the content of a single directory having over 600,000 files
and subdirectories may returned "Input/Output error" or truncated
directory listing.
* Fix: atime mount option didn't work.
* Fix: POSIX file time updates (file access and modification times, inode
change time). Please note, though NTFS-3G updates the times correctly now,
FUSE may keep caching old, invalid time attributes for a short time after
rename(2), creat(2), and some file operations which involve hard links.
FUSE has fixed already the first two cases and they may be included in
the Linux kernel 2.6.24 release.
* New: relatime support was added and made the default atime update method.
* Fix: reads over the end of a file could return "invalid argument" error
using some recent kernels. This problem may have caused Thunderbird
to hang.
* Fix: some faulty versions of the gcc compiler have crashed during
compilation.
* Fix: the utimes(2) system call may incorrectly updated the last status
change filestamp.
* Change: the ntfs.streams.list extended attribute enumerates Alternate
Data Streams (ADS) '\0' separated.
- UTF8 internationalization of filenames, volumes and servers
- proper uid/gid mapping to enable environments with a common or different user directory
- DHX2 UAM from Derrik Pates
- fixes to deal with the Apple Airport Extreme quirks
- fixes to deal with netatalk quirks
- improved status output for debugging
- various small bug fixes
* fix: big-endian problems exposed by the NTFS-3G Test Suite.
* fix: free space calculation may was wrong for >1 TB volumes.
* fix: some faulty Thunderbird versions caused system log flooding.
* fix: uClibc and uClinux improvements.
* fix: logging improvements.
* fix: many other minor fixes.
* fix: unwritten sparse file regions could get corrupted if the end
of a write wasn't aligned to cluster boundary. Sparse files are very
rarely used, most typically by bittorent clients.
* fix: creation of read-only files failed.
* change: free inode calculation was CPU intensive during write activity.
This could reduce write performance when millions of files are used if
a file manager or desktop applet regularly polled for disk space usage.
Now the statfs() system call is constant time.
* change: file creation performance is improved by 40-80%.
* change: smart directory traversal is 20-200% faster for disk based and
300-600% faster for memory cache based traversals.
STABLE Version 1.913 (September 13, 2007) Release Notes
* change: free disk space calculation was highly CPU intensive during
write activity. This could reduce write performance severely, especially
for huge volumes if a file manager or desktop applet regularly polled
for disk space usage. Now the calculation is constant time, around only
10 microseconds on a typical desktop, and it's completely independent
of the volume size.
* fix: hibernation check was too rigid and mount was refused in read/write
mode unnecessarily in some cases.
* all performance improvements included unchanged from the 1.910-RC test
release. Please see those details below.
Version 1.910-RC (September 10, 2007) -- Release Notes
* change: significantly improved the performance of writing many files,
typically by 50-1000% (copying files, unpacking, untarring packages,
compilation, directory synchronization, ...).
* change: improved the performance of writing multi-GB size files,
particularly after the creation of many thousands of other files.
Previous disk space defragmentation is highly advised for the best
possible speed. File level defragmentation is not enough.
* change: improved concurrent write performance.
