1
1
ncdu/ChangeLog
Yorhel 6de0a8ec00 Use correct hard link information after partial recalculation or deletion
Hard link detection is now done in a separate pass on the in-memory tree,
and duplicates can be 'removed' and 're-added' on the fly. When making any
changes in the tree, all hard links are re-added before the operation and
removed again afterwards.

While this guarantees that all hard link information is correct, it does
have a few drawbacks. I can currently think of two:

 1. It's not the most efficient way to do it, and may be quite slow on
    large trees. Will have to do some benchmarks later to see whether
    it is anything to be concerned about.

 2. The first encountered item is considered as 'counted' and all items
    encountered after that are considered as 'duplicate'. Because the
    order in which we traverse the tree doesn't always have to be the
    same, the items that will be considered as 'duplicate' can vary with
    each deletion or re-calculation. This might cause confusion for
    people who aren't aware of how hard links work.
2009-05-11 20:02:46 +02:00

83 строки
2.9 KiB
Plaintext

git - ?
- Implemented hard link detection
1.5 - 2009-05-02
- Fixed incorrect apparent size on directory refresh
- Browsing keys now work while file info window is displayed
- Current directory is assumed when no directory is specified
- Size graph uses the apparent size if that is displayed
- Items are ordered by displayed size rather than disk usage
- Removed switching between powers of 1000/1024
- Don't rely on the availability of suseconds_t
- Correctly handle paths longer than PATH_MAX
- Fixed various bugs related to rpath()
- Major code rewrite
- Fixed line width when displaying 100%
1.4 - 2008-09-10
- Removed the startup window
- Filenames ending with a tidle (~) will now also
be hidden with the 'h'-key
- Fixed buffer overflow when supplying a path longer
than PATH_MAX (patch by Tobias Stoeckmann)
- Used S_BLKSIZE instead of a hardcoded block size of 512
- Fixed display of disk usage and apparent sizes
- Updated ncdu -h
- Included patches for Cygwin
- Cursor now follows the selected item
- Added spaces around path (debian #472194)
- Fixed segfault on empty directory (debian #472294)
- A few code rewrites and improvements
1.3 - 2007-08-05
- Added 'r'-key to refresh the current directory
- Removed option to calculate apparent size: both
the disk usage and the apparent size are calculated.
- Added 'a'-key to switch between showing apparent
size and disk usage.
- Added 'i'-key to display information about the
selected item.
- Small performance improvements
- configure checks for ncurses.h (bug #1764304)
1.2 - 2007-07-24
- Fixed some bugs on cygwin
- Added du-like exclude patterns
- Fixed bug #1758403: large directories work fine now
- Rewrote a large part of the code
- Fixed a bug with wide characters
- Performance improvements when browsing large dirs
1.1 - 2007-04-30
- Deleting files and directories is now possible from
within ncdu.
- The key for sorting directories between files has
changed to 't' instead of 'd'. The 'd'-key is now
used for deleting files.
1.0 - 2007-04-06
- First stable release
- Small code cleanup
- Added a key to toggle between sorting dirs before
files and dirs between files
- Added graphs and percentages to the directory
browser (can be enabled or disabled with the 'g'-key)
0.3 - 2007-03-04
- When browsing back to the previous directory, the
directory you're getting back from will be selected.
- Added directory scanning in quiet mode to save
bandwidth on remote connections.
0.2 - 2007-02-26
- Fixed POSIX compliance: replaced realpath() with my
own implementation, and gettimeofday() is not
required anymore (but highly recommended)
- Added a warning for terminals smaller than 60x16
- Mountpoints (or any other directory pointing to
another filesystem) are now considered to be
directories rather than files.
0.1 - 2007-02-21
- Initial version