QDirStat

Qt-based directory statistics: KDirStat without any KDE -- from the author of the original KDirStat.

(c) 2015-2021 Stefan Hundhammer [email protected]

Target Platforms: Linux, BSD, Unix-like systems

License: GPL V2

Updated: 2021-03-14

Screenshot

Main window screenshot - notice the multi-selection in the tree and the treemap

Overview

QDirStat is a graphical application to show where your disk space has gone and to help you to clean it up.

This is a Qt-only port of the old Qt3/KDE3-based KDirStat, now based on the latest Qt 5. It does not need any KDE libs or infrastructure. It runs on every X11-based desktop on Linux, BSD and other Unix-like systems.

QDirStat has a number of new features compared to KDirStat. To name a few:

• Multi-selection in both the tree and the treemap.

• Unlimited number of user-defined cleanup actions.

• Properly show errors of cleanup actions (and their output, if desired).

• Configurable file categories (MIME types), treemap colors, exclude rules, tree columns.

• Package manager support:

  • Show what software package a system file belongs to.
  • Packages view showing disk usage of installed software packages and their individual files.
  • Unpackaged files view showing what files in system directories do not belong to any installed software package.

• New views:

  • Disk usage per file type (by filename extension).
  • File size histogram view.
  • Free, used and reserved disk size for each mounted filesystem (like df)

See section New Features for more details.

Table of Contents

1. Screenshot
2. Latest Stable Release
3. Latest News
4. History
5. Related Software: KDirStat, WinDirStat, K4DirStat and more
6. Motivation / Rant: Why?
7. Features
8. MacOS X Compatibility
9. Windows Compatibility
10. Ready-made Packages
11. Building
12. Contributing
13. Troubleshooting
14. Further Reading
15. Packaging Status
16. Donate

More Screenshots

Full-size images and descriptions on the Screenshots Page

Donate

QDirStat is Free Open Source Software; you are not required to pay anything. If you want to support continued development of QDirStat, donations are most welcome.

Donate any amount you like via PayPal:

Latest Stable Release

QDirStat V1.7

See the release announcement.

Download installable binary packages for various Linux distributions here: Ready-made packages

Latest News

• 2020-12-21

  • Fixed GitHub Issue #149: Segfault when using a cleanup action with refresh policy AssumeDeleted after a cleanup action with RefreshParent in the same directory.

• 2020-11-23

  • Menu reorganization: They had become a little too crowded, especially on the top level.

    • Moved out some options from the menus; they are still available when editing the config file manually:

      • "Show current path"
      • "Treemap as side panel"

    • "Expand tree level" is now limited to level 5 (formerly 9). Opening that many tree branches means a huge performance drop anyway.

    • The former "Treemap" menu is now a submenu of "View". Most of those actions are available as tool bar buttons and mouse wheel operations anyway.

    • The former "Settings" menu is gone; "Configure QDirStat" is now in the "Edit" menu. There was only that one action in the "Settings" menu, and that is quite wasteful in terms of screen space and toplevel menu complexity.

      I experimented shortly with moving some more settings there, e.g. "Show details panel" and "Show tremap", but I reverted that: The details panel is different for each layout (L1, L2, L3), and the layout switching actions are in the "View" menu where they belong; so "Show details panel" should really remain near them. And not having "Show treemap" in the "Treemap" submenu would be very confusing.

    The new toplevel menus are now:

    • File
    • Edit
    • View
    • Go To
    • Discover
    • Clean up
    • Help

    I.e. it's down to 7 items which is generally regarded as the gold standard by usability experts.

    I am not completely happy yet with how additional views are scattered among several menus:

    • "File" -> "Show installed packages"
    • "File" -> "Show unpackaged files"
    • "View" -> "File size statistics"
    • "View" -> "File type statistics"
    • "View" -> "Show mounted filesystems"

    The rationale is that those views in the "File" menu load completely new content, replacing the scanned directory in both the tree view and the treemap; thus they are on a similar level as "Open directory", so they are right next to that action.

    The other views open in separate windows, so they are add-on views to the currently loaded directory tree.

• 2020-11-22

  • Added a "Discover" toplevel menu to easily find

    • the largest files
    • the newest files
    • the oldest files
    • files with multiple hard links
    • broken symbolic links
    • sparse files

    in the whole displayed directory tree or, if a directory is selected, in that subtree.

    In each case, a non-modal dialog is opened with a list of the results. Clicking on one of them selects it in the main window where you can see more details and use cleanup actions.

    

    Most of that was already there in some way or the other, but now it's easier to discover (pun intended) and to use.

    Finding large files is of course what the treemap is primarily for; just spot large blobs and click on them to find out what they are and where they are hiding.

    Finding the newest files can also be done by sorting the tree by the "last modified" column and then opening the topmost branches. This is often useful to find out where some browser dumped that last download.

    Similarly, to find the oldest files, enable the "oldest files" tree column, sort by that and open branches until you can see a file.

    Files with multiple hard links or sparse files were mentioned in the log; otherwise they were not so easy to find (short of using the command line, of course).

• 2020-10-23

  • Now showing the target of symbolic links in the details panel.

    If it's a short path, the whole path is shown; otherwise without the path (".../somewhere"), and the full path is shown as a pop-up upon mouse click.

    If the target does not exist, a Broken Link warning is also shown.

• 2020-10-15

  • Enabled Refresh Selected now for files as well. In that case, the parent directory is refreshed.

  • Added hotkey [F6] for Refresh Selected.

    [F5] is still Refresh All like in all web browsers. Since window managers / desktop environments tend to consume [F5] with modifier keys ([Shift] [F5], [Alt] [F5], [Ctrl] [F5]), this is the closest we can get, and it's more consistent than using something like [Ctrl] [R].

    This was inspired by the discussion in PR#145.

• 2020-08-03

  • Now ignoring the loopback mounts of installed snap packages in the "Open Directory" dialog and in the "Mounted Filesystems" window.

    Yes, each of them has a separate loop mount, even if it's only installed, not in active use. Those mounts clutter the output of commands like df or mount with nonsensical cruft:

    df -hT | grep snap
    
    /dev/loop0  squashfs  159M  159M  0 100% /snap/chromium/1244
    /dev/loop1  squashfs   55M   55M  0 100% /snap/core18/1880
    /dev/loop2  squashfs   63M   63M  0 100% /snap/gtk-common-themes/1506
    /dev/loop3  squashfs   30M   30M  0 100% /snap/snapd/8542
    
    

    (From a freshly installed Xubuntu 20.04 LTS)

    Thanks a lot. Time to extend my df shell alias to exclude filesystem type "squashfs", too:

    alias df='/bin/df -x tmpfs -x devtmpfs -x squashfs -h $*'
    

• 2020-07-26 New stable release: 1.7

  Summary:

  • Closing the gap between sizes reported by QDirstat and sizes reported by the du command: Now also taking the allocated size into account.

  • Now also displaying the allocated size where it makes sense.

  • New "Mounted Filesystems" window showing output similar to the df command (but without the cruft).

  • New directory selection dialog showing all (real) filesystems.

  • New checkbox to cross filesystems (temporarily) in the directory selection dialog.

  • Workaround for NTFS oddities: Ignoring hard links on NTFS.

  • Added config option to ignore hard links in general.

  • Added a document about Btrfs free size and how different tools tend to show different values: Btrfs-Free-Size.md

  • Bug fixes

  Details:

  • Closing the gap between sizes reported by QDirstat and sizes reported by the du command: Now also taking the allocated size into account.

    There were repeated threads on various social media where users wondered why QDirStat displayed different sizes than the du or the df commands, sometimes slightly different, sometimes off by quite a lot.

    In short, this is mostly due to the difference between a file's byte size and its allocated blocks: On most filesystem types, a file of 49 bytes still consumes a minimum of 4 KB (4096 bytes); disk space is allocated in clusters (typically 4 KB), not byte by byte. On a typical Linux root filesystem with many thousands of tiny files, this difference can add up to a lot.

    Formerly, QDirStat only added up the byte sizes, disregarding the allocated, but really unused part at the end of tiny files. The rationale was that some filesystem types handle that in more intelligent ways, yet there is no documented way to get information from a filesystem if it actually does that. It turned out that in reality, most of them don't bother; they simply let most of that last cluster go to waste.

    Now QDirStat displays both sizes where useful:

    

    • For tiny files, the tree view now shows both sizes: "49 B (4k)"

    • For directories, the tree view now shows the total of the allocated sizes. This may sound a bit inconsistent, but it feels very natural and intuitive: You are typically interested in how much disk space the subtree consumes, no matter if some part of that is really wasted.

    • The details view shows both sizes when there is a difference (in the displayed numeric precision): "Size: 42.1 MB" vs. "Allocated: 42.2 MB", but not "Size: 42.0 MB" vs. "Allocated: 42.0 MB".

    • Treemap tiles now correspond to a file's allocated size. This makes a real difference for tiny files.

    More details at GitHub Issue #134: Size Difference between QDirStat Reports and the "du" and "df" Commands.

  • New "Mounted Filesystems" window ("View" -> "Show Mounted Filesystems" or Ctrl-M):

    

    This shows information about "normal" mounted filesystems (excluding system mounts like /dev, /proc, /sys, bind mounts, Btrfs subvolumes) with usage data as reported by the filesystem itself. The "Reserved" column shows the disk space reserved for the root user on that filesystem, "Free" is the available space for non-privileged users.

    Sometimes the "Used" size reported here may be different from what QDirStat reports after reading that complete filesystem; this can be due to metadata, journals or snapshots. Notice that in those cases, the du command will also display a different value than the df command.

  • QDirStat now has its own custom directory selection dialog; it no longer uses one of the simplistic Qt standard file dialogs:

    

    The "Places" bar on the left shows your home directory and all "real" filesystems (the same as in the new "Mounted Filesystems" window). For each filesystem, it shows

    • The mount point
    • The total size of the filesystem
    • The filesystem type ("ext4", "btrfs", "ntfs", ...)
    • The device name ("/dev/sda2") (in a tooltip)

    In the "Path" combo box you can simply edit the path or copy and paste it from another window. Use the "Up" button to move one directory level up.

  • The "Cross Filesystems" checkbox lets you temporarily override the global configuration option of the same name: QDirStat will then no longer stop when a mount point is found during reading a subtree, it will descend into that mounted filesystem and read it, too; but again, only for "real" filesystems, not for system mounts such as /dev, /sys, /proc, not for bind mounts, not for filesystems mounted multiple times, and not for network mounts (NFS, Samba / CIFS).

    More details at GitHub Issue #129.

  • Added a workaround for wrong size sums on NTFS: Now disregarding hard links on NTFS.

    The (current?) ntfs-3g implementation using fuseblk seems to disagree with Windows tools which files really have hard links.

    It appears that ntfs-3g regards even the MS-DOS compatible short filename (PROGRA~2 vs. Program Files) as a hard link which is of course utter nonsense; that means that almost all files on an NTFS partition are reported as having multiple hard links, so QDirStat displayed them as having only half their real size.

    Notice that this is a bug in ntfs-3g, not in QDirStat.

    See also GitHub Issue #88.

  • Added a config option to ignore hard links.

    This is useful for a very small number of use cases. Hard links are not very common anymore in today's Linux / BSD / Unix-like systems, so most users won't have to bother with this at all.

    By default, QDirStat sums up the disk space for a file with multiple hard links for each hard link's share of the overall size: If a file with 1 MB has 4 hard links, for each of those 4 links QDirStat adds 1/4 of the size (i.e., 256 kB) to the parent directory. If all those 4 links are in the same directory, that's very simple: They add up to 4 * 256 kB = 1 MB, so the sum is correct.

    If those hard links are all in different directories, each directory only gets part of that disk space allocated, because in fact they share the disk space among each other; the total disk space sum taking all those directories into account is still correct, of course.

    The trouble starts when you want to make a backup of only one of those directories: Even though the disk space is still shared with other directories, on the backup medium, you still need the disk space for that complete file, i.e. the full 1 MB, not only that directory's share (256 kB). With a lot of hard-linked files, that can add up to a lot of difference between what QDirStat displays and what disk space you actually need for the backup.

    There was a user who makes heavy use of that, and for that kind of use case there is now the option to ignore hard links: In that case, QDirStat sums up the complete size (the full 1 MB) for each hard link of the file.

    While that is useful for this special case, and you can now see the total size that you will need for your backup medium for that one directory, the total size higher up in the directory tree where more than one of those directories that share hard linked files with each other is off: That file now appears 4 times with 1 MB each, so it will add up to 4 MB.

    There is now a new config option in ~/.config/QDirStat/QDirStat.conf:

    [DirectoryTree]
    ...
    IgnoreHardLinks=false
    

    This is intentionally not available in the GUI config dialog to avoid confusion; use a text editor (while QDirStat is not running) to change this setting.

    Please use that config option only when you are aware of the consequences; this is a specialized option for rare, specialized use cases. It basically makes sense only if the other hard links are all outside the subtree that QDirStat displays.

    More details at GitHub Issue #124.

  • There is now a new document about Btrfs free size and how different tools tend to show different values: Btrfs-Free-Size.md

    Whenever users run out of disk space on Btrfs (which happens a lot, mostly because of snapshots) and they try to figure out where all their disk space went, they are confused about different tools reporting totally different and inconsistent sizes; traditional Linux / Unix command line tools like du and df just like GUI tools like QDirStat. Hopefully, this document will shed some light on that.

  Bug Fixes:

  • Fixed the internal cache writer (the one called from the File menu, not the qdirstat-cache-writer Perl script): For files with multiple hard links, it wrote the wrong size to the cache file: The result of size / links, not size.

    This was part of GitHub Issue #124.

  • Fixed crash when terminating with Ctrl-Q while reading directories (GitHub Issue #122).

  • Now automatically reopening the old current branch in the tree view when using actions like "refresh selected" and "continue reading at mount point (GitHub Issue #135).

  • Not a bug, but a minor new feature: Now automatically expanding the first directory level of a mount point after "continue reading at mount point".

  • Now correctly labelling a mount point as mount point in the "Details" panel.

See DevHistory.md for older entries.

History

This is just a rough summary. For more details, see DevHistory.md.

• 2020-07-26 New stable release: 1.7

  • Closing the gap between sizes reported by QDirstat and sizes reported by the du command: Now also taking the allocated size into account.

  • Now also displaying the allocated size where it makes sense.

  • New "Mounted Filesystems" window showing output similar to the df command (but without the cruft).

  • New directory selection dialog showing all (real) filesystems.

• 2020-02-13 New stable release: 1.6.1

  • Much better handling for "permission denied" errors while reading directories

  • Now showing the exact byte size (134 495 994 Bytes instead of 128.3 MB) upon mouse click

  • New optional tree column "Oldest File" (not enabled by default)

• 2019-07-22 New stable release: V1.6

  • New packages view

  • New unpackaged files view

  • Performance improvements while reading directories

  • Vast Performance improvement for huge directories (100.000+ entries in a single directory)

• 2018-11-07 New stable release: V1.5

  • New details panel

  • Package manager support to show what software package a file belongs to

  • New breadcrumbs navigation

  • Switchable tree layouts L1 / L2 / L3

• 2017-06-04 New stable release: V1.4

  • New file size statistics with histogram

  • Shading for empty space in the treemap for lots of very small files

• 2017-03-05 New stable release: V1.3

  • New file type view

  • Locate files of a certain type (filename extension) in the tree

• 2017-01-03 New stable release: V1.2

  • Improved Btrfs subvolumes support

• 2016-10-31 New stable release: V1.1-Pumpkin

  • Bug fixes

  • Split up the config file into several ones

• 2016-05-16 First stable release: V1.0

• 2016-04-08 Beta 3 release

• 2016-03-20 Beta 2 release

• 2016-02-06 Beta 1 release

• 2015-11-28 QDirStat project start: Ported from the old KDE 3 KDirStat

• Predecessor: KDE 3 KDirStat

  • 2006-06-01 KDirStat 2.5.3: The last KDE3 based version.

  • 2003: Bernhard Seifert wrote WinDirStat based on the KDirStat idea of coupling a tree view and a treemap and providing cleanup actions.

  • 2003-01-05 KDirStat 2.3.3: Treemaps

  • 2002-02-25 KDirStat 2.0.0: Complete rewrite for KDE 2 / Qt 2

  • 2000-01-21 KDirStat 0.86 for KDE 1 announced: First public version.

Related Software

KDirStat and QDirStat

KDirStat was the first program of this kind (combining a traditional tree view with a treemap), also written by the same author as QDirStat. It was made for KDE 1 back in early 2000; later ported to KDE 2, then KDE 3.

QDirStat is based on that code, but made independent of any KDE libraries or infrastructure, so it has much fewer library and package dependencies; basically only the Qt 5 libs and libz, both of which most Linux / BSD machines have installed anyway if there is any graphical desktop installed.

WinDirStat and QDirStat

There are lots of articles and user forum comments about QDirStat being a "nice Linux port of WinDirStat". Well, nothing could be further from the truth: WinDirStat is a Windows port of KDirStat, the predecessor of QDirStat.

So it's the other way round: The Linux version was there first, and somebody liked it so much that he wrote a Windows version based on that idea. That's a rare thing; usually people port Windows originals to Linux.

See also https://windirstat.net/background.html and the WinDirStat "About" dialog.

QDirStat and K4DirStat

K4DirStat is a port to KDE 4 / Qt 4 of the old KDE 3 / Qt 3 KDirStat. QDirStat is independent of that; it is based on the old KDE 3 KDirStat directly.

Other

• Baobab
• Filelight
• ncdu
• du

See Disk Usage Tools Compared: QDirStat vs. K4DirStat vs. Baobab vs. Filelight vs. ncdu (including benchmarks) in the Wiki.

Motivation / Rant: Why?

After having used KDE since its early days (since about 1998), I didn't like the direction anymore that KDE has been taking. I loved KDE 1, KDE 2, KDE 3. When KDE 4 came along, it took me a long time to try to adopt it, and when I did, I moved back to KDE 3 after a short while, then tried again with the next release, moved back again -- several times.

I really tried to like it, but whenever I thought I tamed it to meet my requirements, a new version came along that introduced yet another annoyance.

To name a few:

• A lot of things that used to be user configurable in KDE 3 are not configurable anymore, and when you approach the KDE 4/5 developers about that, they will tell you that this is intentional, and they do not intend to bring those config options back. Well, thanks a lot; this is the Apple approach where they think they know what is good for you, and you are just too stupid.

• Konqueror as the old central tool is as good as dead. It's still there as an alternate file manager (for those who find it), but the primary one is the dumbed-down Dolphin that I consider unusable: It's only useful for complete newbies, not for power users. The web browser part of Konqueror is so outdated that you can't do much with it with most modern web sites, so the great integration of web and local file manager that was the major strong point of Konqueror (and thus KDE) no longer exists.

• I don't like the fact that I can't simply put icons on my desktop anymore -- no, I have to create a plasmoid first as a container, and those things keep doing weird stuff that drives every user crazy. With one false move of your mouse, it might be gone, change shape, move to another place or whatever.

• I also don't like the desktop search that eats resources like there is no tomorrow (disk space, disk I/O, CPU usage) and that for all practical purposes you can't get rid of.

• I don't like the fact that the mail client relies on that MySQL based framework called Akonadi that is not only resource-hungry, but also so fragile that I had to use the akonadiconsole lots of times just to bring it back to life. Seriously, if I as a Linux system developer have a hard time doing that, what is a normal user expected to do?

• Activities vs. multiple desktops. I tried to use both, but they don't integrate well. The desktops previewer is far inferior to the old one from KDE3: Only monochrome rectangles, no real preview. The activities plasmoid keeps rearranging my carefully placed and named activities at random. WTF?!

• Everything is so fragmented that not even the naming is clear anymore. What used to be KDE is now a jumble of the KF Framework, the KF libs, the KF apps and the Plasma desktop. Yeah, great job, folks; people used to know what KDE stood for. Nobody knows what the hell all those components are, and neither does anybody care anymore. You paved your way to oblivion with buzzwords. Great marketing strategy for gaining more visibility!

Then the next generation KDE arrived, Plasma 5. When I was force-migrated to it at work with the SUSE Tumbleweed rolling release, the experience was so bad that I moved to the Xfce Desktop.

Now every time I started my own KDirStat, it started about a dozen KDE processes along with it -- processes that it needs only for minor things like loading icons or translations. I really don't need or want that.

So it was time to make KDirStat self-sufficient; it never used that much of all the KDE infrastructure anyway. Time to make a pure Qt-based and self-sufficient QDirStat.

And while I was at it, I took the chance to add some features that I had wanted for a long time, yet I had never gotten myself to start working on:

• Multi-selection in the directory tree so you can delete several files at once.

• Remove limitations like having only a fixed number of user-defined cleanup actions.

• Properly show the output of cleanup actions, in particular when they reported errors.

• Make treemap colors configurable: Use custom colors and match them to user-defined filename extensions.

• Move away from the arcane KDE build system: Back with KDE 1/2/3 it was the Autotools with custom KDE extensions that only a handful people in the world really understood (I was not among them), later CMake which is little better, just differently confusing.

Yes, there is a Qt4 / Qt5 port of KDirStat called K4DirStat. K4DirStat is an independent project that started when I had not worked on the old KDirStat for a long time (my last KDirStat release had been in mid-2006).

QDirStat is based on that same code from the 2006 KDirStat. It's an 80% rewrite using a lot of newer Qt technologies. And there was a lot of cleaning up that old code base that had been long overdue.

Features

New Features

• Multi-selection:

  • Both views (the tree and the treemap) now support extended selection, i.e. you can select more than one item. This was the most requested feature for the last KDirStat. Now you can select more than one item at the same time to move it to the trash can, to directly delete it or whatever.

  • Tree view:

    • Shift-click: Select a range of items.
    • Ctrl-Click: Select an additional item or deselect a selected one.

  • Treemap:

    • Ctrl-Click: Select an additional item or deselect a selected one.

    • The current item is highlighted with a red rectangle, all other selected ones with a yellow rectangle. If the current item is not also selected, it has a dotted red outline.

• Proper output of cleanup actions with different colors for the commands that are executed, for their output and for error messages (see screenshot above). That output window can be configured to always open, to open after a certain (configurable) timeout, or only if there are error mesages -- or not at all, of course. If things go wrong, you can kill the external command started by the cleanup action from there. You can zoom in and out (increase or decrease the font size) as you like.

• File type statistics window. WinDirStat has it, and users wanted it in QDirStat, too. Since filename extensions (suffixes) don't have as much semantics in Linux/Unix systems as they do in Windows, many files are categorized as "Other". This is a known limitation, but it's a limitation of the whole concept of using suffixes to categorize files by type. And no, checking file headers for magic byte sequences like the "file" command does is not an option here; QDirStat would have to do that for (at least) all the 30,000+ files typically listed under the "Other" category. So we'll have to live with that limitation.

• Locate files by file type window. If you double-click on any of the filename extensions (suffixes) in the file type statistics window, you will get another window that lists all the directories that contain files of that type including the number and total size of those files. You can double-click each of those lines, and that directory will open in the main window with the files of that type preselected so you can start cleanup actions like moving them to trash or converting them to a better format (.bmp -> .png) immediately.

• File size statistics window with histogram, percentiles, buckets and a lot of documentation that everybody should be able to understand. Even if (or, better yet, in particular if) your math teacher or statistics professor never explained it properly, please have a look at it.

• Packages view: Show installed packages and their files in the tree. Supported for all Linux distributions using any of dpkg, rpm, pacman as their low-level package manager or any higher-level package manager like apt, zypper etc.; more details at Pkg-View.md.

• Unpackaged files view: Show a directory tree, but ignore all files that belong to an installed software package. Those ignored files are displayed in a special branch in the tree view, and they are not displayed at all in the treemap. This is useful to find files that were manually installed by a sudo make install command. More details at Unpkg-View.md.

• New macros to use in cleanup actions:

  • %d : Directory name with full path. For directories, this is the same as %p. For files, this is their parent directory's %p.

  • %terminal : Terminal window application of the current desktop; one of "konsole", "gnome-terminal", "xfce4-terminal", "lxterminal", "eterm". The fallback is "xterm".

  • %filemanager : File manager application of the current desktop; one of "konqueror", "nautilus", "thunar", "pcmanfm". The fallback is "xdg-open".

• Which desktop is used is determined by the $XDG_CURRENT_DESKTOP environment variable. Users can override this with the $QDIRSTAT_DESKTOP environment variable, so you can get, say, the Xfce terminal or file manager despite currently running KDE if you set

  export QDIRSTAT_DESKTOP="Xfce"
  

• Of course, you can still simply use your favourite file manager if you simply change %filemanager in the default "Open File Manager Here" cleanup action to the command to start it.

• You can now select the shell to use for the cleanup commands:

  • $SHELL (the user's login shell) - using the same environment, syntax and wildcard etc. behaviour of the shell the user is used to.
  • /bin/bash for well-defined behaviour for wildcards etc.
  • /bin/sh as a last resort (which might be a simplistic dash on Ubuntu).

• Mouse actions in the treemap window:

  • Left click: Select item and make it the current item.
  • Right click: Open the context menu with cleanup actions and more.
  • Ctrl+Left click: Add item to selection or toggle selection.
  • Middle click: Select the current item's parent. Cycle back at toplevel.
  • Double click left: Zoom treemap in.
  • Double click middle: Zoom treemap out.
  • Mouse wheel: Zoom treemap in or out.

• You can configure what columns to display in the tree view and in which order. The only thing that is fixed is the "Name" column which is always there and always the first (leftmost). Use the context menu in the tree header to unlock column widths. Drag columns to the left or right to change their order.

• Exclude rules are now greatly simplified. They no longer always get the entire path to match which requires quite complex regexps; by default, they only get the last path component -- i.e., no longer "/work/home/sh/src/qdirstat/src/.git", but only ".git". You can now even tell the exclude rule to use a simplified syntax: "FixedString" or "Wildcard" in addition to the normal "RegExp". The old behaviour (matching against the full path) is still available, though.

• Configuration dialog for exclude rules -- see screenshots.

• Subvolume detection for Btrfs. Btrfs subvolumes are just ordinary mount points, so normally QDirStat would stop scanning there, leaving a large part of a Btrfs partition unaccounted for. But for each mount point found while scanning a directory tree, QDirStat checks /proc/mounts or /etc/mtab if it has the same device name as its parent directory, and if yes, considers it a subvolume and continues scanning.

• Actions to go one directory level higher or to the toplevel: Context menu and menu "Go To" -> "Up One Level" or "Toplevel". This is useful if you clicked on a file in the treemap that is deep down in some subdirectory, and you want to know what subdirectory that is: Simply click "Go Up" twice (the first click will get you to the pseudo subdirectory, the second one to the real one).

• Open all tree branches up to a certain level and close all other ones: Menu "View" -> "Expand Tree To Level" -> "Level 0" ... "Level 9".

• The total sum of the selected items (subtrees) is displayed in the status line if more than one item is selected.

• Icons are now compiled into the source thanks to Qt's resource system; now it's just one binary file, and nothing will go missing. No more dozens of little files to handle.

• The build system is now Qt's QMake. I got rid of that AutoTools (Automake, Autoconf, Libtool) stuff that most developers find intimidating with its crude M4 macro processor syntax. QMake .pro files are so much simpler, and they do the job just as well. And no, it will definitely never be CMake: I don't like that thing at all. It's just as much as a PITA as the AutoTools, just not as portable, no usable documentation, it's changing all the time, and those out-of-source builds are a royal PITA all on their own with constantly having to change back and forth between source and build directories.

• QDirStat now has its own log file. It now logs to /tmp/qdirstat-$USER/qdirstat.log (where $USER is your Linux user name). No more messages on stdout that either clobber the shell you started the program from or that simply go missing.

• No longer depending on dozens of KDE libs and a lot of KDE infrastructure; it now only requires Qt which is typically installed anyway on a Linux / BSD / Unix machine with any X11 (graphical) desktop.

• It should still compile and work with Qt4. We now have a contributor who is very interested in that (Michael Matz), so it should be possible to maintain this compatibility.

• Slow down display update from 333 millisec (default) to 3 sec (default) with qdirstat --slow-update or qdirstat -s. The slow update interval can be customized in ~/.config/QDirStat/QDirStat.conf:

  [DirectoryTree]
  SlowUpdateMillisec = 3000
  

Old Features

Features ported from the old KDirStat:

• Fast and efficient directory reading.

• Not crossing filesystem boundaries by default so you can see what eats up all the disk space on your root filesystem without getting distorted numbers due to all the other filesystems that are mounted there. If you absolutely wish, you can use "Continue reading at mount point" from the context menu or from the "File" menu -- or configure QDirStat to always read across file systems.

• Efficent memory usage. A modern Linux root filesystem has well over 800,000 objects (files, directories, symlinks, ...) and about 100,000 directories. This calls for minimalistic C++ objects to represent each one of them. QDirStat / KDirStat do their best to minimize that memory footprint.

• Hierarchical tree view that displays accumulated sums in each branch, together with a percent bar so you can see at a glimpse how the subdirectories compare with each other.

• All numbers displayed human readable -- e.g., 34.4 MB instead of 36116381 Bytes.

• All size units are 1024-based, i.e. 1 kB = 1024 Bytes; 1 MB = 1024 kB; 1 GB = 1024 MB.

• In the tree, also displaying the exact byte size as the context menu (right click).

• Each tree level uses another color for that percent bar so you can easily compare subdirectories even if some of them are opened in the tree.

• If a directory has files and subdirectories, all files in that subdirectory are grouped into a pseudo directory (called dot entry in the QDirStat sources) so you can compare the disk usage of files on that directory level with the subdirectories.

• Displaying the latest modification time of any object in each branch. You can instantly see in what subdirectory where any changes lately. You can sort by this column, of course.

• Treemap display. Treemaps are a way to visualize hierarchical data structures, invented by Ben Shneiderman. Basically, the hierarchy is flattened and each level grouped in a rectangle, inside which it is again subdivided in rectangles. The area of each rectangle corresponds to the size of each item or subdirectory. For the purposes of QDirStat, it is enough to know that a large blob corresponds to a large file; you can instantly see where large ISOs or movies are.

• You can zoom the treemap in and out (Ctrl + / Ctrl - / mouse wheel / menu / tool bar) to see more details of directories that are otherwise dominated by larger ones.

• You can move the boundary between treemap and tree view up and down as you like. You can also get rid of the treemap completely (menu "Treemap" -> "Show Treemap" or F9 key)

• Treemap and tree list view communicate. Select an item in one view, and it is also selected in the other. If you click on that large blob in the treemap, it is located in the tree view, all branches up to its directory are opened, and the tree view scrolls to that item.

• Cleanup actions. Once you know what is consuming the disk space, you can start cleanup actions from within QDirStat to reclaim disk space - or to investigate further if you can safely delete a file. You can create your own cleanup actions (as many as you like), and there are some predefined ones:

  • Open file manager here. This will start a file manager in the directory of the current item. QDirStat tries its best to guess the name of the relevant file manager application for the current desktop, based on the $XDG_CURRENT_DESKTOP environment variable. You can override this with the $QDIRSTAT_DESKTOP environment variable.

  • Open terminal window here. In most cases, this is much easier than to navigate to that directory with 'cd' in an already open terminal window and using tab-completion numerous times. As with the file manager application, QDirStat tries its best to guess the name of the relevant terminal window application for the current desktop.

  • Move to trash bin. QDirStat has its own implementation of the XDG trash specification.

  • Delete immediately.

  • Compress: Create a compressed tar archive from a directory and then delete the directory.

  • Delete junk files: Backup files left behind by editors, core dumps.

  • All predefined cleanup actions are fully configurable, of course. You can change any of them, disable them, or delete them.

• You can copy the complete path of the selected file or directory to the system clipboard and paste it to another application.

• Reading and writing cache files:

  • This is mostly meant for remote servers in some server room somewhere: Rather than installing the Qt and X11 runtime environment and running QDirStat over remote X (ssh with X forwarding), you can run the supplied qdirstat-cache-writer Perl script on the server, copy the resulting cache file to your desktop machine and view the content there with QDirStat.

  • For large directories (archives etc.) that don't change that much, you can also generate a QDirStat cache file (either with the Perl script or with QDirStat itself) and save it to that corresponding directory. If QDirStat finds a file .qdirstat.cache.gz in a directory, it checks if the toplevel directory in that cache file is the same as the current directory, and if it is, it uses the cache file for that directory rather than reading all subdirectories from disk. If you or the users of that machine use QDirStat often, this might take a lot of I/O load from the server.

  • If you use the '-l' option of the qdirstat-cache-writer script, it uses the long file format with a complete path for each entry, so you can use the zgrep command with it as a replacement for the locate command.

  • The KDirStat / QDirStat file format is well documented and very simple. It seems to be used by a number of admins and some backup software. See also the specification in the doc/ directory: https://github.com/shundhammer/qdirstat/blob/master/doc/cache-file-format.txt

  • You can specify a cache file to read directly at the command line:

    qdirstat --cache cache-file
    

• Other command line options: See

  qdirstat --help
  

Features that are Gone

(Compared to the old KDirStat)

• Pacman animation
• KIO slave support
• Feedback form

MacOS X Compatibility

find . -name qmake

qmake
make

There be dragons. ;-)

Architecture maintainer wanted for QDirStat for MacOS X

If you are a developer with some prior C++ and Qt knowledge on the MacOS X platform and you'd like to see QDirStat working there, please consider joining the team.

Windows Compatibility

None for the forseeable future. Use WinDirStat.

Ready-made Packages

openSUSE / SUSE Linux Enterprise

QDirStat packages for openSUSE Tumbleweed / Leap (15.x, 42.x) and SLE (15, 12):

• Download page for the latest stable release

• Download page for the current development version (git master)

  Since this version is in development, it may be not quite as stable and reliable as the latest official stable release, although the QDirStat developers try their best to keep it as stable as possible.

Ubuntu

• Recent releases (Artful Aardvark (17.10) or later):

  https://packages.ubuntu.com/search?keywords=qdirstat&searchon=names

• Older releases (Pre-17.10):

  Use Nathan Rennie-Waldock's QDirStat PPA

Debian

https://packages.debian.org/search?keywords=qdirstat

Fedora

https://apps.fedoraproject.org/packages/qdirstat/builds/

Building

Notice that for most mainstream Linux or BSD distributions you don't have to build your own; you can simply install a ready-made package from your normal package manager.

Build Environment

Make sure you have a working Qt 5 build environment installed. This includes:

• C++ compiler (gcc recommended)
• Qt 5 runtime environment
• Qt 5 header files
• libz (compression lib) runtime and header file

If anything doesn't work, first of all make sure you can build any of the simple examples supplied with Qt, e.g. the calculator example.

Ubuntu

Install the required packages for building:

sudo apt-get install build-essential qtbase5-dev zlib1g-dev

Dependent packages will be added automatically.

Recommended packages for developers:

sudo apt-get install qttools5-dev-tools qtbase5-doc qtbase5-doc-html qtbase5-examples

See also

http://askubuntu.com/questions/508503/whats-the-development-package-for-qt5-in-14-04

If you also have a Qt4 development environment installed, select the desired one via qtchooser:

sudo apt-get install qtchooser
export QT_SELECT="qt5"

SUSE

Install the required packages for building:

sudo zypper install -t pattern devel_C_C++
sudo zypper install libQt5Widgets-devel libqt5-qttools zlib-devel

If you also have a Qt4 development environment installed, make sure that the Qt5 version of 'qmake' is the first in your $PATH:

export PATH=/usr/lib64/qt5/bin:$PATH

Compiling

Open a shell window, go to the QDirStat source directory, then enter these commands:

qmake
make

Installing

sudo make install

or

su -c make install

Install to a Custom Directory

The default setup installs everything to /usr. To install to another directory, set INSTALL_PREFIX during qmake.

qmake INSTALL_PREFIX=/usr/local

Contributing

See file Contributing.md and GitHub-Workflow.md

Troubleshooting

Can't Move a Directory to Trash

See file Troubleshooting.md

Further Reading

• Original KDirStat (source code)
• K4Dirstat (source code)
• WinDirStat (for Windows)
• Disk Usage Tools Compared: QDirStat vs. K4DirStat vs. Baobab vs. Filelight vs. ncdu (including benchmarks)
• XDG Trash Spec
• Spatry's QDirStat Review on YouTube

Of course, don't forget to check out the doc directory.

Packaging Status

Repology: QDirStat versions in Linux / BSD distributions:

(click for details)

Donate

QDirStat is Free Open Source Software; you are not required to pay anything. If you want to support continued development of QDirStat, donations are most welcome.

Donate any amount you like via PayPal: