Future plans and TODOs of LXDE in 2010


As many people have seen, LXDE is currently underwent active development, but most of the visible problems are not yet solved and many distros are still shipping out-dated versions of LXDE components. So I think it’s time to give a more clear roadmap about the plans in 2010. Here’s are problems to solve this year:

  • The file manager: PCMan FM 0.5 is quite broken in many parts, and I know that.
    1. Desktop management should be improved:
      1. Icons should be movable.
      2. “My document” icon is a mistake and should be removed or be optional
      3. Drag and drop doesn’t work sometimes
      4. Handling of desktop working area has some bugs in it
    2. Mounting for volumes is broken:
      1. First, I must say, this is not our fault at all and working on this part is a real pain. Recent rapid incompatible changes in freedesktop.org, HAL, and some distros break all older applications previously working well. If your program doesn’t work, complain to them.
      2. HAL will soon be removed from modern Linux and will be replaced by DeviceKit-disks.
      3. What a joke! While devicekit-disk is still under development, they now plan to drop it and replace it with udisk based on libudev, which is currently Linux-only.
      4. So now you can understand why big software companies don’t like to develop Linux versions for their product. All your effort just gets ruined again and again due to frequent incompatible changes in upstream. They just don’t care about breaking others’ work. One of my programs written for Microsoft Windows 10 years ago still work flawlessly today.
      5. In a world governed by two big desktop environments, we smaller projects have no choice but follow their “standards”. Otherwise your software just won’t work on major distros. Even if you are lucky enough and your software is not broken by them, it will soon be broken.
      6. Conclusion: Sorry for saying this, but if we want to co-exist and keep compatible with other new GTK+ software, I think for now we had better use gvfs and accept its gnome-disk-utility dependency. It’s a hard choice, but it seems that we have no choice now. After major distros are upgraded and adopt new freedesktop.org standards/specs, old programs which work well in the past, like the old pcmanfm with HAL support, will stop working. Of course we can directly use udisk and prevent using gnome-disk-utility, but I already tired of fixing my program again and again just because they upstream frequently make arbitrary incompatible changes and break my program. Since they are the people who break the compatibility, they should be responsible for fixing it (in gnome-disk-utility). What if I finally spent my time to implement udisk support, and later they decided to drop udisk and replace it with new stuff again? I don’t want to waste my time on fixing incompatibility again and again.
      7. Complain to your distro or to freedesktop.org developers if you don’t like the situation, but most of the time you may just get ignored. Like it or not, welcome to the real world.
    3. The “Find Files” tool is broken and I know that. With new multi-threading infrastructure of libfm, this will be solved (in later releases).
    4. Lightweight gvfs fork:
      1. GVFS will bring many dependencies, but since it’s free software, we can fork it and remove some gnome dependencies.
      2. This is possible, but the priority is low unless we get more developers.
  • LXPanel, The desktop panel:
    1. While 0.5 series bring new features, it has some new bugs, too. This part requires more testing. After finishing the development of the file manager, lxpanel should be fixed.
  • Improve translation workflow:
    1. We may evaluate the possibility of using Transifex when online editing with correct pleural form support is available.
  • Cleanup the bug trackers
    1. We need to set a “bug day” and do our best to fix fixable bugs.
    2. New rules/guidelines for bug reporting is needed
    3. Duplicated, invalid, or out-dated bugs should be removed.
    4. Many bugs in pcmanfm will be solved by the newly developed libfm and will be closed when new pcmanfm is released.
  • More system tools:
    1. Maybe we need lxde-control-center
    2. Is it possible to integrate this with the file manager?
  • Display manager – LXDM:
    1. Dgod is developing a new display manager named lxdm and this can replace gdm in the future.
    2. Theming support is a must-have
  • Desktop launcher – LXLauncher:
    1. Background image support is not yet added because the time is now spent on the file manager, but we’ll do it this year.
  • Benchmarks, more testings: Can anyone help this part?
  • Code cleanup and optimization:
    1. Fix compiler warnings
    2. Profiling
    3. Developer is needed. Can anyone help this part?

If you like the idea of LXDE, please help the development and join our mailing list for discussion.

If you’re a programmer familar with GTK+, please join us. If you’re not familar with GUI programming, but knows C language and autotools, please help us fix compiler warnings and do profiling parts. If you don’t know programming, please join our translation team to make LXDE available in your language.

We believe that LXDE will be a mature desktop solution in 2010.


9 responses to “Future plans and TODOs of LXDE in 2010”

  1. > Lightweight gvfs fork:
    > 1. GVFS will bring many dependencies, but
    > since it’s free software, we can fork it and
    > remove some gnome dependencies.

    Why do you insist that GVFS has many Gnome dependencies? The only required dependencies are D-Bus and GLib. See comment #4 here:

    http://blog.lxde.org/?p=541#comments

  2. I really want to test arch+lxde on iso.
    can you provide “arch+lxde live cd”?

  3. I glad support this project and developed in parts things.

  4. @anonymous: it depends on many GNOME packages. At least on Arch, these are gnome-disk-utility, gnome-keyring, libsoup-gnome and, most importantly: _gconf_. I have an gconf-free system, and I intend to keep it like that, I disagree with its way and I don’t want my desktop to depend on it. Pure and simple.

    @tiik: please ask in the Arch forums, there are some automatic tools to build live-cd’s. Archers are usually not interested in such things, though, because it is a modular system (actually like LXDE) and no one trusts “standard” selections.

    @pcman: after reading your announcement, I decided to move over from XFCE to LXDE. I always liked your principles but I had the feeling that this project is a bit abandoned. I’m loving it and I hope I’ll help in the future, thanks for your work!

  5. @mihaim: If Arch introduces needless dependencies to GVFS, then that is an Arch problem, not a GVFS problem. GVFS compiles and runs here just fine *without* gconf and other Gnome stuff. Out of the box.

  6. @anonymous: Yes, gvfs is modular. But if you read the blog post, PCMan intends to have gnome-disk-utility handle mounting in PCManFM. And well, gnome-disk-utility brings in gnome dependencies, specifically gnome-keyring and gconf. I don’t like this either, I would prefer for PCManFM to talk directly to udisks.

    @PCMan: Udisks *is* Devicekit-disks. It’s the exact same thing, just a new name.

  7. I like the idea of the control center. But I think it should be a seperate tool and optional, like the rest of the desktop environment.