lubuntu-desktop team mailing list archive

[PCMan <pcman.tw@xxxxxxxxx>]

Thanks for the comment.
Some new features are indeed fixes for old bugs, especially usability bugs.
In addition, some new features require some degree of refactor and
this can affect bug fixes.
For example, in the file manager, after I added directory tree to the
side pane, an old patch in the tracker is no longer needed. Nor can it
be applied.
For the lxpanel stuff, I really think that we need to use libwnck
instead of using our own X11-handling code. Libwnck is well-tested and
is in production use for years and it's built for this kind of task.
Adding this little lib does not make LXDE significantly more
resource-hungry, but it can save much time for maintenance and
debugging. Replacing some taskbar and pager code with libwnck can fix
part of the bugs quickly.

For gtk+ 3 transition, I want to add src/gtk-compat.[ch] to every LXDE
components and add compatibility later there.
Some macros are replaced by functions in gtk+ 3 and this can be easily
fixed in a header with some macro definition.
Maybe later I'll add this to libfm/pcmanfm first.

Previously I tried to make a tool in lxde/devtools/min-lib-ver with
python. This tool is used to determine the minimal gtk+ version
required by the project, but I don't have time to finish it. I want to
scan the source tree and find out all gtk+ APIs used. Then, determine
minimal version of gtk+ required. This one can be very helpful for all
gtk+ using projects. Unfortunately it's half done.
Is there any tool to help gtk2 -> 3 transition?

On Mon, May 23, 2011 at 12:51 PM, Jonathan Marsden <jmarsden@xxxxxxxxxxx> wrote:
> On 05/22/2011 08:41 PM, PCMan wrote:
>
>> After this one is finished, I'll try to find some time to fix other
>> bugs in the file manager.
>
>> As for other lx-* stuff, I currently don't have enough time to fix
>> them. Really glad that you can help.
>
> OK.  Personally, I think that when developer time is really really
> limited (which it seems to be!), we should focus on the basics:
>
> (1) Fix all major bugs that are biting our users, and then
>
> (2) Code cleanup needed to ensure it builds and runs with newer
> compilers, libraries, etc. (e.g. GTK3 transition, and GCC 4.6 is in
> Debian sid already... so we need to check LXDE builds fine there, and
> fix it if not).
>
> After that, if we possibly can, we then do
>
> (3) occasional official source tarball releases (maybe every six months?
>  or every 12 months if every six is too much work).  This is so that the
> work that *is* getting done in (1) and (2), even if fairly small, is
> more easily available to users, and to other Linux distributions,
> without them having to dive into a git tree.
>
> [From a Ubuntu/Lubuntu perspective, we probably need a new
>
> *Anything* at all more than that, things like new enhancements, test
> suite, refactoring code, etc. should be considered a luxury, a *bonus*;
> we should not plan for it at all, unless we have the developer time to
> do it!
>
> This approach is not much fun for programmers (no new fancy
> functionality to design and code!), but IMO it is what allows a project
> to stay alive and remain at least somewhat useful to users.  Hopefully,
> our doing (1), (2) and (3) consistently and reliably (and well!) will
> attract more developers who like LXDE... so we can then do more of the
> fun stuff.
>
> Does this make sense?
>
>> 3. lxterminal: currently we have no maintainer for it. Previously a
>> friend on our mailing list is willing to take over it. So maybe we'll
>> have a new maintainer later. Another option can be using Roxterm
>> instead.
>
> Given my tendency to sit at the command line a lot, I *might* be able to
> take this on, but not yet, and this is not yet a committment, OK!  I am
> not much of a GUI/GTK programmer at this point.  But perhaps once
> Oneiric is out, I could think more seriously about taking maintainership
> of lxterminal on for the longer term.
>
>> These are my current plans. Any suggestions are appreciated.
>
> See above.  I think your plans are fine, but may be ambitious given how
> small a team we are.  I would suggest (you probably will not feel good
> about this!) that you work on bug fixes even in lx* programs before
> adding anything new at all to any of them.  Keeping current LXDE users
> at least reasonably happy is #1 priority, if you want the project to
> survive.  Users do not like significant bugs.  Sorry if that sounds a
> bit negative, but it would be my main suggestion after reading your plans.
>
> Jonathan
>