Basically, we only relod if the file is modified but still handle the case
where the watch gets lost. Maybe we should just ignore IN_IGNORE altogether?
(cherry picked from commit 085c39d394)
This makes sure valgrind doesn't show false memleaks that are caused by ncurses
instead of our code when we do a 'Debug' build. There is a also a disadvantage:
When your ncurses library is not build with the '--disable-leaks' option (this
is the case in the default version of ncurses of most distro's) and you build
conky with ncurses and debug turned on then you will see a segmentation fault
when you stop conky. On the other hand people that want to build conky as a
debug build with ncurses support will probably also have a ncurses with
--disable-leaks ...
Rewrote timed_thread library in C++ using fancy new C++0x features. The
main reason for this is to phase out poor encapsulation and C-style
function pointers.
It seems that closing xft fonts on reload is a bad idea, but this seems
to result in a memory leak. As far as I can tell, the leak doesn't grow
beyond the initial allocation however.
This patch adds the possibility to use a argb visual for the conky
window. Adding a boolean configuration entry 'own_window_argb_visual'.
the 'set_transparent_background' method sets the alpha bytes of the argb
color for transparency, instead of pseudo transparency. This patch
would close following feature request:
https://sourceforge.net/tracker/?func=detail&aid=1563931&group_id=143975&at
id=757311 and would be a workaround for some reported transparency bugs
(https://sourceforge.net/tracker/?func=detail&aid=1968923&group_id=143975&a
tid=757308 for example)
Signed-off-by: Brenden Matthews <brenden@diddyinc.com>
This should close sf #2850092.
Basically, before this change, for strings which includes
SECRIT_MULTILINE_CHAR the computed maximum substring width was getting
summed to the width of the last substring.
The old behaviour is kept but for the case of strings including
SECRIT_MULTILINE_CHAR, for which now the maximum width of all substrings
is taken.
I don't think this will break anything else but blame me if it does ...
- kill trailing whitespace (grep -n ' $' src/*.{c,cc,h})
- eliminate space before tab (grep -n ' ' src/*.{c,cc,h})
(insert tab by pressing CTRL-v first)
- little indenting fixup in configure.ac.in (we indent using tabs, not
spaces)
note: this should not change the actual code at all - if it does, feel
free to blame me personally ;)
