ewmh_process_window_type_atom(...) clears the border for windows that
are to be skipped. However, client_updatewmhints(...) doesn't.
Signed-off-by: Julien Danjou <julien@danjou.info>
Use a single script to generate the version number from git. This script
* strips the v from the start of the "git describe" output
* changes all "-" occurences to "."
* does NOT create a properly sortable version number from "2.1-rc2"
type "git describe" output
* does NOT generate a "proper" "1.2.3" type version number under any
circumstances
* will generate "2.1" in case the "git describe" output is "2.1"
These policy might need closer adaption to awesome's tagging habit
some time.
In dist tarballs, ship a "version-stamp" file with the package version
in it. If the "version-stamp" file is present (i.e. if it is a source
tree from a dist tarball), no git checks will be performed.
Concept from autoconf, but code written from scratch to match
awesome's requirements.
Signed-off-by: Julien Danjou <julien@danjou.info>
the graph_tell() does not modify directly the line-lenght to draw, rather
then calculating the value in percent and in the graph_draw, calculating
the line length ouf of that the percent.
(while trying to fix something little, a whole new scale thing resulted;
also I didn't really understand the old system, my fault that is)
I just noticed that international (reads Greek) text is not rendered at
all in widgets. I traced it down to LC_CTYPE being non-empty in my
environment. Are there any objections to adding the setlocale(...) call
back in awesome.c:main(...)? I could set up a script emptying LC_CTYPE
and then exec'ing awesome, but I guess it's harmless to have awesome do
that internally. Unless I'm missing something...
I tend to prefer to not see a square on the taglist for tags which only
have skipped windows. E.g., if I run 'stalonetray --sticky
-w/--skip-taskbar', in which case it appears on all tags but not on the
tasklist, showing all of the tags as occupied seems somewhat awkward.