This follows the escaped chars in bash completion and allows to pass
filenames with spaces etc.
Using quotes would be another option but requires special handling of
"~" and, what's worse, either hand-correcting the cursor position (into
the quoted area) or more completion mumbo-jumbo to handle the quotes.
While any window can be centered using the apps file, fbrun can serve many
purposes and sometimes (runner) makes sense being centered, sometimes
(button/menu triggered input) near the mouse, sometimes ("application")
regularily placed.
REQUEST: 282
the default is 1024-1025, values are read from the FBRUN_HISTORY_SIZE
environment variable
NOTICE: the limit isn't hard, but will typically be n+1 and only n if
the new entry is already present in the last n entries
REQUEST: 202
- streamline code
- indicate completion by making use of selection
- fix buggy behavior (notably subsequent completions and FS path
following)
- support "~" in paths
- support chunk completion
(ie. "mp[layer] ~/vid[eos/favporn.mp4]" can be completed in both
tokens; buggy with paths including spaces in non-leafs)
REQUEST: 223
xxx_LDFLAGS place the libraries like '-lX11' or '-lXft' at the beginning of
the linker command. Some systems were not able to handle the situation and
this lead to a lot of unresolved symbols. Reading the documentation of
automake:
... you can use LDADD to do so. This variable is used to specify
additional objects or libraries to link with; it is inappropriate for
specifying specific linker flags, you should use AM_LDFLAGS for this
purpose.
In addition we link against -lm in order to satisfy symbols refering to 'cos'
and 'sin'.
This commit fixes issues #72 (brought up + different solution by Mattias
Guns; I received a similar patch by 'Nable 80' via ML and discussed the
issue in #fluxbox with 'Nable 80'), patch #73 (Mattias Guns) and finally
patch #162 (Ulrich Eckhardt; this commit is heavily based upon Ulrich's
work).
The original code was overly complex. It tried to avoid writing
bytes to the disk at the expense of comprehensibility and as a result it
was buggy. I looked at both patches from Mattias and 'Nable 80' which address
the bug with skipping entries in the history-file (my fault: incorrect use
of outfile.ignore(1, '\n')): They provided a proper fix for the problem
but I decided to use Ulrich's code since it improves the whole code by making
it a lot simpler.
So, kudos to all of you.
This commit alters XRANDR (X Resize And Rotate) extension dependency,
which is expected to have at least version 1.4. Earlier old versions of
xrandr were supported, at least to some extent.
Do not try to be too smart which compilations need config.h, as most of
them will simply because of the config.h has information about system
capabilities.
a 'BidiString' holds both the logical content and the visual reordered
version of the content of a string. this helps to reduce the number of
calls to reorder the string before drawing it (as introduced in the patch
from Ken Bloom) and to be more consistent in menus and textboxes (drawing
cursors and underlining text).
on *bsd /bin/sh is not just a symlink to /bin/bash as on most linux's
but a real standalone shell. and it behaves differently from "bash -c"
behavior .. it doesnt exec the command given but waits till the command
finishes. as a result a lot of "rogue" a flying around. solution is now
( $SHELL or /bin/sh ) -c exec <cmd>
- Removed "antialias"-option completly, to enable/disable "antialias"
use either <fontname>:antialias=<bool> in the style or use
Xft.antialias: <bool> in your .Xdefaults
- Added new styleresources:
*.font.effect: <halo|shadow>
*.font.shadow.x : <int> - shadow x offset
*.font.shadow.y : <int> - shadow y offset
*.font.shadow.color : <color> - color of shadow
*.font.halo.color : <color> - color of halo
- Removed 'shadow' and 'halo' options from fontdefinitions:
!! Style authors have to update their styles !!
- Simplified XmbFontImp to not try all possible fonts to match locale
- Style authors may specify multiple fonts:
<font1>|<font2>|<font3>
if loading of font1 fails, fluxbox probes <font2>, etc. The last font is
"fixed". Hints for style authors:
- if xft tries to load a font it will _ALWAYS_ return a valid font,
it doesnt have to look like the one you want to have, read more
about it: http://fontconfig.org/fontconfig-user.html
- export XFT_DEBUG=1 before running fluxbox helps to see
which fonts are picked.
eg:
*.font: Verdana,Arial-12:antialias=true|-artwiz-snap-*-
if fluxbox is compiled with xft this will NEVER try to
load "-artwiz-snap-*-" since xft gives for sure a font,
most likely Verdana or Arial but noone knows. So, if
fluxbox has no xft support the first fontpattern fails
and fluxbox tries the next one, which might be successful.
if everything fails, it will use "fixed"
- Added caching of fonts, fonts are only loaded once.
- Fixed#1090902 (slow utf8 start)
if the xkb-extension is enabled and the user switches between his/her
keyboardlayouts fluxbox's keybhandling doesn't work well anymore because
xkeyevent.state contains also xkb-related flags and thus we have to handle
that with caution.
KeyUtils now contain 'isolateModifierMask()' to really work only on the
modifiers. why not as part of cleanMods() ? because the XLookupString return
false results, eg TextBox's would only print chars from the first
keyboardlayout.
