This patch adds the flags -t and -T that can be used to toggle
multi-selection via Ctrl+Enter. In its current form, it must be applied
on top of the other two patches I sent:
- Add flag to automatically print unambiguous query
- Make Ctrl+Enter print current selection
Eric
>From 9bd22dfeb39ad88a4
On Sun, Apr 12, 2015 at 02:37:41PM -0700, Eric Pruitt wrote:
> On Sun, Apr 12, 2015 at 11:36:19PM +0200, Markus Teich wrote:
> > However I use dmenu in some cases, where I assume, that stdout of
> > dmenu is just one line, so I think another commandline flag would be
> > nice to specify if holding
On Sun, Apr 12, 2015 at 11:36:19PM +0200, Markus Teich wrote:
> However I use dmenu in some cases, where I assume, that stdout of
> dmenu is just one line, so I think another commandline flag would be
> nice to specify if holding ctrl keeps dmenu open for more input when
> pressing return.
Why sho
Eric Pruitt wrote:
> I am using a clean build from git with no patches, and Ctrl+Enter doesn't do
> anything without my changes.
Heyho,
For me ctrl+enter as well as ctrl+shift+enter also do nothing when running:
echo -e 'bla\nblub' | dmenu
(current master)
dmenu stays open and does not
On Sun, Apr 12, 2015 at 07:14:54PM +, non...@inventati.org wrote:
> Try to clone a clean git repo from git.suckless.org. For me everything
> works as expected without any patches.
I am using a clean build from git with no patches, and Ctrl+Enter
doesn't do anything without my changes.
Eric
Try to clone a clean git repo from git.suckless.org. For me everything
works as expected without any patches.
On Sun, Apr 12, 2015 at 11:53:30AM -0700, Eric Pruitt wrote:
> It appears the author wanted to make Ctrl+(Shift)+Enter print the
> current selection without making dmenu exit, and this patch changes the
> input handling so that actually works.
The original patch had a typo in it. The fixed version
It appears the author wanted to make Ctrl+(Shift)+Enter print the
current selection without making dmenu exit, and this patch changes the
input handling so that actually works.
Eric
>From 295ab176ff86c79b309150d463f6510447382539 Mon Sep 17 00:00:00 2001
From: Eric Pruitt
Date: Sun, 12 Apr 2015 11
This patch adds a "-a" flag which causes dmenu to immediately print and
exit if only one entry matches the current query. This feature is simple
and unintrusive, and I'd like to see it committed to the master
repository. If it gets rejected, I'll push the patch to the site
instead.
Eric
>From 8ee1
On Sun, Apr 12, 2015 at 06:32:13PM +, non...@inventati.org wrote:
> The code that actually prints something is
> puts((sel && !(ev->state & ShiftMask)) ? sel->text : text);
Yeah, I saw that line, but the conditional I mentioned determines
whether or not the program actually terminates. The w
On Sun, Apr 12, 2015 at 11:18:02AM -0700, Eric Pruitt wrote:
> While looking at the dmenu source code to make a patch for it, I noticed
> this conditional on line 371: "if(!(ev->state & ControlMask))". It looks
> as though pressing Ctrl+(Shift)+Enter is supposed to print the current
> selection to
While looking at the dmenu source code to make a patch for it, I noticed
this conditional on line 371: "if(!(ev->state & ControlMask))". It looks
as though pressing Ctrl+(Shift)+Enter is supposed to print the current
selection to stdout without terminating dmenu, but pressing
Ctrl+(Shift)+Enter doe
Hello everyone,
Most webmails out there are messy stuff written in PHP and with lots
of dependencies. Memory-hungry monster filled with bugs.
After lots of research I found one small webmail written in 100% C by
some University of Cambridge professors. It's called Prayer.
http://www-uxsup.csx.ca
On Sun, 12 Apr 2015 11:36:00 +0200
koneu wrote:
> It is not a nop though, but two copy operations. Even with i=0 the
> function call might be more costly than a simple je.
The assembly is rather irrelevant in this case. Let's write the code
the way it is most understandable and clear.
I'd remove
On Sun, Apr 12, 2015 at 11:35:42AM +0200, Silvan Jegen wrote:
> I just wonder if it really will be copying the data around into a
> temporary array for no reason when i == 0 (either in glibc or in another
> libc).
Well, it does not really move data to temporary array, memmove just
chooses the righ
On April 12, 2015 10:56:40 AM CEST, non...@inventati.org wrote:
>On Sun, Apr 12, 2015 at 10:41:36AM +0200, Silvan Jegen wrote:
>> I was thinking about this option too but in that case you would be
>> calling memmove with an identical src and dst when i == 0. The man
>page
>> for glibc memmove(3) do
Heyhey
On Sun, Apr 12, 2015 at 08:56:40AM +, non...@inventati.org wrote:
> On Sun, Apr 12, 2015 at 10:41:36AM +0200, Silvan Jegen wrote:
> > I was thinking about this option too but in that case you would be
> > calling memmove with an identical src and dst when i == 0. The man page
> > for gl
On Sun, Apr 12, 2015 at 10:41:36AM +0200, Silvan Jegen wrote:
> I was thinking about this option too but in that case you would be
> calling memmove with an identical src and dst when i == 0. The man page
> for glibc memmove(3) does not mention anything about this being a noop
> so I am not sure th
On Sun, Apr 12, 2015 at 08:29:55AM +, noname wrote:
> ---
> st.c | 6 ++
> 1 file changed, 2 insertions(+), 4 deletions(-)
>
> diff --git a/st.c b/st.c
> index 109122e..f183803 100644
> --- a/st.c
> +++ b/st.c
> @@ -2788,10 +2788,8 @@ tresize(int col, int row) {
> free(term.
---
st.c | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/st.c b/st.c
index 109122e..f183803 100644
--- a/st.c
+++ b/st.c
@@ -2788,10 +2788,8 @@ tresize(int col, int row) {
free(term.line[i]);
free(term.alt[i]);
}
- if(i > 0) {
Sorry, forgot to include my work so far.
Ciao,
Markus
>From 96c65781f6d0d7b4f914156f520efe86ce193975 Mon Sep 17 00:00:00 2001
From: Markus Wichmann
Date: Sat, 11 Apr 2015 21:21:34 +0200
Subject: [PATCH] Implement most ICCCM rules for selection handling.
ICCCM mandates the use of real timestamps
21 matches
Mail list logo
