The suggested command 'echo $(tput smkx) /dev/tty' works fine.
However, setting enable-keypad in .inputrc has strange results. It
doesn't make a difference to the delete key functionality, but does
result in the 'k' key not being recognised... Also it wasn't clear
from the FAQ whether I
You should put enable-keypad On if you want Delete key works, but I have
tested it now, and you are right, it doesn't work properly and causes 'k'
key doesn't work. I think it was working, so maybe we have a regression. I
will try a git bisect and I will send the results here
I have done
On Mon, Feb 25, 2013 at 04:44:11AM +0400, Alexander Sedov wrote:
Commenting on this even further:
- tested whether st rightfully extracts selection contents. It does, as
debug output told me, so the problem is in nano.
Usually we try don't add code for fixing bugs in others programs.
On Sun, Feb 24, 2013 at 08:35:06AM +0100, Andreas Marschall wrote:
hello,
since I use alpine as mail cleint I noticed that when I hit the
del button while writing a mail to delete the text that's ahead it
doesn't work in st giving me an [ Unknown Command: ]. Doing the same
in urxvt works
You meant nano :P. Well, we are pasting text using keyboard emulation
(writing to fd). Keyboard sends \n, not \r. So it makes total sense.
Sorry, you are right, I mixed this mail with other (and like nano is the
editor of alpine it was confusing for me). Maybe you are right and we should
On Tue, 26 Feb 2013 18:27:14 +0100 Roberto E. Vargas Caballero
k...@shike2.com wrote:
So it is more logical that pasting send '\r' instead of '\n', but it is a
better place in selcopy
800 if(is_selected y sel.e.y)
801
1. Replaced strtok_r with strsep in strparse, because strsep can handle
zero-width fields (manpage says so) and because resulting code is
clearer.
As far as I know strsep is a BSD extension and it is not a POSIX function,
so I think we should look for other solution.
Best
TO
4.4BSD.
And the standard used is POSIX (I doubt C99 admit 'read', 'write' or
'gettimeofday'), although we never have decided what version of POSIX...
Best regards,
Roberto E. Vargas Caballero
Since a long‐time goal of suckless is to have its own better libc, is
this a selection which is done to find the best functions. Strsep seems
to be superior to strtok[_r], so I’m using it now.
Ok, I only hope that users of systems without strsep (for example solaris
On Thu, Mar 14, 2013 at 10:40:52PM +0100, Markus Teich wrote:
If you realy need a .gitignore, you can use your own locally:
just add
echo /.gitignore .gitignore
and the .gitignore file itself as well as everything else matching
the ignore patterns will not show up in git status anymore.
I'd say this qualifies for a move to config.h.
I might be back later with a patch for that.
Umm, no. If you want to use st with x, y, z programs, set this to
\r, if you want to use a, b, c programs, set \n is a terrible idea.
Best to figure out what is more correct, do that, and patch any
Yeah, it makes sense for pasting from st to st.
But pasting from, for one, surf to st suffers from ^J instead of ^M
(mcabber, nano, who knows what else), and pasting to file and, as
topicstarter states, to gvim is pain, too.
Yeah, the problem here is you will have problems in both cases, some
On Fri, Apr 12, 2013 at 12:28:43PM +0400, Alexander Sedov wrote:
Now, on selcopy, st generates traditional Unix newlines ('\n'), because
apparently that's what GUI applications are used to see. But some
terminal applications do not handle '\n' very well, so, when pasting to st,
line feeds are
+ while(--last = gp !(selected(last - gp, y) \
+ strcmp(last-c, ) != 0))
/* nothing */;
I think you could write here only:
while(--last = gp strcmp(last-c, ))
/*
On Tue, Apr 23, 2013 at 10:50:32AM +0200, Christoph Lohmann wrote:
Greetings.
On Tue, 23 Apr 2013 10:50:32 +0200 Otto Modinos ottomodi...@gmail.com wrote:
Howdy Comrades.
I'm new to this mailing list.
Just found st couple of days back and really loved it. It lacked however
the ability
On Tue, Apr 23, 2013 at 02:51:58PM +0300, Otto Modinos wrote:
It means first of all vim. I have also tried mocp, and that didn't work
either. What apps are you using that work?
All the applications which admit a meta-key should accept 8bit, that is the
real meaning of meta. Send ESC
No, it is the intended behaviour.
http://fsinfo.noone.org/~abe/typing-8bit.html
The fact that someone discovered it, _thought_ it was intended, and
showed other people how to do it does not mean that it actually was
intended.
It is the expected behaviour. As far as I know the first
The nice thing about ASCII and the QWERTY layout is that it has
all ASCII characters easily available, so you can enter *all*
latin1-subset-of-Unicode characters with just it.
Because all the latin1 symbols begin from a0, that is the smaller value you
can generate using a meta key, because 20
From: Roberto E. Vargas Caballero k...@shike2.com
The commit b78c5085f72 changed the st behaviour enabling BCE capability,
that means erase regions using background color. Problem comes when you
clear a region with a selection, because in this case the real mode of the
Glyph is not the value
On Tue, Apr 23, 2013 at 03:38:32PM -0400, random...@fastmail.us wrote:
On Tue, Apr 23, 2013, at 14:34, Roberto E. Vargas Caballero wrote:
From: Roberto E. Vargas Caballero k...@shike2.com
The commit b78c5085f72 changed the st behaviour enabling BCE capability,
that means erase regions
On Tue, Apr 23, 2013 at 05:12:47PM -0400, random...@fastmail.us wrote:
On Tue, Apr 23, 2013, at 17:05, Roberto E. Vargas Caballero wrote:
What _exactly_ is the behavior you are observing? Are you sure it's not
_actually_ staying selected, rather than simply drawing that way?
Did you see
Maybe the behavior is wrong - but if the problem is that it is
_still selected_ (i.e. hilight goes away when you select something
else), it's not something that can be solved with anything to do
with visual attributes only.
That was why I was asking for clarification whether it is _still
We can test it in tclearregion in this form:
for(y = y1; y = y2; y++) {
term.dirty[y] = 1;
for(x = x1; x = x2; x++) {
if(selected(y, x))
selclear(NULL);
term.line[y][x] =
Too late -- I already sent almost that exact patch :P.
Yes, I think this is the right thing to do, because st currently
clears selection on putting character inside it. So, changing a
characted with tclearregion() should clear it as well.
I can not see your patch, I only can see you send
@@ -1425,6 +1425,9 @@ tclearregion(int x1, int y1, int x2, int y2) {
for(x = x1; x = x2; x++) {
term.line[y][x] = term.c.attr;
memcpy(term.line[y][x].c, , 2);
+ if(selected(x, y)) {
+
From: Roberto E. Vargas Caballero k...@shike2.com
The commit b78c5085f72 changed the st behaviour enabling BCE capability,
that means erase regions using background color. Problem comes when you
clear a region with a selection, because in this case the real mode of the
Glyph is not the value
On Thu, Apr 25, 2013 at 12:38:02PM +0200, markus schnalke wrote:
[2013-04-24 17:14] random...@fastmail.us
A) Why do min and max need to be macros at all?
That's the point, IMO. All the problems are introduced by performance
optimizations that might not even necessary. If you write it as a
On Sat, Jun 22, 2013 at 11:39:19PM +0200, Markus Teich wrote:
And of course i only did this patch for
XK_KP_Delete and forgot about
XK_Delete.
Sorry but these patches are incorrects.
+ { XK_Delete,XK_ANY_MOD, \033[3~, 0,0,0},
{ XK_Delete,
From: Roberto E. Vargas Caballero k...@shike2.com
There were two problems with match denfinition.
1) There was a forward declaration in the form:
static inline bool match(uint, uint);
but later the function was defined as:
inline bool
match(uint mask, uint state
From: Roberto E. Vargas Caballero k...@shike2.com
draw is the function which update the Xwindow with the information st has,
and it is designed in a way that it must be called once in the main loop
(run function), and calling it in other places it is a waste of time.
---
st.c |1 -
1 file
From: Roberto E. Vargas Caballero k...@shike2.com
The way st knows if there is a selection activated is checking if sel.ob.x
is equal to -1. In some parts of the code the way of disabling the selection
was only setting it to -1, but after it you can't be sure if the selection
is clearing from
Uhmmm, the patch is not good because it breaks a lot of configurations (when
we use XK_ANY_MOD). I am going to send a new version of the patch.
On Sun, Jun 23, 2013 at 10:44:46AM +0200, Roberto E. Vargas Caballero wrote:
From: Roberto E. Vargas Caballero k...@shike2.com
There were two
From: Roberto E. Vargas Caballero k...@shike2.com
The way st knows if there is a selection activated is checking if sel.ob.x
is equal to -1. In some parts of the code the way of disabling the selection
was only setting it to -1, but after it you can't be sure if the selection
is clearing from
From: Roberto E. Vargas Caballero k...@shike2.com
There were two problems with match denfinition.
1) There was a forward declaration in the form:
static inline bool match(uint, uint);
but later the function was defined as:
inline bool
match(uint mask, uint state
Ignore this mail, it was error.
If on any system other than linux, I would consider loading into ram,
but because of memory overcommit, malloc never fails, the whole system
crawls to a halt, and the oom killer takes 20 minutes to put
everything back together. No thanks.
Okay so you are discussing a problem with your
Maybe a better solution could be using mmap.
Perhaps. I still wouldn't know how much to map, so I'd need
mremap, and from the manpage:
This call is Linux-specific, and should not be used
in programs intended to be portable.
If you write a portable patch making use of mmap,
Sponge's killer feature is that it doesn't open the output file until
after the input is finished. Using it in a pipeline removes this,
because it's something else instead of sponge writing to the file.
There is something I don't understand, why the temporal file is not deleted
after writing
Files created by “tmpfile” are automatically deleted after they are
closed. Also, how do you delete a file you don’t know the name of anyway?
I didn't know it, thanks for your reply.
I would suggest a subset of POSIX tar options. It could be possibly
amended with z, Z, j and J options, though using
A $ tar c dirname | gzip -9c filename.tar.gz
or
A $ tar cf - dirname | gzip -9c filename.tar.gz
I usually use:
gunzip file.tar.gz | tar xf
On Mon, Jul 08, 2013 at 09:47:10AM +0200, Kurt Van Dijck wrote:
Hello,
I just remembered this thing from an embedded 16bit CPU.
Why would you do
unsigned char XXX :1
and not
unsigned int XXX :1
I think the latter may benefit code size (a tiny bit)
because it avoids an
not a strong opinion though, just an idea.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
niche).
I agrre that this tool is useful only in corner cases, when the
number of files in the directory is huge, and usually the number
is very small, so it is not enough general for a coreutils package.
--
Roberto E. Vargas Caballero
k...@shike2.com
http
,
attr[i]);
}
break;
Best regards,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
sense or not. Do you know how radare2 uses it?.
Other related problem with this kind of sequences is that they are used
only if TERM=xterm-256 or something similar, because there isn't any
terminfo capability for them.
--
Roberto E. Vargas Caballero
k...@shike2.com
E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
From: Roberto E. Vargas Caballero k...@shike2.com
OpenBSD defines strlcpy function, and the declaration is a bit different
of the static declartion found in tar.c. The duplication of symbol name
with different type (one extern and other static, one returning int and
the other returning size_t
I've been testing this patch for some time, but not really sure if this
is stable enough. Also, this patch is base on 40e4d76 (the current HEAD)
It seems good for me, and I think we could add it to HEAD and it could
help others could test it to. If nobody says something about it, I
will add it
STRLCPY.
If we are going to use only in this file then maybe is better use the snprintf
directly.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
From: Roberto E. Vargas Caballero k...@shike2.com
It is impossible allocate all the RGB colors using a static
array, so it is necessary move DC.col into a pointer and use
dinamic memory.
Since the color definition is not used to much is not a bad idea
use realloc each time. It means the color
From: Roberto E. Vargas Caballero k...@shike2.com
OpenBSD defines strlcpy function, and the declaration is a bit different
of the static declartion found in tar.c. The duplication of symbol name
with different type (one extern and other static, one returning int and
the other returning size_t
returning size_t) caused tar couldn't compile in OpenBSD.
I have applied this. In the future, could you send your patches as
attachments? It would make my life easier.
Ok, but take a look of git am (it applies a patch from a mailbox).
--
Roberto E. Vargas Caballero
, 0, Blksiz-l);
fwrite(b, Blksiz, 1, tarfile);
}
One solution could be ignore the tar file (maybe using
realpath(3)), or call to stat and get the size of the file
before archiving it.
--
Roberto E. Vargas Caballero
k...@shike2.com
This patch fixes this issue using the inode number of the tar
file.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
From e5b989e7899c87f316597e6b8ed39adfa97959b2 Mon Sep 17 00:00:00 2001
From: Roberto E. Vargas Caballero k...@shike2.com
Date
numbers) for true colors, and wrapping colour
access in xdraws(). Much cleaner, doesn't touch heap.
I agree with you, your solution is far better than mine, it removes
the malloc and I think it will take less code. I will try implement
it this night.
Thank you.
--
Roberto E. Vargas Caballero
From: Roberto E. Vargas Caballero k...@shike2.com
This patch uses the bit 24 in the color descriptor as an indicator
of RGB color, so we can take the values and generating the XftColour
directly in xdraws.
---
st.c | 119 ---
1 file
The -std=c99 patch is to suppress warnings about using the %llu format
specifier. It's kept separate in case C99 is undesirable, though.
Since we are using long long in other place, I think the best
option is change the standard version to c99.
--
Roberto E. Vargas Caballero
value is a printable code.
This suggests that we should still add 32 to the button value when in
MODE_MOUSEX10.
It is no clear for me if we have to add it or not, did you test it with
other emulators?
--
Roberto E. Vargas Caballero
k...@shike2.com
http
On Thu, Jul 25, 2013 at 04:24:16PM -0700, Michael Forney wrote:
ATTR_BLINK is an attribute for a Glyph and will not be set in term.mode.
Good catch, I'll apply it.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
is running 'tput rmm' before of mc, or you can remove smm and rmm
in the terminfo definition.
Maybe could be a good idea adding a command line switch for removing the
8bit-meta
behaviour.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
prop);
XSetWMName(xw.dpy, xw.win, prop);
+ XFree(prop.value);
}
Good catch, I'll apply it.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
It seems good to me, if nobody says other thing I'll apply it next week.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
of tdefcolor instead.
Could you amend the patch and send it again?
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
When a line doesn't fit in a row of the terminal it is wrapped, so next
characters of the line are shown in the next row. st adds always a newline
character when rows are copied, but it causes that in wrapped lines
a non needed newline is appended. The solution is easy, don't add the
newline when
On Sat, Aug 10, 2013 at 04:15:49PM +0400, Dmitry Potapov wrote:
Thanks a lot for quick response.
Looks like this is already fixed in master. My bad, I haven't checked it
first.
You are right, my patch was unneeded.
Thanks.
--
Roberto E. Vargas Caballero
k
*ptr++ = '\n';
which is used only in the last line of the selection, so it is not related
to the problem of Dmitry Potapov. Like we have seen, the problem was
already fixed and it isn't necessary any change.
--
Roberto E. Vargas Caballero
Some programs don't check the value of km and use smm and rmm
capabilites, and they cause the terminal change to meta enabled
mode even in cases where is not desirable.
Allmost all people is using the terminal waiting that meta sends
escape, so rmm and smm are not needed. If someone needs meta
it, because he knows the font part better than me.
Thanks,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
+ while ((r = read(fd, c, 1) 0)) {
+ offset++;
+ if (isprint(c)) {
+ span++;
I think is better use the buffer interface here. Calling getc
will save a lot of system calls and it will improve the performace
of the application.
--
Roberto E
,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
], and it is the way that I usually use it.
[1] http://pubs.opengroup.org/onlinepubs/009604499/utilities/strings.html
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
;
}
} while (c != EOF);
if (ferror(fp))
eprintf(%s read error:, fname);
}
}
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
agree it is a bug. It seems
related to the bce capability but I have to look it better.
Best regards,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
.
Please let me know what you think about my proposed changes.
I only like the int - size_t in strlcpy change.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
it instead of
using other local variable for it). I use const only when I want
show that the function is going to accept constant string which
should not be modified, like for example:
printf(const char *fmt, ...)
--
Roberto E. Vargas Caballero
k...@shike2.com
http
I have applied the patch, thanks.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
After pull the patch I have problems with graphical characters (for example
lines and arrows in mutt). is there someone which could test it?
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
it stores
it in ~/.terminfo like is usual to. But I don't know why it seems
don't read the definitions in /usr/share/terminfo and ~/.terminfo,
and I am not be able of modify the default definiton.
Is there some way of fix this problem?
Regards,
--
Roberto E. Vargas Caballero
From: Roberto E. Vargas Caballero rvar...@bsccs347.bsc.es
\a is the character for bell, and st is only marking the window as urgent
if it is not active. This patch adds an audible bell which can be disable
with bellvolume variable.
---
config.def.h | 6 ++
st.c | 2 ++
2 files
.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
I'm fairly certain that we are supposed to always add 32 unless in SGR
mode. I've just now confirmed with the xterm (version 296) source code:
Ok, patch applied. Thanks!.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
Hi,
Hi,
I attach a simple patch to st, to enable bracketed paste mode (
http://invisible-island.net/xterm/ctlseqs/ctlseqs.html#Bracketed%20Paste%20Mode
).
Could you put an example when it is useful?
Best regards,
--
Roberto E. Vargas Caballero
See e.g.
http://stackoverflow.com/questions/5585129/pasting-code-into-terminal-window-into-vim-on-mac-os-x
(especially the 2nd comment)
hope that helps,
It seems there are some users that could be interested in this feature,
so I will apply it next week.
--
Roberto E. Vargas Caballero
of returning directly
the char representation.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
))
If you agree with this small difference I will commit your change.
Best regard.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
WebKitGtk+ usually means install a lot
of suck software, like for example dbus. This new port of WebKit
can reduce a lot the number of dependencies of surf.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
On Mon, Sep 23, 2013 at 12:07:06PM +0200, Maurice Quennet wrote:
Roberto E. Vargas Caballero k...@shike2.com wrote:
If you agree with this small difference I will commit your change.
I don't mind.
Applied.
--
Roberto E. Vargas Caballero
k...@shike2.com
http
to check compability with real terminals, so it
doesn't know about X11, window managers and these kind of things.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
we should try catch
what it is wrong in this test of vttest.
Best regards,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
of running this test without 24x80.24.
You applied two patch (one in my first mail and next in second mail)?
Yes, I have applied both patches.
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
I meant: 2. Test of screen features (last test of the section: SAVE/RESTORE
CURSOR feature). Be careful of running this test without 24x80.24.
Sorry, s/without/with/
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
vt100 has support for two defined charset, G0 and G1. Each charset
can be defined, but in each moment is selected only one of both
charset. This is usually used selecting a national charset in G0
and graphic charset in G1, so you can switch between graphic
charset and text charset without losing
I applied the patch to tip ( eeae9b0 ) but compilation fails due to
a redefinition of a the tsetchar function:
Sorry, I don't know why my gcc version doesn't give me the error (some
type of overloading). This version uses a different name for this
new function.
-- 8 --
Subject: [PATCH]
, XK_NO_MOD, \0, .len=1 },
I cannot see when could be interesting send a '\0'. It someone can give
a case when it is necessary we can think about this patch.
Isn't there strnlen() function?
This was the same I was thinking before of reading this new version,
but it is not necessary now.
--
Roberto E
I was thinking the same until I checked back on the ISO C11 specification and
- there's no strnlen function :D
Yes, it is true, but take a look to strnlen(3):
The strlen() and strnlen() functions conform to IEEE Std 1003.1-2008
(``POSIX.1'').
;).
--
Roberto E. Vargas Caballero
On Wed, Oct 02, 2013 at 09:36:27AM +0200, Mark Edgar wrote:
On Tue, 1 Oct 2013 23:04:16 +0200, Roberto E. Vargas Caballero wrote:
I agree here, good catch and good implementation. Could you send
a mail with the patch and a proper commit message?
No problem!
I think we could avoid some
all the things
we write to the tty with ttywrite. Let's me test it with some emulators and
with my real terminal. I will send the results.
Best regards,
--
Roberto E. Vargas Caballero
k...@shike2.com
http://www.shike2.com
101 - 200 of 700 matches
Mail list logo
